Speculation On AMD Buying Transmeta 119
Nuke Skyjumper writes: "According to a report on CNBC, it appears AMD is interested in purchasing Transmeta. I wonder what the implications for Linus, and in particular, Linux would be?" With the recent agreement between them, some people see them working even more closely together. But there's been a lot of hot air about this before - I think at one point people had been talking about Transmeta buying AMD. But, as always, time will tell.
Its a good idea... (Score:2)
Mark Duell
why would this effect linux? (Score:2)
I really don't see how such speculation merits front-page treatment. Slashdot's pro-Transmeta propoganda should have died off as soon as it was clear that Transmeta's chips are not the cure for Microsoft, world hunger, and freedom for all.
--
Cheap PC's... (Score:1)
Re:.sig redirects to goatse.cx (Score:1)
Mark Duell
Speculation...Speculation (Score:2)
AMD's interest in Transmeta... (Score:2)
There is other technology there they would be interested in, I know... but AMD is doing pretty poorly against Intel in the mobile market, for now anyway.
"Everything you know is wrong. (And stupid.)"
Bidding war? (Score:2)
I'm interested in buying Transmeta too.
Take Paypal?
Big fish little fish (Score:2)
As for Transmeta, they have their own niche in the market and should stick to their focus, many businesses with strong policies on their business models and focusing on those models do rather well, unlike others who try to carry the world on their shoulders such as Lucent who is getting slightly pounded, possibly from jumping into too many different segments and forgetting their core business model. (sounds correct although I am tired)
How significantly would AMD captive anything in the processor sector buy purchasing Transmeta? Not much, they should save the money for hard times and focus on their own stuff instead of trying to have their cake and eating someone else's.
Home sweet home [antioffline.com]
Re:why would this effect linux? (Score:1)
You're thinking of Intel... they were the ones who tried to block other CPUs using their CPUID.
Too bad for them, no developers thought that limiting their market to people who owned a certain CPU was a good idea.
Besides, its pretty much impossible to limit Linux to a certain CPU: Its open-source. Of course, Linus could change that, but I don't think AMD could convince him to do any of this.
Re:All this A buys B, B buys C is *bad* (Score:1)
Interesting, but not likely... (Score:5)
Not to mention that AMD already is licensing the most important thing it could get from Transmeta -- the code-morphing tech so that they can simulate their upcoming chips.
If Transmeta had some fab capacity of their own, it might be different, but IBM produces their chips, and I don't think AMD's quite ready to buy IBM. :-)
Maybe in a couple years when AMD has an extra couple billion dollars sitting around and/or Transmeta's stock crashes, but for now, AMD is selling every Athlon they can produce without any help. There's just no good financial or technical reason for AMD to do this right now, not in their current position.
Re:Interesting, but not likely... (Score:1)
-Andy
Re:All this A buys B, B buys C is *bad* (Score:1)
Enigma
I wouldn't worry about Linus. (Score:4)
Regarding Transmeta, my company is developing some internet appliances using Crusoe. Maybe you'll not believe me, but the prototype boards are the primary development platform. People are even compiling XFree86 on them!
By the way, am I the first in the world to run Digger under Linux+DOSEMU on Crusoe?
Won't happen... (Score:1)
Well... (Score:1)
IMHO, Transmeta has come up with one of the coolest ideas in computer history: they're making a platform designed for dynamic recompilation. If it gets buried, I'll be sad.
---
pb Reply or e-mail; don't vaguely moderate [ncsu.edu].
Transmeta to cool down AMD (Score:2)
My Duron melted my copper heat sink...
Re:Its a good idea... (Score:2)
Intel will release it too. First it made multimedia come alive. Next it sped the internet up. Maybe with the Very Long Words it will make english reports better.
Re:Interesting, but not likely... (Score:2)
Re:why would this effect linux? (Score:1)
Re:All this A buys B, B buys C is *bad* (Score:1)
Remind me never to trust you on anything requiring an understanding of statistical trends. Because of a 6 month trend, put against the backdrop of 20+ years of massive mergers all over the board, you conclude "mergers are over for good"? Jeez.
Tired of Linux Stories, and /. Rumors (Score:3)
I hope this doesn't offend the poster, or any of the slashdot "elders", but I am sick and tired of this type of post.
