Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror
Hardware

3Dfx No More -- NVidia Purchases Video Card Maker 320

Posted by timothy
from the alas-we-hardly-knew-ye dept.
Julius X writes: "This just came out, from Yahoo, 3dfx has announced that they will be sold to NVidia as soon as the deal is approved by its shareholders. From the release, "After aggressively pursuing a wide range of options that take into consideration the interests of our creditors, our shareholders, our employees and our customers," said Alex Leupp, president and CEO, 3dfx Interactive Inc., "we strongly believe that to reduce expenses, sell our assets and dissolve the company provides the highest return to our creditors, shareholders, and employees." I think we all saw this one coming. For more details, go to the press release." Actually, tossing in some details early is [hk]doogie, who writes: "Nvidia bought the patents, pending patent applications, trademarks, brand names, and chip inventory related to the graphics business of 3dfx. Get the full scoop [here]."
This discussion has been archived. No new comments can be posted.

3Dfx No More -- NVidia Purchases Video Card Maker

Comments Filter:
  • Don't laugh too much. One of ATI's big advantages is that they already have an "SLI"-type technology. Although the Radeon is a bit slower than the GeForce 2 (original, not ultra) it has some nice extra features, so they're roughly comparable. But with the ability to put two chips on one board, ATI should have been able to keep up easily, at least for a while.

    Now, though, nVidia has access to 3dfx's SLI technology. Dual Geforce 2 Ultras aren't very likely, but how about 2 (or 4?) Geforce 2 MX's on a card? It'll take some time before they can integrate SLI into their present products, maybe it's not even possible, but maybe this buyout will take away one of ATI's big advantages.
  • We don't want their drivers, we want the specs to their cards so others can make independent drivers. Unfortunately, their driver developers have equated NVidia's driver development with their job security, and don't see releasing specs as welcoming competition (never mind that quite afew of open source projects that were started by companies are still worked on by paid in-house developers).
  • Something breaks here...

    Underdog bests incumbent, knocks them silly, outmaneuvers them, and then destroys them, finally purchasing all the relevent patents and technologies:

    Nvidia == Underdog
    3dFX == Incumbent

    AMD == Underdog
    Intel == Incumbent

    Microsoft == Incumbent
    Apple == Underdog

    I suspect I have your pattern wrong; what pattern causes Microsoft buying Apple insightful, instead of confusing. Apple hasn't been an incumbant, unless you're counting the days of Apple II...

    Your VA Linux crack should give some insight, but all I could glean was that an upstart who has no technical prowess is able to purchase VA Linux, who will probably crumble in a way analgous to 3dFX...

    The only thing that comes to mind is Slashdot's ties to VA Linux(whatever they are), and that a kid who delivers newspapers are somehow... more relevant than Slashdot?

    I'm sorry, I know asking about a joke will often kill the humor.

    Geek dating! [bunnyhop.com]
  • ...and see releasing specs ...

    blah, gotta hit that preview button more often
  • If this does lengthen the "ultimate card->crap worthless card" video cycle, it will be somewhat better for consumers, in my view. For too long, game companies have been babied by these 6 month card releases, and never have to compete in the gameplay arena, or on price. They just try to be first out the door on the new cards so they can put out hotter screenshots and bigger polygon counts in the game fan 'zines. Slower graphics chip cycles will also lead to better, more mature drivers before the driver team moves on to the newest chip.

    The best thing that could happen for regular, non-bleeding-edge consumers like myself is for the few remaining companies to raise prices on the newest cards by $50 or so, and have them reign for a year, getting down to current new release prices at around 4-6 months.

  • NVidia IMHO makes the best current 3d hardware, but they have nothing in the business/SOHO/laptop/OEM market that I'm aware of, whereas Matrox and ATI have vast sums of revenue from those markets.
    >>>>>>>>>>>>>>>
    Where have you been the past 6 months. Take a look at almost every PC sold at CompUSA, go to Dell.com and look at their PCs, try Micron, Gateway, etc. They all use NVIDIA cards. There used to be a time when ATI owned this segment (RagePro) but these days, most OEM/cheap PCs come with integrated TNT2-VANTA class processors. ATI is starting to make a little bit of a comeback here since Dell recently started using the Radeon on some of its machines, but NVIDIA owns the market right now. Neomagic and ATI still rule the notebook roost, however, but there is no word on how long that will last. The reason NVIDIA is dominating is simple; their price/performance kicks ass. NVIDIA cards are cheap and fast. A Radeon DDR is usually about the same price as a GF2, but the GF2 is faster. There is only one weak segment in NVIDIA's line, that is the GeForce2 MX vs Radeon 32 SDR (or DDR for a slightly uneven match)The Radeon is only slightly more expensive, but much more powerful.
  • That's what Aureal said about their Vortex drivers, just before they went bankrupt and bugs that could only be resolved in the binary portion were found with VIA based Athlon motherboards. They too were bought by their competition, Creative Labs, but inversely creative had released open source drivers for their competing cards. hopefully they'll give the same treatment to the vortex driver once the acquisition's complete, but I gave up waiting and got a SBLive!
  • Does anyone else think 3Dfx's naming conventions are completely stupid? There would be the model, which is a number, and the product, which is another number. Voodoo3 2000, 3000, 3500, V4, V5 5500, V5 6000...
  • by FFFish (7567) on Friday December 15, 2000 @02:04PM (#555525) Homepage
    I gotta really apologise to y'all.

    I went and bought a V3-2000 videocard last weekend. Finally gave up on trying to pump Unreal through 4-year-old technology.

    And I really, really should have alerted the world about my purchase.

    You see, this sort of thing happens to me on a regular basis.

    Call it the Purchase of Deth syndrome. The reverse Midas touch. With friends like me, what company needs competition?

    Needed a sound card. Picked out Gravis as the best. Company went under a few weeks later.

    Needed a video card. Picked out a Diamond Monster. Company quit the video business shortly after.

    Needed a sound card upgrade. Picked Aureal A3D. It shut its doors a few months later.

    Needed a new video card. Picked out a 3DFX Voodoo3-2000. Bang, within ten days, they fold.

    Tell you what... I'll make up for all that.

    I'm off to purchase some Microsoft products. Hah! That'll teach the bastards...

    --
  • This is too bad. This is also the reason why getting an IPO isn't the best idea. At anytime the shareholders can hold a meeting and liquidate your assets. Nvidia's code has been buggy. I know forst hand, my son's TNT2 Riva never worked right. Glide problems, OpenGL problems, the promise of new drivers to fix the problems, and eventually the card is no longer supported and the new drivers never materialize. Nvidia sucks... 3DFX was great, and the code was pretty solid. They will be missed.
  • by Nailer (69468) on Friday December 15, 2000 @02:06PM (#555528)
    ATI has long had a reputation of putting out awful drivers. If they manage to put out decent drivers and improve their technical support, then they might be a worthy contender.

    Bzzt! Thanks for playing! As of two weeks ago, Rage 128 and Rage 128 Pros running OpenGL under Linux Quake 3 [as an example] now generally outperform Windows, thanks to drivers ATI commissioned from Precision Insight. Download them from ATIs site.
  • by 2nd Post! (213333)
    You forget ATI
    Oh, and Matrox

    They aren't dead yet.

    I think ATI still has a very good presence in the market, so NVIDIA still has a lot of fighting (and thus competition) to do.

    Geek dating! [bunnyhop.com]
  • I suspect the market will narrow as it becomes a footrace towards efficiency, refinements, and generational improvements.

    NVIDIA and ATI will settle into the top 2, with Matrox hovering around the edge...

    Some no-name will come up with something stellar and exciting, 2 years from now (not BitBoys), and knock some excitement into the display adaptor market, until ATI or NVIDIA catch up, 2 years later, giving said competitor 2 years to build itself up to a frenzy... then a third competitor will jump in, with a further refinement, and perhaps topple ATI in the process... then there will be a competition between the incumbent NVIDIA, the newly grown Radical, and freshly fed Upstart+ATI...

    Something like what happened just two years ago, when 3dfx bowled everyone over (s3, Rendition, ATI, and Matrox)

    It's just business, as usual.

    Geek dating! [bunnyhop.com]
  • Holy crap, I never saw this coming. I've owned 3Dfx cards for years, and their hardware kicks ass - not necessarily performance-wise, but reliability, and compatibility. Drivers were never an issue (as opposed to some other card maker).

    I currently have a V3 3K ... and it runs sweet on Linux. I'm not about to change it for any NVidia product anytime soon, till someone can prove to me that NV works as well as 3Dfx.

  • There's a definite cycle in this part of the biz - companies come and go - I think it's mainly because the product life cycles are short compared to their design times (ie the chips they depend on) - this makes designing the best chip/card a pretty hit-and-miss operation - between the time you're commited to a silicon architecture and those chips are in boxes on the shelves the whole world can change around you - IMHO (having designed graphics accelerators in the past) there's a lot of luck involved - make the right guess and you're on top, miss it by even a little and you're toast - and in this biz you don't get to screw up twice.

    As an example at a previous employer many many years ago we once bet the company on a [then] new and untested chip packaging technology - it worked and we had an accelerator design that walked all over the competition for almost 2 generations - made over $100M in sales off of it - but management wouldn't spend the money to do the short term re-engineering to keep our lead and we were toast - and by the time they figured it out it was of course too late ....

