
AMD's Fab 30 Revealed 145
Harkids writes "AnandTech posted a
sweet article about a recent trip to AMD's newest chip manufacturing plant,
Fab 30 in Dresden, Germany. The article is more than just a walk around the building;
it includes juicy details of AMD's microprocessor manufacturing process, innovations,
and even has some Hammer info in it. A good read if you are interested in CPUs
or simply what AMD has up their sleeve."
wow (Score:2, Funny)
Re:wow (Score:2)
Re:wow (Score:1)
Re:wow (Score:1)
Re:wow (Score:1)
Re:wow (Score:1)
fab 30 (Score:3, Funny)
Re:fab 30 (Score:1)
Re:fab 30 (Score:1)
Of course, thier banners have been stripped because of NCAA violations, so it's arguable how really fab they were, but I digress.
Re:fab 30 (Score:1)
Of course, the fact that fab *anything* means something is due to the Beatle reference.
Re:fab 30 (Score:3, Funny)
Re:fab 30 (Score:2)
Slashdot: News for Sweaty Nerds, Stuff that doesn't wash.
*shudder*
Re:fab 30 (Score:1)
Re:fab 30 (Score:1)
Re:fab 30 (Score:1)
Re:fab 30 (Score:1)
chipmakers vulnerable.. (Score:4, Interesting)
This makes me wonder how vulnerable the chipmakers are... One good fire, flood or earthquake and we're without cpu's?
Imagine the sabotage posibilities... (hey Intel, for 1M, I'll wipe out amd ?)
Re:chipmakers vulnerable.. (Score:5, Informative)
The city's infrastructure was affected though, which may have caused some trouble for the people working at Fab 30.
http://www.amd.com/us-en/Corporate/VirtualPress
http://www.geek.co
However, floods and other unforeseeable events can of course cause a major ripple in the economic situation of an enterprise such as AMD. Sometimes even rumors suffice...
Re:chipmakers vulnerable.. (Score:2)
Re:chipmakers vulnerable.. (Score:2)
Essensially, yes... (Score:5, Interesting)
While the AMD fab is quite a bit away from the river, up in the hills, other natural or unnatural disasters could be pretty severe.
However, it's not like the marked would totally collapse, AMD may have one main plant, Intel has a couple, UMC or whatever AMDs partner was called has some and I'm sure there are more flash etc. plants that could be converted on a longer timescale.
Of course if you had coordinated gro... *stops* *pullls out his Men In Black-zapper and zaps any Al-Quida associates reading this*
Kjella
Re:chipmakers vulnerable.. (Score:2)
Earthquakes probably won't happen here; floods are not a great risk as the Elbe valley rises very quickly as you walk away from the river. The plant is not right at the river's edge so it will not have much trouble, I think...
Re:chipmakers vulnerable.. (Score:2)
Imagine the sabotage posibilities... (hey Intel, for 1M, I'll wipe out amd ?)
Suppose that both Intel and AMD engineering headquarters were destroyed, either accidentally or by someone deliberately trying to cripple the US tech market. With the extreme complexity of what those companies are working on, Qhow long would it take to recover? Quite possibly it would be the end of the x86 line, period.
There's much to be said to diversifying a bit, and not having all of our collective eggs in one basket. Yes, there are other CPUs out there, such as the PowerPC and MIPS, but think about how much we're really dependent on the x86 in so many ways.
Re:chipmakers vulnerable.. (Score:1)
Re:chipmakers vulnerable.. (Score:2)
Naturally, but it's more than just that. It put either company into a deep financial hole from which it couldn't recover. It could result in the loss of the engineers who really understand how things work and where the the next gen designs are headed.
Re:chipmakers vulnerable.. (Score:1)
Although 1 major Pacific NW Earth quake near portland (which, although not california is still an "Earthquake area" )and Bye bye Intel R&D. Of course the buildings are made to withstand 8.0 earthquakes. They were going to build stronger but found it cheaper to build a new plant in case of disaster.
Thank God! A story for true nerds! (Score:5, Funny)
it includes juicy details of AMD's microprocessor manufacturing process
Yes! YES! JUICY DETAILS OF PROCESSOR MANUFACTURE! YES!!
Slashdot is SAVED!!
Direct link (Score:5, Informative)
Nice to hear (Score:5, Insightful)
"There is one thing that's for sure; AMD has not done a very good job of telling the public exactly what they're capable of from a manufacturing standpoint. Intel has been talking about their manufacturing capabilities for quite some time now and has left the market with the impression that they are the only leader in the x86 manufacturing world. It is a shame because in reality, AMD has quite a few accomplishments of their own to talk about but it's just a matter of getting them to loosen up and let you all in."