1. Does anyone really care if a failing, over-hyped company such as transmeta may get bought out?
2. Is anyone else sick of "What will this do to Linux" posts?
3. C'mon people, there are plenty of good stories, and I would like there to be a Senseless speculation and/or Linux fears category.
Every day, it seems, there is a new post, about a big evil corporation with plans to fork linux, or perhaps spoon linux, or perhaps make Linus a job offer...
At least discussion of the MPAA gets me fired up & frustrated with stupid laws.
I urge others to speak-up if they too would like a seperate board/section for these stories.
Please - no offense to the poster, this is about the quality and usefullness of slashdot only.
I survived the hype of survivor.
Re:Interesting, but not likely... (Score:1)
overpriced (Score:2)
Transmeta is all vaporware, a couple of well known figures, and a chaos.
But, hey! in 3 years they could bring us another PC chip ;)
Transmeta: A Solution in Search of a Problem? (Score:1)
The article notes the point that who is going to want to run a server with Transmeta's watered-down chip? Furthermore, is the ability to extend battery life a couple hours a good trade-off for processing power?
While I can see several niche markets where this might be useful, I'm wondering if Transmeta can survive in the long term with what its done so far.
Anyone have any thoughts on this?
Re:Interesting, but not likely... (Score:3)
Re:Interesting, but not likely... (Score:1)
It might well be a great long-term investment, but AMD just couldn't afford it without strangling themselves for the next several years.
Re:I wouldn't worry about Linus. (Score:1)
I think theres a mistake here (Score:3)
Transmeta + Linus != Linux.
Linus has said many times that he's not interested in the commercial parts of Linux, it's just a hobby for him. If AMD were to buy out Transmeta and try to force Linus to put pro-AMD code into the kernel, he'd likely quit Transmeta, or at least tell AMD to go screw themselves. Transmeta has nothing to do with Linux, aside from the fact that one of the people working there happens to ALSO work on Linux in his spare time.
Re:Won't happen... (Score:4)
All of you people talking about how the DOJ would get involved must be on crack. AMD has less than 25% of the processor market. Transmeta is barely even a blip. The DOJ won't give a rat's arse about them merging.
Re:Its a good idea... (Score:2)
For example, Linus Torvalds => Steve Jobbs in 5 steps.
1. Linus Torvalds works for Transmeta.
2. Transmeta get's bought out by AMD.
3. AMD CEO W.J. Sanders III held a variety of positions in the engineering, sales, and marketing departments of Motorola Semiconductor.
4. Motorola chips are the processors used at Apple.
5. Steve Jobbs is the CEO of Apple.
Can you do it in fewer than 5 steps?
Yeah, right ... (Score:1)
Less competition (Score:3)
Right now there aren't many processor manufacturers in the whole computer industry (PC, Apple, pda's, web appliances, etc). You have Intel, AMD, Transmeta, Motorola, IBM, Sun, Alpha. I know there are more, but those are the major players atm. If I missed any, feel free to add them to this list.
So what happens when one of them dissapears? Less choice, less competition. And no matter what these companies will promise (saying that they will keep their speed in developing new technology, and keep costs low) it won't happen. Because developement is driven mainly by competition, and not much else.
I lived for a long time in a comunist (now ex-comunist) country, Romania. And I've seen what happened in an economy that lacked any form of competition. In the 1970s, during a brief period of change in politics, the government bought a lot of top of the line technology (cars, computers, and a lot of other high-tech stuff). But only a few years later, things turned to worse, because the president of the time, Ceausescu closed off the trade with the Western Europe and the US. The decision was to run everything within Romania, without any world contact.
Fast forward to the 1990s... The car designs that were bought in 1968, the Renault 12, was still being built, in it's original shape and form (nothing changed on it). The top of the line computers were some 8086 clones. Everything had stagnated, as if for the whole 20-something years that Romania closed off its borders nothing had happened.
Now granted, this is an extreme case, and the chances of this happening in a capitalist society are very slim. But some of these effects can (and do) take place every day. So I really think that AMD buying Transmeta (or the other way around) can be a very bad thing.