    I think that in the long run NVDA and ATI have more to worry about from Intel than anyone else (Intel's 810 is already hurting them both) - they now own the largest pieces of silicon in a PC outside of Intel's control - luckily for them Intel has already been burned by trying to go the graphics route and may be somewhat reluctant (just talk to the C&T people who were absorbed by the iBorg ...)

  • A slowdown in development cycle can have benefits elsewhere. If the hardware development slows down, then software development can ramp up knowing that they will have a longer window of viability before hardware improvements makes older software obsolete. Also, a more stable hardware environment benefits open source and alternative operating systems, who often get frozen out of hardware improvements and lag behind Windows in terms of driver support.
  • Am I the only one that finds it unsettling that:
    • Once there was 3dfx and NVidia, but the latter swallowed the former
    • Once there was Creative Labs and Aureal, but the former swallowed the latter

    I own a Voodoo 2 (Guillemot), Voodoo 3, and Vortex 2 (Diamond). I find it quite amusing:

    • Aureal marketed A3D, an essentially proprietary API. It sued Creative Labs for patent infringement, with mixed results. Creative Labs bought Aureal, including all the patents it allegedly infringed.
    • 3dfx marketed Glide, a proprietary API (going so far as to sue Glide wrapper developers, including Creative Labs). It sued NVIDIA for patent infringement, with mixed results. NVIDIA is buying 3dfx, including all of its patents.

    Some New Year's resolutions:

    • Micron: buy Rambus
    • Barnes & Noble: buy Amazon
    • AMD or VIA: buy Intel
    • Prodigy: buy BT and/or Unisys

    Moral: developers are cheaper and more effective than lawyers.

  • Actually, with closed source drivers, you are usually guaranteed it will get worse.
    What if Nvidia decided they would not support the 2.4 kernel? glibc versions >2.2? Xfree 4.1? With closed source drivers, all work comes from paid employees - think of it not as them deciding NOT to support these, but deciding not to expand their support to these new technologies.
    If I want to try BSD/Hurd, as far as I know I am up sh*t creek as far as support with an NVidia board. Same with trying to get it working under LinuxPPC.

    With closed-source drivers, every new technology amounts to 'expand support budget, eliminate support for old technology, or ignore'. With open-source drivers, at least people who want say, their Riva 128 to work with new drivers actually *can* implement the support.
  • Oh! I remember you! Well, I remember seeing your name on various bits of software that I ran on my (lost, lamented) Amiga 1000.

    You did Llamatron, didn't you?

  • This is very bad news for Mac users. For quite some time now, 3Dfx has been the only one putting out a combination of decent cards, drivers, and support for Macs at a reasonable price (ATI skimps on drivers, Matrox skimps on support, Formac is way too expensive even with the neato 3D glasses, etc). There are rumors of NVidia supporting Macs, but those have been around for nearly two years and nothing has surfaced from them (plus there have been announcements by NVidia about their commitment to a single-platform environment).

    So to say the least, I'm very dismayed by this one. Here's hoping NVidia will finally deliver the support they've been promising. If not, the future of 3D on the Mac looks rather bleak.

    Besides which, while NVidia was known for delivering excellent framerates, the renderer itself is also known for having the worst quality of The Big Three if taken on a frame-by-frame basis (ATI tends to come first, which may be part of -if not most of- the reason for the framerate problems their cards tend to have).


    ----------
  • I couldn't really care about Direct3D compatibility, as I use 100% Linux on my machine. What was important to me was a stable X server and excellent 3D rendering, both of which I obtained when choosing the V3 3000. I still enjoy it on a daily basis, as Q3A, UT and SoF run beautifully on Linux / V3 3000.

  • I'll put in my 2 quatloos re: point 1 -
    My Matrox G200 (and my old 1995 vintage Matrox Millenium) can drive my 21" monitor cleaner at the higher resolutions in 2D (ghosting, edges, etc) than my TNT. Comparing the G400Max and the GTS, there's the same comparison in 2D quality, which is where I spend most of my time (3D /. ?!) In 3D, the image quality with matching settings comes through sharper on the Matrox, but that GTS is a killer for speed, though Triple Play and Madden don't really cause nearly as much pain as Q3.

    That, and it took the other video card makers until late 1998 to make a card that performed better than the "old" 4MB Millenium at 1024x768/32bpp and better... shouldn't have been a problem, but 3D was the focus. Oh well... now I'm just into my crotchety old man phase again (once you hit 23, it's all downhill ;-)
    --
  • Does anyone else think 3Dfx's naming conventions are completely stupid?

    It's pretty annoying, but not as bad as ATI's ("New ATI Rage Fury Anger Pissed Hate MAX PRO 2000"....)

    I actually do think ATI makes good all-around cards (I'm using one of the OEM R128 cards who's specific confusing name I forget right now, and it works just fine), but the names are rather obnoxious in my opinion.


    A vote for the lesser of two evils is still a vote for Evil.
  • There's a very real chance that the NDA issue WILL go away with the 3dfx purchase. I may just be blowing smoke here, but it seems to me that, because 3dfx comes with fullblown OGL drivers, it's not outside the realm of possibility that NVidia will adopt them and stop paying SGI.
    --
  • I dont know what kind of problems YOU are having, but nvidia's drivers KICK ASS, and my GeForce2 can destroy any 3dfx card under Linux. I dont care if I dont have full and open source, just that the drivers work without problems, and for me, they have.

    Sure, I'd like to see more openness from video/sound/other card companies but I cant see this purchase of 3dfx as a step backwards. It can and will only get better.

    siri

  • 3dfx basically took a Voodoo 2 and added the Banshee extentions to it, which NO ONE used, so it's now basically an overpriced Voodoo 2.

    It's not even that. It only has a single TMU, so performance in multitexturing games (i.e. all of them) is worse than V2.

    The Banshee was yet another product showcasing 3dfx's utter lack of desire to improve their technology. They stunned the world with their amazing original Voodoo Graphics chipset, and then coasted right up 'till today. They incrementally improved the original (adding 2D, adding 32 bit color) ONLY long after the rest of the industry forced them to realize that these features were required.

    3dfx arguably gave birth to the consumer 3D industry. But after delivery, they sure did a lousy job of rearing their child.

  • by Guppy (12314) on Friday December 15, 2000 @07:35PM (#555588)
    Nvidia == Underdog
    3dFX == Incumbent

    As a long-time investor in 3D graphics stocks (Unfortunately for me, 3dfx mostly), I'd have to say this is not quite right. At the start of the 3D graphics industry, the order would have been something like this:

    ATI, S3 == Incumbents
    3dfx, nVidia == Underdogs

    Yes, us geeks derided the Virges and Rage boards as crap, but they sold to the people that mattered (OEMs, corporate sales), and as a result both ATI and S3 grew absolutely huge. Either of them alone dwarfed nVidia and 3dfx combined in almost every measure -- market cap, revenue, profit, units sold, overall market share, etc... The only important exception was probably in the (tiny) retail market, where 3dfx was #1 for a while. 3dfx was only the leader in performance (for a time) and mindshare--important to geeks but not to suits.

    As the market evolved, S3 faltered in their transition from the Virge to Savage chipsets. nVidia began to take away big chunks of ATI's core business (OEM sales). 3dfx manages to hold onto a slim lead in retail and brand name recognition, but the retail market is too small to support the company, and they begin to struggle financially. The field changes into something like this:

    Incumbent == ATI, nVidia
    Underdogs == 3dfx
    Has been == S3

    Then, nVidia went for ATI's jungular. The GeForce was originally a high-end, low volume part, but the MX version is making big inroads into both the retail and OEM market. Now, nVidia is about to release both mobile laptop chipsets (Where ATI currently is dominant, having wrested control away from Neomagic) and two versions of a motherboard chipset with integrated graphics. nVidia has continued to take business away from ATI, and is now on top of the pile by most important gauges. So, the modern order looks something like this:

    Incumbent == nVidia
    Underdog == ATI
    Has Been == S3, 3dfx

    Now, I've sort of neglected some players here. Matrox, Videologic, etc., but most of those have been niche players.
  • I remember the old days of certain swinging dicks in some online communities I frequented blathering on and on and pissing themselves talking about how 3dfx was always king and always would be. 3dfx this. 3dfx that. I guess you studs are chugging some monster crow now. Heh heh heh. Hope you don't mind if I chuckle just a bit more over that. Heh heh heh heh.
  • "It will be VERY interesting to see if NVidia continues 3dfx's commitment to the OSS arena."

    and in other news, NVidia announced today to continue support of the Voodoo line of video cards under linux, with a binary-only kernel module.
    --

  • "When they dissolve the company, the remaining assets (after liabilities) will be divided among the shareholders. You'll probably wind up with some NVidia stock."


    How much? Suppose it's 1/100th of a share of nVidia per share of TDFX?