It is nice of AMD not to scream about what they can do in their labs, but actually rely on their current products.
Re:Nice to hear (Score:5, Informative)
Re:Nice to hear (Score:2, Flamebait)
Re:Nice to hear (Score:1)
actually, you can add US movies, food, fashion, music, TV, constitution, business, law and education to that, too.
Re:Nice to hear (Score:1)
The problem ISN'T the Constitution, it's the nitwits that the uneducated, and uncaring, masses keep electing into office...people who are looking for ways to make end runs -around- the limits placed on them by the Constitution.
So instead of blaming the USC, blame yourself. And educate yourself about where your representatives stand on -all- this issues, not just the ones you care about.
For the proponents of Parliamentary-style governments: Fuck off. The US has been more successful both politically and economincally under our Constitutional Republic than -ANY- of your Parliamentary systems.
Re:Nice to hear (Score:2)
Re:Nice to hear (Score:1)
Re:Nice to hear (Score:1)
You ever think Canada's success might have more to do with the non-aggressive tendencies of the neighbor to their south, and the commensurate protections involved therein than it does with their socialist, populist parliament?
What is your beef with the U.S. Constitution. It's far from perfect, as is the government. However, it does effectively enable a continuous revolving revolution.
There is no hereditary Presidency. In fact, we've found that bastard degenarates can be President, in spite of their personal weaknesses.
I don't know where you find nepotism in the electoral college. Each of the seperate states has their say in how delegates are picked. The only time this changes is when the legislatures change it or, in the case of Florida, when their State Supreme Court thinks it should be a democrat.
Get off your high horse. It's great you've got it in for the U.S., but realize, most of us could care less. Leave us out of it and pay your bills when they're due.
Jackass.
Re:Nice to hear (Score:1)
Re:Nice to hear (Score:2)
Really. Without the Empire (and guys like Mark Sykes) we wouldn't have Palestine, Israel, Lebanon, Syria, Beirut, Iraq, Iran, Afghanistan and Pakistan.
Depends on how you measure success. Personally, I think the Empire was an abject failure seeing as it ultimately FAILED.
The U.S., on the other hand, has no Imperial aspirations. In fact, we really couldn't care about other countries (you call it ignorance, we consider it apathy.)
Nevertheless we have to keep going abroad and fixing your messes (see Chamberlain, Neville).
You can get on your high horse, however, Monty Python and Black Adder are not enough to justify your imagined sense of cultural superiority.
Re:Nice to hear (Score:1)
Re:Nice to hear (Score:1)
Re:Nice to hear (Score:2, Interesting)
Can you say SuSE? They don't seem to make or get nearly the noise level as a certian colored item of headwear. Yet their distribution, IMHO is vastly superior.
Re:Nice to hear (Score:3, Interesting)
Then how come Alpha is dead? (Score:2)
Nobody bought it.
Their sales people couldn't convince consumers that superior technology was worth the price. They generated no excitement! They hoped that the technology could speak for itself.
I've been told by a long time Alpha engineer that long time clients would call him directly asking about new products, bypassing the sales rep. At the same time Sun sales team was attacking like wolves and stole the high end server market.
The result is that Intel now owns most of Alpha technology including the engineering teams. The last Alpha EV7 system can do up to 256 MP, and 256GB RAM, all on high speed ultra-wide packet switched interconnect bus. HP has to play down its performance to justify moving to Itanium servers.
The first Itanium has problems with more than 4 MP. Thank you Mr. Capellas.
That's some scary stuff... (Score:5, Interesting)
Intel may have tried to make the guys in the bunny suits cute, but after learning some of the dangerous stuff they're around I came to respect them for sure.
Re:That's some scary stuff... (Score:3, Interesting)
3 days? That's it? When I worked for TI it was a full week... and you had to retake the course yearly.
You're not kidding about the chemicals though... nobody was allowed to wear contacts in the fabs where I worked because they used a chemical with the tradename Pirhana. It was used beneath the floor in the air cleaners, but if it ever backblasted then it would melt plastic - and thus your contacts to your eyes.
Realistically, fabs are incredibly safe as far as chemicals go. They have to be because of the incredibly nasty shit being used. You're much, much more likely to fall 20' onto concrete through an open floor tile then you are to get exposed to a dangerous chemical.
Re:That's some scary stuff... (Score:1)
Well, this was the class for folks who wouldn't be on the fab floor - and they still had to tell us all about that death stuff!
The one cool thing was that I learned how to read those diamond signs with the red, yellow, and blue. That's pretty cool.