Re:Interesting, but not likely... (Score:2)
Also, acquisitions usually are at a premium to the current stock price, so AMD would probably have to pay well over TMTA's current $3+ billion to acquire them. They just can't afford it.
Re:overpriced (Score:1)
It's easy to flame and/or whine here on slashdot. The reason AMD has a market cap of $7B while Transmeta is $3B is the market's belief (or lack thereof) in the two companies.
If you think the market is on drugs (and it wouldn't be the first time), _do_ something about it.
Re:why would this effect linux? (Score:2)
Re:Its a good idea... (Score:2)
also have a look at the 32 bit benchmarks for Merced, sorry, Itainim if you wanna see true Intel "Inovatoin"
first the P4 is slower then a P3 (yet costs more) then Merced is slower then a P1 75MHz (yet costs lots more!) whats next? a 128 bit chip thats slower then the origonial 8080??
Re:what it means ? (Score:2)
Oh, come on. Can you imagine the public outcry if they tried something like that? Not that it would make any difference - Linus has little to do with IA64 support in the kernel. Besides can you see Linus accepting that? Personally I credit him with a bit more integrity than that. It's not like he'd have difficulty finding another job.
Re:AMD's interest in Transmeta... (Score:2)
What I want! (Score:1)
You know what I'm desperately hoping for?
Computers that are Fast Enough. Are Macs? I dunno, I guess they might be.
Not only are they *fast enough*, but they are engineered. Quiet. Cool. Easy to upgrade and fix. Useable. Reliable.
I'm drawing qualities from cars; at first they were simple, crude, and very expensive. Elite, even. Eventually they became extensions and expressions of our personality. They became specialized, tailored, fashionable, and above all, reliable. They don't regularly crash and stall, now.
I want computers the same. Notebooks, desktops, TiVos, communications devices, consoles, etc.
Geek dating! [bunnyhop.com]
Re:AMD's interest in Transmeta... (Score:3)
Don't get me wrong... I think that the power consumption of the Transmeta chips would be great for something like the Athlon... but the real kicker would be code morphing for the Sledgehammer. Everybody knows the trouble that Intel has had with the Itanium x86 emulation... and AMD wants to correct that ASAP, especially with the 32bit-64bit hybrid that they have planned.
An added bonus: the combination of the two companies would be so large that Microsoft could not afford to ignore supporting it with their 64-bit OS (which they have not yet committed to for the Sledgehammer). And I don't care what you think about Microsoft, but they are DAMN important to much of the IT world.
-rt-
Re:I wouldn't worry about Linus. (Score:1)
Gerhard
Re:Less competition (Score:1)
Advanced Meta Devices (Score:3)
AMD-TM
Meta Advanced Devices
Athlon, Duron, Cruson?
can we expect a barrage of silence from AMD now? :)
rr
Re:All this A buys B, B buys C is *bad* (Score:1)
2. Assigning blame isn't the point of the exercise when the issue is monetarily getting screwed over. Joe sixpack doesn't go to court if coke is priced too high, he just stops drinking it. If Intel, AMD, & Transmeta would collude to raise CPU prices 10x you would have a renaissance of mac sales, Alphas would become a force in the consumer market, and even Sparcs would have a renaissance period.
In short, you sound like some bitter socialist who's sad that Gorbachev's teaching in San Francisco instead of running the USSR.
DB
Re:Its a good idea... (Score:2)
Linus Torvalds wrote Linux.
Linux runs on Macs.
Macs are produced by Apple.
Steve Jobs is the president of Apple.
Anybody else want to try it? ; )
Re:Its a good idea... (Score:1)
Transmeta PowerPC emulation? (Score:3)
AMD's volume profits could comfortably absorb the cost of selling the chips at prices competitive with Motorola immediately, and even with the lower-than-average performance of the Transmeta chip, a 1.5-2GHz Code-Morphing Athlon would likely whup any Moto. G3/G4 in non-Altivec benchmarks.
If I were looking for the fastest way to support MacOS 9 (will be important for at least the next year) and MacOS X on the x86 platform, then something pretty similar to a Duron with a Transmeta PPC-emulation layer might just be the way.