    Again, details are spare and there is much room for interpretation of what little info we have. However, it appears that the shares of nVidia do not go to TDFX shareholders. If my interpretation is correct, they instead first go to 3dfx, which still exists as an "independent" company (Although stripped of everything of value), which apparently will use them to pay their creditors first. Then, whatever is left (Could end up being very little) will go to shareholders, possibly directly, or maybe as cash after liquidation. Probably the latter, as it appears the management will attempt to bail in their golden parachutes first.
  • There is not a single good reason from a software engineering point of view for NVIDIA to open source their drivers. Not one. Software engineering is not a religion, it is an engineering discipline. General users of an operating system should not need to have the source code for their drivers. NVIDIA is more than capable of producing very good drivers and distributing the source code would only give its competition valuable information and algorithms. When you have such a high technological edge with drivers (just look how bad ATIs are), why give it away? Why give millions of dollars of research away to competition for free? Why should I, as ATI, even *bother* with driver research and development if NVIDIA, the current market leader, *gives* it away? I mean really, think a little bit...get off the pathetic and completely useless tangent of NVIDIA or other big software companies releasing their software under open source. NVIDIA is supporting linux, and quite well it seems from Tom's Hardware benchmarks, they are quite close to windows framerates with the only reason that's holding them back being XFree86. The GNU movement is all about re-inventing the wheel. You said you fear for the 3d revolution because of NVIDIA's closed source drivers under linux. That is so blatantly narrow view. You do not take *any* factors into account. NVIDIA has continually broken 3d barriers with their hardware, they have continually produced excellent products with exceptional drivers. It appears you comments were not based on anything rational...
  • by Galvatron (115029) on Friday December 15, 2000 @02:38PM (#555599)
    I think what most people are saying is not that this is some horrible disaster, but undeniably, competition is good. Even if 3dfx makes an inferior product, merely by existing they force everyone else to work to keep ahead of them.

    On top of that, as others have pointed out, 3dfx is much better about open drivers than Nvidia. It's the same reason most of us want Netscape to triumph over MS, because mozilla is open source, and IE will go open source sometime around when Hell freezes over.

    Yes, this is the natural result of 3dfx not doing a good job. Yes, there are some other graphics card makers out there. However, there aren't a lot of them, and Nvidia may be able to use their market position to drive up the price of cards, which is obviously bad for all of us. What do you expect people to say? "Hooray, now I won't be as confused by all the choices next time I buy a video card?" This may not be horrible, but there's really nothing good about it either.

  • by jmcmurry (3759) on Friday December 15, 2000 @04:21PM (#555601) Homepage

    Wait a second...

    But what if NVIDIA tomorrow does an "ID Software" and says that they won't support Linux in the future, due to that niche being to small?

    This is a misleading statement. id said they wouldn't provide a separate Linux-only CD-ROM for purchase at your local software outlet. They'll still be supporting Linux through downloadable binaries. And if you think your Q3TA CD will be good for anything besides reinstalling graphics and models after the first patch to the binaries, you should think again.

    Sure, nVidia should allow open source developers to see specs on their hardware so free like speech drivers can be developed. You have a great point with a non-great example.

  • by _|()|\| (159991) on Friday December 15, 2000 @02:44PM (#555603)
    it's quite possible that [NVIDIA] will continue to support the open effort of current 3dfx cards

    I doubt it. NVIDIA's Q&A [nvidia.com] says, "The structure of this deal allows NVIDIA to purchase certain assets that are consistent with our business model without acquiring 3dfx liabilities." One of these liabilities, apparently, is the current Voodoo line: "The remaining 3dfx entity is responsible for their current product lines and retail channel. ... The 3dfx product in the channel and installed base and customer support remain the responsibility of 3dfx. It is best to check those details with 3dfx management."

    After this deal, there isn't going to be a "remaining 3dfx entity." According to the 3dfx press release, the "board of directors will recommend to its shareholders that they ... approve a plan to dissolve the company following completion of the asset sale." The creditors will lick the plate clean, any remaining employees will be fired, and Nasdaq will remove TDFX from the ticker.

  • by ewhac (5844) on Friday December 15, 2000 @02:46PM (#555605) Homepage Journal

    You think you've got it bad...

    Once upon a time, I fell in love with the Amiga computer. I bought one as quickly as I could, and was happily hacking on it for years. I thought it was a really neat system, and it died.

    Roughly parallel to that, I got to work on CDTV, which was a "consumerized" version of an Amiga 500, intended to directly compete with Philip's CDI. I helped create what is still probably one of the best CD audio players ever done for a "home" gaming/multimedia system. I though it was a really neat system, and it died.

    After that, I was fortunate enough to be invited by RJ Mical and Dave Needle to join NTG (New Technologies Group) who were working on what was to become the 3DO Interactive Multiplayer. It was based on the ARM-60, had ass-kicking graphics and sound capabilities, and a nice tiny OS that owed much of its heritage to the Amiga. I thought it was a really neat system, and it died.

    Undeterred, we went on to design M2, the 64-bit follow-on to the 3DO Multiplayer. This thing had a 3D chip that did 32-bit rendering and outperformed 3Dfx's PC offerings at the time. It was also slated to have two 66MHz PowerPC 602 chips running the show. I thought it was going to be a really neat system, and it died.

    I now work for Be, Incorporated... And I think it's a really neat system.

    Schwab

  • They deliver a solid product, good drivers, and are open source sympathetic.

    I have to say I disagree with every point you just mentioned. I purchased a G400 Dual Head card last year, and it fried its own BIOS when I tried to install the Win2k drivers that were available at the time. The Win98 drivers that I tried were dog slow and had some weird quirks (such a graphic artifacting with certain Direct3D games that always worked fine with the Voodoo3 I bought afterward). And frankly, most consumers don't give a rat's ass about open-source sympathy or not, so I consider that a non-issue when it comes to predicting the successs of a video board company.

  • Forgot to mention something--nVidia has agreed to pay $70 million plus 1 million in nVidia stock. Now, assuming all of that stock goes to TDFX shareholders (The amount of debt TDFX has is still unclear to me at this moment, so I'm unsure how many shares will have to be liquidated to pay creditors), there are 39.4 million shares of TDFX outstanding. So let's assume 1 share nVidia for every 40 you own of TDFX. That's still a hell of a lot of suckage for TDFX investors, as it currently puts my 3dfx stock at less than $1 in value.
  • Is there anyone really left that can compete now in the video card market? Unless Matrox pulls something fantastic out of their hats, I wonder what's going to happen, especially with prices.
  • Actually, the GeForce2-class chips already blows away the Oxygen series in terms of performance. If you take a look at the Intense3D benchmarks [intense3d.com] you'll find that the WildCat (designed by Intense3D, acquired by Intergraph, then acquired by 3Dlabs) is the fastest midrange workstation card out there. However, take a look at the Elsa Gloria II scores. The Gloria II is almost exactly a GeForce running at 130Mhz. (10MHz overclock) with some anti-aliasing and other features enabled. Its already around 50-70% of the performance of the WildCat, and I wouldn't be surprised if a GeForce2 Ultra comes within 20% of the performance of a WildCat. At around 1/4 the cost. While 3DLabs might have some tricks up their sleeves, they better get those tricks out fast, because NVIDIA is not only taking over consumer space, but has a great chance at the workstation market as well. As for Matrox, you have to respect them. The G400 MAX was a little late, but had the best visual quality of any card (and still does) and was most of the speed of a TNT2 Ultra.
  • by itchytr0n (189250) on Friday December 15, 2000 @12:56PM (#555618)
    Hey! We're just gonna sell chips [slashdot.org] now!

    No wait, sorry, we're still gonna sell cards [slashdot.org]!

    Ah screw it, we suck, let NVIDIA have it all.

    And I thought the wife's moodswings were bad. Sheesh.
  • by Devil (16134) on Saturday December 16, 2000 @12:42PM (#555620) Homepage
    A lot of people (especially shareholders and Voodoo_ owners) are discussing the sale, but there are a number of causes for this, not just one or two. Here's how history worked for, and ultimately against, 3Dfx. Apologies if I don't get the timeline exactly right:

    1. In the beginning... 3D graphics chips were *really* expensive, running to the thousands of dollars and only working on certain hardware.
    2. Then simple 3D graphics chips like S3's ViRGE (shudder) came along. They were slow and had a horrible framerate (the ViRGE is commonly referred to as the world's only 3D decelerator), but actually made the pictures look a bit more smooth. This trend continued for a while, until...
    3. 3Dfx released the Voodoo, ushering in really nice second-generation hardware acceleration. With high-res, high-color-depth and smooth framerates, they became the overnight champ. They scared the daylights out of companies like Number 9, S3, ATI and all the other 3D chipmakers. The only card these companies had was that the Voodoo was a secondary graphics card; it didn't do 2D. The other problem was that the Voodoo really only worked with Glide; most other graphics chips could work with OpenGL.
    4. 3Dfx tried to fix this problem with the Voodoo Rush, which was essentailly a 2D chip hot-glued to a Voodoo. The Rush failed miserably (First slip: Don't cobble together a half-assed product. Gamers and reviewers always know the difference.), because 3Dfx did not take into proper account the interaction and integration needed to make the 2D and 3D work together. 3Dfx then released the Voodoo2 (essentially a sped-up Voodoo), to much critical acclaim. Still, Glide was the only graphics system that it really worked with, and 3Dfx was convinced that Glide would eventually replace OpenGL. (Second slip: Hubris has brough down more empires than one can imagine. Always know that you are mortal.)
    5. Around this time nVidia was starting to show signs of becoming a good chipmaker. While early chips like the NV1 were absolutely laughable when compared to the mighty 3Dfx chips, nVidia was busy trying to prove that they were fast learners (the Riva 128 was considered a great chip for gamers on a tight budget).
      Also around this time, most other 3D chipmakers (like Number 9, PowerVR and S3) were frantically trying to stay in business. A couple of them made it, but most either quit the 3D market or fizzled. 3Dfx was riding high, and went public amidst the dot-com market inflation. Things were looking good.
    6. The release of the TNT was the first real threat to 3Dfx's business; it was fast, ran OpenGL & Direct3D (at the time a horrible choice for anyone, but there nonetheless) and also did 2D. At first 3Dfx tried to convince itself (and us) that the future would be powered by Voodoo/Glide, but when reviews of TNT chips essentially matched the Voodoo2, 3Dfx knew it was in for a fight to the death.
    7. 3Dfx had to do something; people were buying TNT cards, and that spelled trouble. 3Dfx ultimately decided that their best (only?) option was to do it all: the chip and the board. They bought STB (right after I bought my PC which had an STB graphics card and a TNT chip -- doh!) and stopped selling Voodoo technology to third parties, most of whom immediately jumped into bed with nVidia (remember how pissed off Creative was?). The Voodoo3 would come from only one source: 3Dfx. (Third slip: Don't alienate the very people who provide most of your revenue, i.e. boardmakers)
    8. As the Voodoo3 and TNT/TNT-Ultra waged war, gamers were slowly coming to realize that nVidia's chip was just as fast (or faster), on par with price, and didn't require the proprietary Glide. PC makers also realized this and started shifting towards nVidia graphics boards. And since they came from a variety of boardmakers, the PC makers could pick and choose their vendor while still using the nVidia core. Meanwhile, 3Dfx watched their sales plummet as they realized that shipping the Voodoo3 without support for 32-bit color was turning many gamers away. (Fourth slip: Of course we didn't need 32-bit color support then; most games turned into slideshows at that depth. But the first rule of business is to make the customer happy.
    9. nVidia also got gamers horny by releasing a new chip approximately every six months; 3Dfx was releasing at a rate of one new chip about every year to year-and-a-half. 3Dfx had failed to realize that hardcore gamers demand speed, and that gamers will do just about anything within their power to get that little extra boost of speed from their machines. (If you don't believe me, visit Tweak3D [tweak3d.net] and see for yourself.) 3Dfx was also having serious financial problems; between the collapse of the dot-com-saturated market bubble and disappointing sales, 3Dfx must have realized that going solo was a terrible mistake.
    Since then, 3Dfx has been in free-fall, and the announcement of their sale is just the SPLAT of a once-mighty graphics giant hitting the pavement; even their open-source drivers were not enough to save them. 3Dfx never developed (or at least never released) a new architecture; the V2 through V5 were essentially speed upgrades and hot-glue-and-duct-tape "new features" to their now sorely outdated and overworked architecture. (Why do you think they have required so much more power and cooling? I mean, a power brick for the V5? Are they kidding me?)

    Meanwhile, nVidia continues to dominate the market; they will provide the precious GPU for Microsoft's upcoming XBox, and the GeForce 2 Ultra GTS Pro Whizbang 7 Foomlegricken Supersize (or whatever the hell they call the latest chip) consistently garners rave reviews among journalists and gamers alike. Even the lack of open-source drivers hasn't stopped people (even open-source advocates like myself) from going with them. In short, 3Dfx got beat by a company with a real third-generation graphics chip, but not without help from itself.

    I am sorry to see them go; they were nVidia's biggest rival, and I am always a fan of good competition. My only hope is that nVidia open-sources their own drivers, and perhaps learns from 3Dfx's mistakes. No, I hope we all learn from 3Dfx's mistakes.

    Robert Dumas (robertdumas@hotmail.com)

  • by JanneM (7445) on Friday December 15, 2000 @12:56PM (#555621) Homepage
    The question is of course if Nvidia will take a page from 3dfx:s book and further open their own drivers, or if this is the end of the (relative) openness from 3dfx. Then again, it's quite possible that they will continue to support the open effoert of current 3dfx cards, while continuing to produce closed drivers for their core Nvidia line.

    As Nvidia has claimed that a big reason not to open their drivers is that they are forbidden to do so because of NDA:s with technology partners, one possibility is that as they now own 3dfx technology can use that in place of (probably quite expensive) 3rd party stuff. That would mean more the possibility of increasing the openness towards developers.

  • Are you sure this [slashdot.org] time [slashdot.org]?
  • by Guppy (12314) on Friday December 15, 2000 @02:48PM (#555623)
    First, let me describe how I came to be a 3dfx shareholder. I first started investing while I was still in college, using a little money I had saved up (in part earned by serving as a guinea pig in clinical trials). I began buying 3dfx back before they merged with STB. As shares gradually dropped from my initial purchase price of ~$15, I doubled down, and doubled down again. Eventually, I ended up with about 1,000 shares, at an average purchase price of $10. I really wanted to believe in the company, in the engineers that were always so earnest and hopeful when you spoke to them. And in the products that always seemed to get slammed around by web reviewers, but really kicked butt if you actually took the time to try one out yourself.

    OK, details on this are pretty sparse, but at first look it looks really, really bad for us shareholders. The Motley Fool board has the most active and knowledgable group of 3dfx investors out there, and on that board, some of the mostly highly recommended (ie., plus moderated) messages there right now happen to be pure profanity. Some folks there have suggested we may be getting as little as 0.30 a share, if anything.

    It looks like nVidia may not actually be "buying" 3dfx. Rather, it looks like they will be cherry picking the few assets worth anything, like the designs for 3dfx's next products including Rampage (And it's associated T&L chip, Sage) and Mosaic, and leaving 3dfx as a hollow shell containing nothing but a near-worthless boardmaking plant and lots of debt--in other words, completely screwing over the shareholders in the worst way possible. This hollow shell would also probably be responsible for for providing support, warranties, and driver updates for anybody out there owning a 3dfx card.

    Basically, it looks like I may as well write off my entire investment as a loss, and consider it tuition in the school of hard knocks.
  • by Julius X (14690) on Friday December 15, 2000 @12:57PM (#555630) Homepage
    Here are some more things places you can go.
    • Nvidia's 3dfx Q&A [nvidia.com]
    • 3dfx's third quarter earnings webcast [corporate-ir.net]
    • A letter [3dfx.com] from 3dfx founder Scott Sellers about the sale


    Its really a shame with all the problems that 3dfx has had that they couldn't pull it out of the gutter...they started out great, and made the best products back in the day...if they hadn't bought STB, I doubt this would have ever happened.

    -Julius X
  • "Well.. it seems like the shareholders aren't digging the news too much. NVidia shares fell 10% today."

    Well, us 3dfx shareholders are digging it even less. The stock looks like it's going to be worthless, and shareholders may end up getting a pittance. There are calls for blood out on just about every active TDFX message board out there.
  • It is now nVidia's fiduciary duty to their stockholders to stop wasting money on tech research when it is expensive and unnecessary.

    With the combined nVidia and 3Dfx patents, they can sue any potential competitor in the gaming market who even tries to compete with them, so no possible competitor will get funding and no existing competitor will be able to approach current nVidia performance levels.

    It's an interesting question whether there is any reason to support Linux or the Mac- with Microsoft shaking and issuing earnings warnings, it's not stupid to hang onto whatever support for alternate platforms you have, so binary-only support for Linux and Mac is likely to continue indefinitely. The current nVidia offerings will make their way to a mostly-working support of Linux and Mac, and they will stay there- because it'll be at least five years before we see any significant improvements. No financial return in wasting money on development, remember? When no other competitor _can_ arise because you have the field locked up with patents and can afford to use them as a weapon, it's payback time.

    I'm quite glad that I'm not a serious gamer today. I can play 'X-Plane' quite happily on an old ATI rage128-based card, and don't need to play new games. If I want more flash I can get a PS2. That's the smart bet now- because there's no reason for nVidia to sweat too hard making X-Box that great, even if it does ship. It's PC-based, and the only PC-based 3D vendor of note is nVidia now, so nothing will come along to make X-Box look bad compared to the PC platform. I honestly thought that the PC was going to far outclass X-Box by the time X-Box is supposedly out, but now everything changes because the PC development will stop (fiduciary duty, remember? They can hire some ad-men to go with the lawyers- that'll do. Cheaper than techs) meaning that nVidia can actually cooperate with Microsoft to ensure that PC gaming does _not_ exceed X-Box- assuming of course Microsoft _wants_ X-Box to beat PC gaming, which I guess is up to Microsoft and not you.

    Welcome to the world of the future. Chess anyone?

  • by Menthos (25332) <menthos.gnu@org> on Friday December 15, 2000 @02:51PM (#555638) Homepage
    No. With closed drivers you don't know if it will get better.

    With open drivers and open specs, you know that support for your precious card will only get better in the future - as time goes, someone will improve it, and they have full access.

    With closed drivers, your card will work fine today. But what if NVIDIA tomorrow does an "ID Software" and says that they won't support Linux in the future, due to that niche being to small? Then your card is just a worthless piece of crap... the binary drivers won't be upgraded and you are suddenly stuck with having to use old kernels and old XFree86:s.

    Openness is not as much as about today as it is about tomorrow. Personally, I like to decide myself when I want to scrap my hardware.

  • Consider learning a bit about economics and history before you start moaning about this being bad.

    Several points:

    • In a competitive market, somebody has to be the loser. And loser's tend not to survive very long
    • The GPU market has been shrinking (number of firms producing) for the last few years. Losing one more is not that big of a deal.
    • Sure, 3dfx was responsible for bringing about the whole 3D acceleration thing, but they got lazy, and they got their asses handed to them. Anybody with eyes could've seen this (or something like it, where 3dfx ends up not existing) coming way back with the introduction of the Voodoo2.
    • If nVidia were to increase prices (which is actually economically impossible for them to do, given the state of the market), competitors such as ATi and Matrox would benefit, not nVidia. Add to that the fact that if nVidia were to increase prices, it's very likely we'll see new firms entering the market to take advantage of the situation. In short, nVidia might raise the prices on their chips, but they don't have the market clout to increase prices overall.

    In short, this is neither unexpected, nor bad, and in fact has a number of good qualities.

  • I have some problems with your "timeline"... as someone who works in the game industry and got to work hands-on with each generation of chipsets, both for our games and for others, perhaps I have a better memory for things. But, in any case...
    The other problem was that the Voodoo really only worked with Glide; most other graphics chips could work with OpenGL.
    Actually when the Voodoo Graphics came out pretty much everyone had their own independent library. Glide was one among many. OpenGL wasn't anything more than a twinkle in Carmack's eye at this point as far as games were concerned. Remember, DirectX wasn't even available (we're talking Win95A era), and OpenGL wasn't supported until OSR2 in late '96. And Microsoft didn't support any way of adding it to 95A until much later.
    Still, Glide was the only graphics system that it really worked with, and 3Dfx was convinced that Glide would eventually replace OpenGL.
    Again you're making the weird assumption that OpenGL was the 3D architecture of choice from square one, and was being used by everyone all along (see previous point about OpenGL not even being present in first rev of Win95, which at this point in the timeline comprised the VAST majority of user systems - by my then-employer's research it was at least 90%), and Glide was some weird proprietary architecture that everybody hated to use. Again, Glide was one of many architectures, and certainly was easier to work with than Direct3D (still horribly hobbled by it's Talisman-oriented design), which is what most games in this era were using for a card-independent interface (though for the most part everyone supported Glide). OpenGL wasn't used by anyone outside of Id at this point, and even they were using that werid "miniGL" interface. (again, keep in mind I'm talking about games - just because your insert-whizbang-3D-modeler-here app used OpenGL back then doesn't mean jack)
    As the Voodoo3 and TNT/TNT-Ultra waged war, gamers were slowly coming to realize that nVidia's chip was just as fast (or faster), on par with price, and didn't require the proprietary Glide.
    Funny, I got a Voodoo3 within a couple weeks of it being available, and out of the box I had OpenGL, Glide, and Direct3D support. Obviously you got the BEST performance by using Glide, but that makes perfect sense if you understand what Glide is. A heavily optimized 3D interface/library that understands the silicon and only supports features that are implemented in silicon. Hence, it's fast. Implementing "features" in software introduces performance penalties, just witness NVidia's current FSAA "solution".

    Personally I bought into the whole "TNT is equal to Voodoo3, and the TNT2 is out and blows the Voodoo3 away!" hype myself, and replaced that Voodoo3 with a TNT2. Know what? My framerates didn't increase, they went down. The TNT2 required a faster CPU to get those higher framerates everyone was raving about. On my box (not exactly a slouch, an Asus P2B-L with a P2/400) the Voodoo3 worked better. And later I wished I still had the V3 still in there when I moved to an Athlon and had to suffer the consequences of that decision.

  • by Guppy (12314) on Friday December 15, 2000 @09:02PM (#555657)
    In the past few years we've seen a lot of graphics companies go under. I've tried to put together a list, although I'm sure there may be a few names missing.

    Dead:
    Tseng Labs - gone.
    Western Digital - out of the business.
    Number 9 - gone (?)
    Orchid - gone (?)
    Hercules - gone in all but name.
    Cirrus - Out of the graphics business.
    Neomagic - Moving out of the graphics business.
    Alliance - Out of the graphics business.

    MIA:
    C&T - Purchased by Intel.
    3D Labs - Purchased by Intel.
    Rendition - Purchased by Micron.
    S3 - Purchased by VIA, now focusing on "integrated chipsets".
    Trident - Still making low end cards, mostly living off proceeds from well-timed investment in UMC.

    Still kicking:
    Matrox - Appears to be focusing on corporate and 2D markets. Private company, so little info on internal status.

    Videologic - in Sega's Dreamcast, still attempting to break into PC market, active R&D. Partners with STMicro.

    ATI - Looks like will be in Nintendo's Dolphin, financially still strong (for now). Large (but shrinking) OEM and retail share, dominant in mobile graphics and Macintosh market. Active R&D.

    SiS - Low end cards, apparently active R&D.
  • by RzUpAnmsCwrds (262647) on Friday December 15, 2000 @12:59PM (#555659)
    This is VERY BAD for consumers. In the past, competition between 3DFX and NVidia who both shared the lead position in the 3D card battle have helped prices crashed. Now, it's just NVidia, ATI, and Matrox. Matrox can't really harm anyone with their current 3D technology, and ATI can't convince people that RAEDON cards are up to the level of Geforce2 or Vodoo5 cards. It's really a dissapointment.

    Outlook for the future:
    NVidia will rule the graphics card business. The Geforce3 will be at least $400 and will stay that way due to the lack of an alternative. I can only hope that ATI or Matrox can come out with a comperablie card at a lower price, or we'll all be paying a lot more for 3D cards.

    Doesn't this sound like AOL TIME WARNER? Two giant industry leaders joining together to become a giant? Monopoly is bad for consumers, with media or 3D cards. Like I said, this is bad.
  • with a very, very small foothold in the workstation market

    The 3D workstation graphics board market has been run over by the gamers. E&S has retreated to big simulators, Lockheed doesn't seem to be in 3D boards any more, and Fujitsu exited long ago. NVidia bought out ELSA. 3DLabs is still selling "high end" boards, but their product line is still stuck in 1999, and the latest game boards outperform them. We all know what happened to SGI.

    NVidia's high end line, the "Quadro", is actually the GeForce line with a jumper change. The gamer boards have caught up to the high end, and there's no real distinction any more.

    Well, actually there is. If you buy a Quadro board through ELSA, you get a real warranty, drivers that have been tested, and reachable tech support.

  • I don't care as much about blistering speed (I'm more than happy with 60fps, why do I need 160?), as much as having a card do what I want, doing it reliably and well enough.

    I used to use an ATI Rage Pro (came with the machine, not my choice). For 2d, it did what I wanted well and did it pretty reliably. A refresh rate of 100 was more than enough, and I didn't really consciously think of the video card, as there was nothing wrong, it worked, and so I didn't think about it. For 3d though, well, let's just say I always used software mode rendering.

    Since I like playing all the latest games, and software mode was cutting it less and less, I upgraded to a Voodoo 3 2000. I could have gotten a faster, more showier TNT, but it didn't do what I want, and it didn't do it reliably. What did I want? Glide support for all my old games that had Glide or Software mode, nothing else really; decent 3D performance, which the V3 provided and good 2D, refresh rate of 100; many other things more than good enough and for this iteration, good enough Linux support to do various work & game related things. Grabbing the latest version of XFree86 at the time, there it was. At the time, I was hearing that while NVIDIA was faster, 3dfx was more stable. Hearsay or not, I heard it enough from websites, discussion groups, friends, etc., to lean towards 3dfx in this matter too (although glide support was a bit higher up than small issues, graphics bugs can be very annoying).

    Time moves on again... This time I have a bit more cash to spend, and I'm also very interested in having TV-in capabilities (for various reasons). Again, I hit the reviews, reading between the "200 fps" hype, and looking for what I want, and what I want done reliably. I hear that ATI's drivers, instead of being almost never stable, are now almost never unstable. Performance wise, they appear to score pretty well compared to the GeForce's I'm looking at (I have more money, not infinately more money). And guess what else Radeon comes in? All-In-Wonder! The best TV-in solution, or so everywhere seems to say (even new and better with the Radeon version supposedly). So I get very decent 2D (ATI's strong point since way back), very decent 3D (new to ATI, but still very good) and also very decent TV-in. I bought my ATI AIW Radeon 32MB DDR, and haven't looked back since.

    3dfx + NVIDIA may make a really good card in the future that will make me change my mind, but for now, I'm in the ATI camp. As far as I see in the immediate and near future, ATI's going to be the big competition to NVIDIA. They've got the T&L, got all the whizzbangs, even have some that GeForce's don't (yet, they all end up on the next rev of cards). Not only am I happy with the card, they're also Canadian, like myself. Gotta support those local multinationals. ;)
  • "Really, what do you think? That nVidia should be concerned about the financial well-being of 3dfx investors?"