Apple have proved they have the marketing department and design group from hell, and an OS (Mac OS X) that needs to pick up serious attention outside the existing Mac market to bring Apple profits up. That means selling lots of machines into the hands of people who have never before owned a Mac.
Not easy when your fastest model runs at 700MHz and costs US$5000 without a monitor.
I don't think i can buy an AMD or Intel chip less than 600MHz at the moment, and Motorola are not going to be able to double their clockspeed this year.
Apples biggest problem currently is MHz... even though it might only perform like a PPC of half the clockspeed, it would be good for Apple to be able to advertise '1.5 GHz Macs'.
Apple could continue to offer Altivec models to the scientific, creative and education industries
while targeting the G3-alike AMD chip at corporate/home users.
It might not encode MPEG-2 in realtime in software, but it'll run MS-Office like a raped ape.
And, sadly, thats all the computer buying public seem to give a shit about these days.
Re:why would this effect linux? (Score:1)
DB
Re:What's your problem with Intel? (Score:1)
-- Eat your greens or I'll hit you!
Re:Its a good idea... (Score:2)
Linus Torvalds wrote Linux.
Apple Tweaked Linux (see http://www.mklinux.org [mklinux.org])
Steve Jobs is CEO of Apple
thats 3!
Intel, AMD and Linux (Score:2)
Come on people... If anything it would get a major hardware maker in the linux corner - plus the chips they make at Transmeta are x86 emulated chips.
Hell we could see linux shipping with new pc's that have AMD chips... but i doubt it.
AMD isn't buying Linus.
Oh, and doesn't AMD already have a bigger market share than Intel, I heard that on CNBC like 6 months ago, and people are still talking about them taking charge.
I run an AMD and love it, my 750 Athlon is fast enough to reboot into linux after my roomates run a windows session it doesn't bother me they are still in the dark ages of computing.
Linus & Commercial interests (Score:1)
I suppose there's also the issue of the Linux trademark... While Linus is a benevolent dictator, it's ok, but there are plenty of risks here...
rr
I doubt it's worth it (Score:2)
Re:Less competition (Score:3)
Government action is a special case of non-competition. When Ceausescu wanted to stop something from happening, lots of people with guns, poisons, and other instruments of violence went into motion. In a corporate world, this doesn't generally happen. You *can* have corporations rent or hire governments to do the violence for them but if you have a decent constitution and a limited government, the amount of violence on sale is really quite limited. Free elections also tend to make it essential that such deals remain unpublicized.
Getting back to the chip market, Transmeta was formed as a startup around 5 years ago because somebody got a bright idea and a lot of bright people bored with the existing chip companies joined him to make it work. Nothing is stopping that process from repeating an infinite number of times except for a lack of imagination on the part of hardware engineers. As long as the process of forming new startups continues to be available, your nightmare scenario won't happen or won't happen for long.
Another romanian on slashdot
DB
A Duron emulating a PPC (Score:1)
Re:AMD loves Linux (Score:1)
Maybe Linus should just buy AMD himself (I'd donate a couple bucks) and implement the whole thing and in the mean time kick Bill Gates out.
Re:overpriced (Score:1)
Re:Transmeta PowerPC emulation? (Score:2)
As for advertising 1.5GHz Macs, if Apple/Motorola had wanted to do that, they would have needed to do little more than add a 2x clock divider to the processor (I'm sure they could have found some technical justification).
Re:All this A buys B, B buys C is *bad* (Score:2)
Haha. You undermine your own argument. The US has the biggest government of any country, and the largest economy. How come then, since the US government is less limited than in any other country in the world, you imply that people in he US don't live worse lives?
Assigning blame isn't the point of the exercise when the issue is monetarily getting screwed over.
Who was talking about monetarily getting screwed over? I'm talking about *really* getting screwed over-- being shut out from information, being denied needed medical care, being denied insurance coverage you paid for, being poisoned by toxic waste, etc.
Re:Its a good idea... (Score:1)
(Now this is a wasted post.)
Re:overpriced (Score:1)
Selling stock you don't own and buying back once the quote dropped. Can be a real pain if the quote doesn't drop.