    Don't be an idiot. Try reading my post again. TDFX should be concerned about the financial well-being of TDFX investors. There are a lot of really fishy things going on in the details of this agreement, and quite a few of us suspect that, in this end game, the 3dfx management may be trying to benefit themselves at the expense of their own shareholder who are the true owners of the company, and that we believe there are ways to dispose of 3dfx's assets that would return more of it's value to us.

    Now, I'm a young fellow who's only been investing (As a long term buy-and-holder) for a few years, and has been frantically saving for the next stage of my life. If I lose money due to my own bad decisions it's my own damn fault (And I've lost quite a bit of it in this market lately), but in the months prior to today's announcement, 3dfx had been making some statements that some shareholders might call misleading, and this latest chicanery is really the last straw.
  • 3DFX could really no longer compete as the Windows/Intel/Nvidia alliance grows stronger by the day... Xbox anyone? Forget about AOL+Time Warner, lets see these three come together... Suddenly The Matrix doesn't seem so far fetched.
  • There have been a number of strange things happening in the 3D market lately. One time industry leader DiamondMM decided to pull out of the board manufacturing industry and focus on its MP3 player. Now 3dfx has also pulled out. It seems to be hard times for the board makers. Two major manufacturers have stopped making products. Is this part of a general trend in the marketplace away from high powered graphics boards?

    Perhaps it is just that those who entered the market first did not make good enough products to survive when the competition got serious. I was never impressed by 3dfxs decision to focus on its own proprietary format instead of OpenGL. Was there ever a 3dfx product that had full OpenGL support? Also, I own a Diamond Viper V770 Ultra and it was a real pain to get working. I was never impressed by their customer service and I recommend against them.

  • ATI has long had a reputation of putting out awful drivers. If they manage to put out decent drivers and improve their technical support, then they might be a worthy contender.

    Bzzt! Thanks for playing! As of two weeks ago, Rage 128 and Rage 128 Pros running OpenGL under Linux Quake 3 [as an example] now generally outperform Windows, thanks to drivers ATI commissioned from Precision Insight. Download them from ATIs site.

    Exactly. ATi's drivers are so bad that some 3rd party drivers can beat them! The fact of the matter is, while we'd all like good Linux drivers from everyone, what's going to determine whether there is any competition in the high-end 3D market is ATi's Windows performance and hence their Windows drivers. If their Windows drivers are so bad that some other company can make better drivers for Linux, then that doesn't bode well for competition in the 3D market.

    On the other hand, it was my impression that the Radeon drivers, while not up to nvidia's standards, were actually not half bad. Now that Matrox, S3 and 3Dfx are out of the consumer 3D market, and now that nvidia is moving into ATi's traditional monopoly market of 3D chipsets for laptops, there is little doubt that ATi is going to make a stronger move towards the high-end consumer 3D space (mainly retail and DIY). Indeed, they began that move with the Radeon, a product ATi took much more seriously than their previous high-end 3D cards. Unlike the cheap-chip-on-a-motherboard space which ATi has traditionally dominated, drivers are important here, and ATi knows it. I would be shocked if bringing their drivers up to nvidia quality is not one of ATi's major goals going forward, and I bet they'll do a decent job at it too.
  • by Primer 55 (263965) on Friday December 15, 2000 @01:03PM (#555707)
    At this rate:

    In one year, AMD will buy out Intel (or vice versa).

    In two years, Microsoft will buy out Apple...

    In three years, VA Linux will be purchased by a twelve year old that has been with the money he has earned delivering newspapers for two months.
  • GeForce + Voodoo =

    VoodooForce!
  • I dont know what kind of problems YOU are having, but nvidia's drivers KICK ASS

    They work marginally well if you have a BX board and one of a subset of nVidia cards and then it still enjoys locking up here and there for no apparent reason. Other chipsets may have AGP issues or might not work with AGP at all. Some people report PCI cards not even working anymore.

    SMP? Forget it.
    Doublescan modes? No chance.
    Proper Modeline Handling? In your dreams.
    Memory Leaks? Sure, have a double helping.
    XINERAMA? Not with GL support.
    Geforce 2 Twinview? Not even if you can get the card to work.
    TV Out? No way jose.
    XF86 RENDER Extension support? Hah, not even in the next version.

    To be fair the linux driver developers are in short supply and are trying their best even though the nVidia market-droids enjoy touting their extremely useless and bloating common codebase which they have 100 developers working on. Out of those 100 developers maybe 5 work on the linux drivers.

    nVidia needs to open these drivers up, if anything the nvidia_drv.o (XAA module) needs to be opened, if you wanna keep the kernel module and GL implimentation closed, great, but man, that XAA module needs help.

    -- iCEBaLM
  • by republic (86647) on Friday December 15, 2000 @01:06PM (#555725)
    I fear that this is a step in the wrong direction for open drivers and open source in general. At least 3dfx supported linux back in the dark ages, then finally saw the light and open sourced their drivers. Nvidia on the other hand gave empty promise after empty promise. They provided us some crappy driver with obfuscated source for almost a year. Now they want us to use their binary drivers because of other peoples IP. After having used 3dfx under linux and struggling for many months with a TNT, I would have to say this is a sad day indeed for linux, the 3D revolution, and open source. I guess from now on I will support Matrox or ATI, because I have had it with Nvidia. My only hope is that the 3dfx employees can enlighten their new collegues.

    Republic
  • Its not like Nvidia won because of a Patent war (hello Rambus!),

    Not only that, but they won in spite of a patent war -- with 3dfx!

    They won by simply flat out making a better product.

    No question at all about that. Here's hoping that they stay hungry and keep driving the industry forward.

  • If you hadn't heard, S3 got out of the Graphics Accelerator market (except for the FireGL series), and changed their name to Sonic Blue.


    -Julius X
  • by Fervent (178271) on Friday December 15, 2000 @01:08PM (#555744)
    Unfortunately, a company that cannot be beat by anyone. Combine NVidia's better technology with 3DFX's better and more experienced marketing division (they managed to sell Voodoo 5's on decent print ads alone) and you have a killer combination.

    Big problem though: with only one large company the market will disappear. It's as if Intel bought AMD.

    Competition in graphic cards, we hardly knew ye.

  • He must have done it, too- Microsoft shares are freefalling after they released a profit warning. The next couple weeks could be verrry interesting...
  • Folks,

    You can rant and rave all you want, but let's face one fact: high end graphics cards are a very limited market item.

    Think about it: how much do the best graphics cards using the nVidia GeForce2 series chipset cost? Well over US$300.

    The thing here is that for the majority of computer users, they don't NEED that level of 3-D graphics acceleration.

    With the advent of the Intel i81x series of motherboard chipsets with built-in graphics functions, that is more than enough for the average home user who uses the computer to surb the Web, run business applications, and the majority of games. The current Intel i815E chipset actually has a pretty decent on-board graphics accelerator--Anandtech and others who have tested the on-board video were pleasantly surprised that it worked reasonably well.

    The reason why 3dfx failed was because their products never recovered from the debacle of the Voodoo3 series, which were overtaken by the nVidia Riva TNT/TNT2/TNT2 Ultra series a few years ago. The arrival of the nVidia GeForce series effectively sealed the doom of 3dfx.

    Right now, only ATI and Matrox are the serious competitors left. ATI's latest Radeon chipset is actually very good, and Matrox's G450 chipset is also quite good.
  • I could see this being a good thing. While Glide is a pretty cool technology, it kind of created a big devide. Standardizing on OpenGL will ease cross-platform development!

    .. or maybe it will make it harder. Let's face it -- I'm no genious. ;)

    ------------
    CitizenC
  • Monopoly is bad for consumers, with media or 3D cards. Like I said, this is bad.

    Your analogy is a little flawed there. What AOL/TW is attempting to do is called vertical integration, which is just as bad. Vertical integration basically means they are eliminating any middlemen in their delivery of "content" to consumers. A comparable situation is that of TV networks also owning studios that produce shows, thereby having their own studios sell them shows at lower prices than competing studios, forcing smaller studios out of business.

    However, you are correct in stating that 3dfx/NVidia is the beginnings of the consolidation of a monopoly. There will still be competition, but as you stated, it will be trivial.

  • This situation will mean only one thing. That soon we may find only bloatness and stagnation in this market.

    NVidia's is surely a winner here. And the quality of their cards much better than 3Dfx's. But how many of this quality was made under the market presence of 3Dfx?

    Let's note one thing. 3Dfx was never a real market dominator. It had a golden time with its 3D accelerator cards. But when it came into the video market it was already a looser, much like the big old gamers here. NVidia made a great move by combining 2D+3D into one card and beat everyone else on this. Meanwhile, it should be noted that this was done having a huge concurrence from two parties 3Dfx and the 2D market. Now they are nearly gone...

    Yes, NVidia did a big job. Their technology is just great! But is everything so cool? No. The 2D is worser than many of its concurrents. I'm sure of this because I did a lot of design and noted that some features on NVidia are even buggy. Specially irritating are some features with colours. Even an old S3Virge manages to produce a much cleaner pallete than NVidia. It is funny but I still keep an old Diamond exactly for this case.