Re:All this A buys B, B buys C is *bad* (Score:2)
Absolute garbage. People in Western Europe have a much better quality of life (Britain excepted as the 51st state anyway) due to the government providing things that corporations wouldn't provide, such as quality universal health care, good quality and cheap public transport and housing for all. No-one has to live in a trailer in Europe for example.
Re:what it means ? (Score:1)
Re:Its a good idea... (Score:2)
2. Steve Jobs is CEO of Apple.
Re:Transmeta: A Solution in Search of a Problem? (Score:1)
Speaking of laptops, the performance of the current CPUs is overkill for many applications. For someone who mainly uses an office suite, the extra battery life might be more important than another increase in performance. Actually, I see more and more talk about underclocking in online forums, with the goal of running your desktop without these noisy fans.
Re:Interesting, but not likely... (Score:2)
well, that's just the thing, isn't it? Transmeta could be in the position to crash at once. I personally feel transmeta will tank by summer. The hype they generated in the year+ prior to any actual product anouncements vastly inflated the float prices, co-mingled with the dearth of licensees and such hokey manufacturing arrangements all conspire to suggest immenent critical failure.
call it a hunch.
If I were uppr mgmt at transmeta, I would be looking to find a solution while the price is still high. Selling to AMD at a slight premium over market would be like striking gold. (Otherwise I would be cashing out what I could before Lerach & cronies could finish us off.)
I am gearing myself up to short them when the time is right. The indicator I'm watching for is the first wave of talent leaving. These will be the smart folk seeking security. I wonder if Linus will be in that wave? I even wonder if this is the week we see it beginning...
:)Fudboy
Re:Less competition (Score:3)
The processor industry, on the other hand, is one of the most competitive that exist. That's why Moore's Law continues to apply. Now I agree that the loss of choice is a bad thing, but you got to keep perspective. Transmeta is insignificantly small in the processor marketplace, despite the hype and their market cap. Frankly I think the loss of 3dfx is more concerning.
Re:Less competition (Score:2)
This is just a partially-complete list of proccessor vendors with products found in machines that will run Linux or NetBSD. There are many, many, many more that deal exclusively with embedded processors or processors for special applications like rad-hardened and DSP.
To be ruthlessly honest, compared to RISC architectures like ARM, MIPS and especially PowerPC, Transmeta's price/performance/efficiency is nothing to write home about. No big loss to the competitive market if they get swallowed up by AMD...and it would give AMD a bigger club to go after Intel on the low end with.
SoupIsGood Food
Re:Its a good idea... (Score:2)
Linus and future employment (Score:2)
Yeah, I doubt that he'll have much problem. :-) I can only imagine what his resume looks like (not that he'd probably need one).
Seriously, just for starters I imagine IBM would hire him (and the rest of the kernel bigwigs) in a heartbeat, as would Intel, as would AMD, as would Motorola, as would any of the big linux solutions providers, as would any of the national laboratories in several countries (educational or "other"), as would the technical branches of several government's intelligence agencies, etc. etc. etc.
So yeah, I don't think that any of these folks would be out beer money if {Transmeta|VALinux|redHat|whatever} fell over and died tomorrow. ;-)
--
Fuck Censorship.
Re:Its a good idea... (Score:1)
Steve Jobs [tuxedo.org] is Linus Torvalds [amazon.com]? Nah... don't think so. :-)
--Joe--
as per uusual (Score:1)
Pathetic
---
>80 column hard wrapped e-mail is not a sign of intelligent
Re:All this A buys B, B buys C is *bad* (Score:1)
Re:I wouldn't worry about Linus. (Score:1)
Cheers...
Not so dumb (Score:1)
Re:why would this effect linux? (Score:1)
Re:I wouldn't worry about Linus. (Score:3)
I can do much better than that. The full source code to Digger is available at http://www.digger.org and has been ported to Unix using the SDL library. I think the "remastered" DOS version has been known to work under DOSEMU, but without sound.
Re:Interesting, but not likely... (Score:1)
Actually, this makes sense (Score:2)
Yes, this does make more sense for Transmeta than for AMD, but a merger is still plausible. AMD bought NexGen for their cool technology, and wound up with the P3-whomping Athlon. They could do it again.