    On 3D everything seems quite cool. Yes everything does seem very good. They do beat all concurrents. But... Bloatness is already in its way. I can't understand NVidia's sticking to make "it's own drivers for X". Yeah cool, it is great but it suxx quite a lot sometimes. For game players, this may seem strange and weird. However OpenGL does not start or end with an X interface. In fact we don't need X to produce 3D. And in scientific work this is BLOATNESS. I need something more than a X driver. I know that the card can shoot more and better than 3Dfx. But it is DAMN slow and buggy when i get out of the game play. Because it sticks too much to present things in a X+games environment and everything else is less cared. and I can't use Windows for such stuff. One file on Windows is enough to overkill the machine, even a Win2000 based. So I have to stick to a 3Dfx V3 to do my work. Now 3Dfx is no more...

    If after acquiring 3Dfx these guys kill Glide and 3Dfx's almost Open Source policies then this will make a huge blow. Yeah you windowsers will be happy. Quakers, Starcrafters, Counterstrikers will surely be happy for some time longer than us. but the fact is that having a company sticking into one trend (games) and not releasing specs will immediately have a blowing effect on *NIX world.

    However this is not the worst. the worst will be to see this company sticking to its own rules and disregarding everyone else. Even a 5% 3Dfx presence on the market was enough to push NVidia further. 3Dfx were the forefathers of 3D acceleration, a standard to overcome. Now there are no lights around. A few concurrents and a market that looks more as a bunch of play-hunger users + some irritating Open Source hackers. NVidia will surely turn over the first. But it will have no clear incentives to do something more than 3D gaming. Later it may not have any incentive to improve 3D gaming itself.

  • I agree. 3DFX cards easily were the most reliable I'd ever seen. And XF86 loved them.
  • Granted, many probably think 3dfx hasn't been much of a challenger to NVIDIA since who-knows-when, and the purchase may really not mean all that much in the end, but still...

    Am I the only one that finds it unsettling that:

    • Once there was 3dfx and NVidia, but the latter swallowed the former
    • Once there was Creative Labs and Aureal, but the former swallowed the latter

    Or is it just me being bitter because I purchased hardware from both 3dfx and Aureal, only to find each company disappearing?

    To be fair, I miss Aureal (and its tech) more than I will miss 3dfx (and whatever may come to pass for its tech), but still...

    I wish I had the illusion of choice, or at least the opportunity to delude myself into finding it. Ah, well, back to Matrox. :P

    Me,

    being bitter

  • (I apologise if this was answered in one of the articles. I just skimmed them.)

    So will there continue to be two distinct product lines? Or will the 3dfx tech be assimilated and used to help create an entirely new generation of graphics accelerator/processor?

    If you ask me, these two schools of graphics technology could learn from each other. With nvidia's recent trend towards speed rather than eyecandy, and 3dfx doing much the opposite (speed => effects), we could soon see blazingly fast, awesome-looking graphics headed our way, courtesy of the new VoodooForce card.

    Or, perhaps they will choose to keep the brands seperate. Any speculation?

    --Psi

    Max, in America, it's customary to drive on the right.

  • The way I see it, the card's nVidia release in the next 2 years will be absolutely amazing. They'll have the best of 3dfx and nVid tech in them. (and hopefully we'll see the end to 3dfxGlide, that just got annoying for non 3dfx'ers)

    Yet after the initial burst, I do see the company getting lazy and just releasing "amazing new features" that do nothing, but just keep the customers buying.

    Alternatively development could continue just as well, but prices could get too high. Either way, I'm gonna buy a great 3D card in 2 years because it's going to have to last me a while.
  • by drsoran (979) on Friday December 15, 2000 @01:10PM (#555766)
    Well, since Nvidia has acquired the patents to the 3dfx stuff, we'll probably see them shift pace a little bit and start to use some of their technology. Expect the NV20 to use 6 GeForce2 GPU's on one card and require a seperate external 250 watt power supply. :-)
  • by Temporal (96070) on Friday December 15, 2000 @01:12PM (#555771) Journal

    I am a big NVidia fan (as many of you know). I really think they have better hardware and I love working with the special features they add to their stuff in my 3D game engine. But this news worries me.

    I hope NVidia will continue to advance the industry at the same rate as they did in the past. Without 3dfx as competition, their incentive may not be so great as it was before...

    However, there is plenty of reason to believe that these concerns are misplaced. The ATI Radeon is a good card, having some features (like the third TMU) which not even the GeForce 2 has. Also, NVidia hardware is now being used in consoles as well as computers. Tough competition in the console arena is pretty much gaurenteed for them. So, as long as they continue to use their console gaming hardware in their video cards, we can continue expect new, better hardware from them.

    This really could go either way. We'll have to watch and see what happens. If they do stop advancing their hardware, or charge too much for them, I will stop supporting them. Let's hope it doesn't come to that.

    ------

  • Glide is, for all practical intents and purposes, dead. If you think about how many major glide-based games that have come out recently, you get something like 2 (Deus Ex & UT, and how old is UT now?). Can anyone think of any bigger (i.e., not crap like Extreme Bull Farting 2001) games that are coming out, where the developers are advertising glide support?
  • It is indeed an honour to meet someone who is not in awe of 3D card-board effects in the gaming industry.

    What do you need to create a good game? Good enough representation to get your ideas across, this can be 2D, 3D isometric, text, or (depending on your idea being completely useless without it) perspective, texture-mapped, 3D-vector graphics.

    The top-10 best selling game franchises include: The Sims, Championship manager, Command and Conquer, (and here in the UK) Who wants to be a millionaire? Some of the other best sellers are variation of these. These games are either 3D isometric or text (with 2D backdrops and coloured fonts). The vast majority of games probably are 3D cardboard effect games with little chance of success, as Good game developers, and public in general, realise that you do not need Cardboard effects to be a good game.

    Good ridance to the 3D-cardboard industry, and heres hoping for a restoration of balance and sanity in the game industry once again. Here is to definitions of gameplay which do not include superfulous graphics. Here are to games of the future whose representation is not predefined before the game is even written.
  • by huie (148646) <mhuie.netcom@com> on Friday December 15, 2000 @01:30PM (#555781)
    3Dfx wants to substantially reduce its costs in order to best conserve its resources. These cost-cutting measures include a reduction of substantially all of the company's workforce by early next year so that the comapny provides the highest return to our creditors, shareholders, and employees.

    So, they're laying off everybody in order to preserve the stock value for the employees?

  • If anybody can, it's Bitboys! I mean, they did just announce their mega-merger [somethingawful.com] with Rambus.
  • by CaptainCarrot (84625) on Friday December 15, 2000 @01:31PM (#555786)
    So now all the lawsuits can be kept inhouse, right? They only need one legal team so costs are kept way down, the discovery phase will take nowhere near as long since all the documents are kept in the same office, and they win no matter which way the ruling goes.

    As far as I can see, this is a win-win deal any way you look at it!

  • I have a pending 'lifetime' warranty claim on my 3DFX Voodoo3, I wonder what's going to happen with that..

    *looks out the window at the Matrox bldg.*
  • Well, now.. The NViddy kiddies now officially rule the high-end PC graphics market. And with PS2 going 'thud', next year they stand to also rule, with M$FT, the console space with XBox. This is disquieting at best. They are in a position to not only stifle PC/console graphics capability competition, but to start rolling out new graphics tech at a rate best suited to their financial gain (instead of as demanded by a state of healthy competition). I like NV, I like their cards, but this state of affairs can only be bad for the consumer. Of course, AI co-processors may well be the next big thing in gaming. If NVid makes graphics less interesting, there are other plcaes gamers can put their money. My $.03.
  • by Junks Jerzey (54586) on Friday December 15, 2000 @01:17PM (#555792)
    Gotta give 3dfx credit; it jumpstarted 3D hardware for the PC. They weren't the first, but they brought inexpensive workstation class 3D to home PCs with a bang. Even today, the Voodoo 1 is an impressive piece of hardware.

    Just speculation here, but I can't help but wonder if the performance video card market is much smaller than assumed. I expect most people just use whatever video card came with their machine. I've read that 80% of all video cards out there are from ATI, because they dominate the OEM market. Having a bunch of big fish fighting over the last 20% would be pretty rough.
  • Its worse than that. A lot worse...

    There may be a reason why this sale will not go through, but I'm still hoping(shareholders can say no...). 3dfx and Nvidia together completely lead the retail market. (Not OEM/laptop sails here, retail.) Now with 3dfx being consumed this leaves us with Nvidia. Now here is the kicker: ATI Market Share (Currently) 26% Nvidia market share (currently) 20% 3dfx market share(currently) 7% These numbers may be a bit old or off. But, with nvidia + 3dfx you get a new market break down where(through simple addition), 3dfx/nvidia will have a lead. Now that leaves ATI in a bit of a bad spot, and I *really* doubt that 3dfx is going to release anything with open source drivers now. Which leaves only 1 company(ATI) with a good, open source gaming board which can be easily obtained in the retail market.