Reality check? (Score:1)
Joe McGuire
tinfoil.music
Re:Its a good idea... (Score:1)
Re:No... It's a good move... (Score:1)
So large ? (Score:1)
AMD + Transmeta ~= AMD
(neary equals to)
IMHO
Re:Tired of Linux Stories, and /. Rumors (Score:1)
Re:Tired of Linux Stories, and /. Rumors (Score:1)
Re:Well... (Score:2)
The Transmeta 'code morphing' is not actually that great an intellectual leap! OK, now I've annoyed the fans of it I'll try to pacify them.
It takes a few reasonably well used components - hardware, firmware, and software - and then implements them together without requiring any software component.
I'm sure that the brains behind it were fully aware of the intricacies of :
- Microcode.
I have it from an Motorola engineers mouth that there were 8 different test 68000 implementations that differend only in microcode - one of which would decode x86 instructions (endianness-corrected, of course)!
The x86 of course also having plenty of micro code too (for stuff such as 'REP', 'LOCK' etc.) x386->uops has of course been around for 5 years in itself.
- Emulation:
What did Apple do when they moved over to PPC? Got some UK wizzkids to implement a JIT emulator (rewriting in machine code, so only translated one) for the old 68K code that was already out there.
What did Dec do when they got in bed with Microsoft? Implemented a feedback-optimising
emulator. It runs great, and the feedback is so good it runs better on only the second time round (further improvements are very small, as it's optimised all the most important bits already). This Fx32 emulator/translator/optimiser is _brilliant_. The engineers were rocket scientists.
OK, so Transmeta have rolled it all into one, with the cherry on top being the fact that the self-optimisation is in real-time.
Yes it's cool. Yes it is one of the coolest recent developments. However, don't forget the Motorola/Intel/DEC guys who've done some equally brilliant work in their time.
(Now is not the time to bring management-oriented Intel muck-ups into the argument. Even brilliant engineers produce broken products if the budgets and deadlines and other requirements are unfavourable. So praise where it is due, they do have _some_ good engineers you know...)
FatPhil
-- Real Men Don't Use Porn. -- Morality In Media Billboards
And the generation after that. . . . (Score:2)
Re:overpriced (Score:1)
FatPhil
-- Real Men Don't Use Porn. -- Morality In Media Billboards
Re:Its a good idea... (Score:1)
Re:Transmeta PowerPC emulation? (Score:1)
Apples biggest problem currently is MHz...
"
MHz is as useless a benchmark as 'bogomips'.
If Apple want to market their machines as _performance_ machines, then they ought to use a _real_ benchmark.
PPC isn't even really a 'braniac' chip. The HPPA chips are the real brainiacs, and they out-perform x86 chips at numerical simulations even with a 4:1 clock ratio. Look at the SPEC/FP. Until recently HPPA were the only systems to come anywhere near Alpha performance (they beat it occasionally).
Apple can't compete in the meaningless MHz war - so they should _educate_ the market. 15 years ago noone even knew (or needed to know) what a MHz was (excpet that it was on their radio dial). Trying to beat a MHz claim with a 'photoshop filter' claim is _not_ a valid parry, it's an equally misleading statistic. Tell people to demand 100 SPEC/FP from their next system.
At the moment, SPEC is as independent a benchmark as you can find, so they ought to compete in that arena - if they can.
(It's not truly independent, but as I say it's about as good as you can get)
FatPhil
-- Real Men Don't Use Porn. -- Morality In Media Billboards
Re:Intel, AMD and Linux (Score:1)
This could be 3 or 4 different sentences, I thought my runons were bad!
:-D
Re:Interesting, but not likely... (Score:2)
Now one thing that _could_ do well for the future is dumping lots of resources into fab technology, but I'm sure they're already allocating as much as they think is wise. Excess too has diminishing marginal returns, as well as possibly being counter-productive.
What AMD needs (and they know this) is to break into different markets, such as that of the server. They've already got their business strategy, but here's an interseting point of view.. Their next x86-64 design tentatively will be a dual core. If this succeeds, then one thing will be clear from the outset - heat will be a _massive_ issue. Not to mention suceptibility to defects. If the chip is slated for the server market, then they have an interseting possibility. Focus on UNIX platforms where the code can all be garunteed to be recompiled for the new architecture. Then remove much of the compatibility hardware in each core and potentially resort to software emulation... Getting closer and closer to RISC at each juncture. In a server with Gig's of memory, we can afford to take the crusoe approach of segmented processor memory. The main advantage of such an emulated approach is that code isn't required to be recompiled (a la 32bit emulation mode), and the hit hopefully wouldn't be as bad as what the Italium seems to make it. Additionally, I speculate as to how much CPU surface area can be reduced. In the situation of specialized 128-node NUMA super-computers, the savings MIGHT be noticable.
Essentially this is an argument for AMD's futher path. It doesn't really suggest that a purchase of Transmeta today would be wise - what does their crusoe archetecture really do for AMD? My guess is that the personelle themselves might be of value.. Again, pure speculation.
-Michael
Interesting w.r.t. Itanium IA-32 emulation (Score:2)
There's a heck of a lot of x86 software floating around out there, so well-performing emulation (in my opinion) will be critical to acceptance by large businesses.
Re:Interesting, but not likely... (Score:1)
The reason given by AMD is that they want to offset some shares coming on the market through an employee equity incentive plan. However, could this be a cover for a plan to do a stock swap with Transmeta?
Re:AMD's interest in Transmeta... (Score:2)
Don't get me wrong... I think that the power consumption of the Transmeta chips would be great for something like the Athlon...
Ok, let's get something straight. The crusoe was designed from the beginning with power-consumption in mind.. This shaped the sort of trade-offs made.. Namely number of functional units, or ports on registers. Obviously the use of VLIW was key, so on and so forth. The Athlon was designed with features such as maximal redundancy to reduce latency (such as stalls for lack of execution units). I'm much more up-to-speed on the design of the Athlon than the crusoe so I won't continue the comparison.
From what I remember, the crueso's sort of truely novel features (as opposed to design trade-offs) had to do with MHZ ramping and possibly choice of transistor design (which again probably trades speed for efficiency). Since Transmeta didn't actually fab the chips themselves, I'm doubting that they actually innovated with the fabrication process.. Sooo the only thing crusoe technology could bring to the [mobile] Athlon are things like MHZ ramping. And I doubt the specifics of this are worth $3.4B.
Future designs, on the other hand as referenced in my other comment [slashdot.org], might provide interest - though mainly through IP, which would probably be cheaper to lease than outright own.
Incidently, I love all this hype about OS transparent code-morphing.. I don't understand why a big server couldn't just require something like Alpha's Fx32. I'm sure it needs to tie directly into the OS, but wouldn't a server want it's OS to be natively compiled? Transparency isn't always the answer.
-Michael
Re:Well... (Score:1)
In a sense, they are moving the microcode into software, similar to how RISC architectures simplified hardware instruction decoding already.
You're right, FX!32 did sound like an similar (brilliant) idea as well, but I didn't hear nearly enough about that, even at the time. (and Microsoft still dropped NT on Alpha, and Linux is native on the Alpha...)
I'm sure Intel does have some good engineers, too; and many of them are probably pissed off, or working for someone else... Intel doesn't have the best track record for working with their employees--in fact, I have yet to hear something good about them in this respect.....
I am a big fan of x86 assembler, though. Intel was in the right place at the right time, they made something cheap that the market adopted, and they implemented all the real functionality starting with the 386. It's really another Microsoft success story. But I like programming in CISC-style assembler better, even though x86 doesn't have (pre/post) increment/decrement built-in as a memory addressing mode...
---
pb Reply or e-mail; don't vaguely moderate [ncsu.edu].
Wow... talk about creative journalism... (Score:2)
Did you think people wouldn't notice?
-Todd
---
Re:Transmeta PowerPC emulation? (Score:3)
This is incorrect. The 1.5 GHz Pentium 4 gets a SPECfp2000 score of 549, handily beating the fastest HPPA (PA-8600 at 552 MHz) which scores only 400. The only chip faster than P4 at FP is the 833 MHz Alpha which scores a barely better 571 (and the P4 still beats the fastest Alpha at SPECfp2000 - and is 1/4 the price to boot!).
Combine this with the old Disney/Apple rumor (Score:2)
Imagine how powerful Tux would become! Hell, he'd even start showing up on lunch pails ;-)