  • by will_code_for_beer (224193) on Friday December 15, 2000 @01:23PM (#555798)
    ATI has long had a reputation of putting out awful drivers. If they manage to put out decent drivers and improve their technical support, then they might be a worthy contender. If you ever visit the ATI newsgroup you'll see post after post about poor support and shoddy performance due to drivers/software.
  • 1: Lose the profanities
    2: Yes, they do. TIME WARNER owns the CABLE INTRENET SERVICE in many states ("RoadRunner"). AOL now owns a virtual monopoly over the internet market in those areas. TIME WARNER owns some of the most visited websites on the internet - CNN.com comes to mind, it's #4 - AOL now holds a virtual monopoly over the news services. CNNHN and CNN are the #1 and #2 most watched news channels. AOL.com and CNN.com are the #1 and #2 most used internet news "portals". AOL.com + CNN.com = GIANT NEWS SITE.

    Get the idea?
  • by Bulldawg2000 (264224) on Friday December 15, 2000 @01:24PM (#555805)
    I'm not sure about 3dfx's assets, but if it is below 112M they were running the risk of being liquidated. At least this way, we'll see their technology incorporated into future products.
  • Maybe you shouldn't use slashdot for your only source of news... I have a feeling that the moodswings were more a "slashdotian" thing, not one from 3dfx.
  • check out:
    http://www.voodooextreme.com/
    They have more information on the subject and lots of news about 3d gaming.
  • Well, look at the 3d hardware business from the manufacturing angle: a state-of-the-art 3d chip is as complicated (if not more complicated) as a general-purpose CPU, with similar die-sizes and transistor counts. Unlike Intel and AMD, however, they can not sell the newest stuff at a hefty premium, instead they sell the chip complete with memory, bus logic and ports -- cheaper than a comparative general processor. Then add the requirement to get something entirely new (not just a clock speed tweak) out the door every six to nine months... No wonder that companies are falling faster than the Microsoft stock price.

    On the other hand, the graphics companies are not encumbered with keeping up with legacy systems -- their stuff isn't really programmable from the user level, so all they need is a new set of drivers to support their new hardware. This should make for much speedier development.

    It would be interesting -- but not entierly surprising -- if graphics hardware makers would start driving the IC business the same way that games are driving software today.

  • by Tridus (79566) on Friday December 15, 2000 @01:29PM (#555812) Homepage
    What is it with all the posts in here moaning about how bad this is, because of how there is no competition and other such...

    People, please... are you living in a fantasy land? 3dfx hasn't been any serious competition for Nvidia since the GeForce. As much as you might like to think that the Linux market matters, it doesn't. Nvidia destroyed 3dfx in the Windows market, aka the one that matters.

    Contrary to several posts in here, this is not a bad thing. Its the natural course of business.

    - 3dfx dominates market
    - Nvidia enters market
    - 3dfx gets lazy
    - Nvidia makes better products
    - Nvidia dominates market
    - Nvidia continues to make better products, and 3dfx crumbles against the competition.

    Its not like Nvidia won because of a Patent war (hello Rambus!), because of backdoor shenagians, or whatever else thats bad. They won by simply flat out making a better product.

    People who think this is some kind of disaster want a market where nothing changes. If you actually want innovation and competition, you had better expect that some players will loose at some point! If you want competition but without the potential to loose, you don't really want competition, you want to live in a fantasy world.

    ATI is still there, as a better managed company then 3dfx with a better product, they stand a better chance in this market.

    Now please... quit bitching about how this is the end of the world. Its not. After the 40th post moaning and whining about it, it really does start to get tiresome.
  • by -=[ SYRiNX ]=- (79568) on Friday December 15, 2000 @01:43PM (#555815) Homepage
    In a capitalist market, survival of the fittest is a rule, and 3Dfx just wasn't fit anymore. They were releasing crappy products, failing to license out their technology, and selling their cards only at retail (rather than working to get them integrated or included with new PCs).

    A few people here have mentioned the Voodoo 1, 2, and 3's good cost-performance ratios and their stable drivers, and those are very true observations. However, 3Dfx shot themselves in the foot when they stopped marketing their products in any ways that bring in strong revenue. They stopped running television ads and stopped licensing their chip technology to other board manufacturers--and all the revenue dried up right there.

    Someone here said that ATI owns 80% of the PC video card market because they get their chips and boards included with OEM systems. ATI has traditionally made somewhat crappy products, but as a business they have thrived because they know how to work themselves into revenue-making positions in the marketplace. They have become so successful as a business, in fact, that they have finally gotten back around to investing more resources in R&D and QA, and the quality of their products has improved radically in the last 2-4 years.

    nVidia is especially interesting because they have a pretty even mix (IMHO) of product and marketing excellence. They have figured out how to achieve rapid growth in both areas. The only problem is that they now have only one worthy competitor left on their radar (ATI; Matrox just doesn't have enough market share or technical superiority). When one company in a given industry pulls far, far ahead of all the rest, it means they run the risk of getting lazy.

    That hasn't happened with all large behemoths (Sony and Microsoft continue to work hard and produce excellent stuff, for instance), but it has happened with many (General Electric, Phillips/Magnavox). Let's hope that nVidia doesn't get too cozy as it approaches the top of the food chain.

  • by StandardDeviant (122674) on Friday December 15, 2000 @01:44PM (#555816) Homepage Journal

    I see a large number of posts saying that this will lead to stagnation in the video card market, increased prices, blah blah blah, the usual bad things that come with monopoly. I don't think these fears are grounded in a solid grasp of reality.

    The video card market is much broader than the high-end-home-user-gamer-speed-freak niche. Although I do not have exact figures to back this up, I'd wager that the total amount of cards sold as integrated solutions (part of a Dell or the like) to both the business and non-gamer household market exceeds the gamer market by a large integer multiple. NVidia IMHO makes the best current 3d hardware, but they have nothing in the business/SOHO/laptop/OEM market that I'm aware of, whereas Matrox and ATI have vast sums of revenue from those markets. With that kind of revenue stream, they could probably each buy NVidia several times over.

    In short, don't assume that becuase NVidia has become the de facto monopolist in the gamer market (with a very, very small foothold in the workstation market[1]) they are somehow the totality of the video card market. They will continue to face competition from Matrox and ATI for the forseeable future.

    [1] quadro and somebody was telling me the new sgi vpro line of graphics chipsets was based on NVIdia tech


    --

  • My numbers were a tad out of date, here are the updated numbers:

    http://www.tomshardware.com/graphic/00q4/001129/ky ro-01.html

    So, in reality we have:
    ATI at 25% right now
    NVIDIA at 15% right now
    3dfx at 5% right now

    After the sale, NVIDIA will have 20%, ATI 25%... Still a sizeable lead. And as was noted before, NVIDIA isn't really buying 3dfx, they are just getting their chips, which means that NVIDIA will NOT get all of 3dfx sales. These numbers just show, and still, that this will put NVIDIA much closer ATI than the were before.

  • would perhaps be to get together with Matrox?
  • You really have to give a lot of credit to NVIDIA. Just two years ago, they were an obscure chip maker, and now they are the 800lb gorilla of the market. This whole 3DFx buyout is just symbolic; 3DFx was dead in the water anyway. However, it symbolizes the total defeat 3DFx has suffered at the hands of NVIDIA. I can remember when they released their original NV1 chip. It was non-standard (quads instead of triangles) was slow, and had very limited support. Then I remember the Riva128. It had pretty bad picture quality, but for several months held the speed crown. When the TNT came out, everyone pretty much knew NVIDIA would go to the top. Thankfully, they haven't lost their small-company image through all this. They still make great cards at great prices, have awesome customer support (eg. their latest drivers still upgrade even old cards like the TNT) and they are one of the few consumer vendors to get OpenGL right. If it weren't for NVIDIA, the rise of OpenGL would have been severely hindered. As I recall, they were the first major chip maker to publish a full, pro-caliber ICD. Otherwise, all those nifty OGL apps might still be useless on a market of consumer cards with "Quake drivers." All I can say is NVIDIA is 'da bomb!

    PS> No, I did not get payed by NVIDIA to say this. I even have reasons to dislike NVIDIA (they won't give 3D specs to Be) However, I can see some logic in their desicion (BeOS might be a therat to the SGI-blessed Linux for 3D ;) and don't hold it against them. Of course, I would not be at all dissapointed if they would suddenly change their minds, and my new NV20 would accelerate GLTeapot for me ;)
  • You've got it the other way around -- CL sued Aureal for patent infringement. ... Aureal did turn around and sue CL on the basis of their supposedly frivolous lawsuit

    It's clear from its press release [creative.com] that Creative Labs is a patent aggressor. CREAF boasts that "EMU's ... patent was upheld and found to be valid and enforceable," and pledges to appeal (or buy Aureal, whichever comes first).

    I can't find any details on Aureal's lawsuit.

  • The real story here is that 3dfx achieved dominance back in the day and handed control over to marketing and bought STB. Their product declined as a result. In the mean time nvidia got their shit together and started pumping out excellent product with 6 month product generations. 3dfx responded with more marketing and got their ass handed to them. Back when STB was purchased by 3dfx and Diamond by S3 everybody figured Nvidia was not long for the world, instead they are the last one standing. If Nvidia tries the same thing they may last a bit longer, but not by much, either matrox or ati or even intel will get their shit together and hand them their ass. The one problem is that GPU development is nearly as complex as CPU development and doesn't have anywhere near the profit margin. As a result Nvidia may try to up the price in response.

Those who can, do; those who can't, write. Those who can't write work for the Bell Labs Record.

Working...