Re:That's some scary stuff... (Score:2)
I don't recall the diamond sign bit... must've slept through that part
And I meant the statement you quoted as a bit more tongue in cheek than it came across.
Re:That's some scary stuff... (Score:1)
I'm remembering when I was working at an HP building that had some sort of chip etching facilities in it. I think that it was not for manufacturing, but for design (Corvallis, OR). Nonetheless, we were evacuated four times in one week because the chemical alarms went off. It turned out that the dumping station for the janitors' chemicals was situated in such a way that the air intake system for the fab area was sucking in the fumes from the cleaning chemicals and circulating throughout the air system! *Gulp* Even though it was not the dangerous chems causing the alarms, I had nightmares about my bones rotting away from the insides for days.
Re:That's some scary stuff... (Score:2)
I work at a fab now (that shall remain nameless), and one of the nastiest liquids we have around here is HF (Hydrofluoric acid). The stuff has a NASTY desire for the calcium in your bones.
Re:That's some scary stuff... (Score:1)
Yeah, that's the one that put the fear of God in to me. Not only does it mess up the calcium in your bones, but it also messes up the calcium in your heart tissue, And it takes just a relatively tiny amount on your skin to croak you...or at least that's what I saw on ER...
HF is scary because it is small (Score:1)
In other news... (Score:5, Funny)
Sure is a cheap fab (Score:5, Insightful)
The combination of the Fab and the Dresden Design Center (DDC) was said to require a $2.3 million investment, with close to $2M already spent...
At that price, any Tom Dick or Harriet could set one up!
Of course, the next sentence makes it clear they're talking BILLIONS, not millions:
and the remaining $300M due to be used by the end of 2003. For an advanced microprocessor fabrication facility like Fab 30 this is the going rate for start-up costs, which is a major part of the reason why there are only two big competitors in the desktop CPU market; with such high barriers to entry, it's very difficult to become a mass market competitor in the CPU business.
Re:Sure is a cheap fab (Score:2, Informative)
Probably thats where the confusion derives from.
Re:Sure is a cheap fab (Score:1)
slashdot math (Score:3, Funny)
By any chance were the /. editors on this tour?
(BTW here's a tip: Click on "Print this article" to see the whole article at once, ad-free, without having to wait for 7 pages of ads to download.)
Ad free? (Score:1)
And cheat the advertisers of their pound of flesh? Get rid of the adverisers and the alterntive is pay-per-view.
It doesn't cost look at the ads.
I'll bet the "skip commercial" button on your VCR remote is worn to a nub.
Re:Ad free? (Score:1)
it sure as hell is! and when i rip/encode dvd's to vcd i don't include any of the previews at all. and when i back up my vhs tapes i skip those too! bad consumer.. bad..
and my linux satelite tv software descrambler ensures that pay-per-view comes across quite nicely
they have a "print article" button for people to use. I don't think there's an EULA you have to click through to agree that you will actually print the article and not just read it thouroughly to avoid the cumbersome advertisements and even if there were, and it stood up as being legitimite in the highest and most ethical court in the land, does that make it moral?
Fab 30? (Score:5, Funny)
John, Paul, Ringo, George.. who are the other 26?
Re:Fab 30? (Score:2)
who are the other 26?
Spares for the first four, just in case.
Juice? (Score:1)
Check out the article on Hardware Analysis (Score:3, Informative)
http://www.hardwareanalysis.com/content/article/1
A bit short but they've got two pictures of the plant.
- hama -
Re:Check out the NEW article on Hardware Analysis (Score:1, Informative)
The most important part of the article... (Score:3, Informative)
SOI (Score:1)
I thought that the Prescott was going to be released on SOI, though maybe I got that mixed up with strained silicon.
Anybody know?
Why not buy Alpha from HP? (Score:2)
In addition, many still depend on the Alpha, and this would give AMD the server market it's always wanted, and has been trying to secure.
All this is not mentioning that the Alpha's perfomance has always been far better than x86. Just imagine what would happen if AMD came out with a 8GHz processor a year from now...
Just a little ranting from someone who does not want to see the Alpha go, and still remembers the potential they hold.
Re:Why not buy Alpha from HP? (Score:2)
In addition, many still depend on the Alpha, and this would give AMD the server market it's always wanted, and has been trying to secure.
All this is not mentioning that the Alpha's perfomance has always been far better than x86. Just imagine what would happen if AMD came out with a 8GHz processor a year from now..."
I think the problem with this is that Intel bought most of the Alpha engineers, from Compaq, before HP bought Compaq (or however that deal worked...) so AMD would have to do some serious work to get those people. However, from my understanding (and I may be completely wrong) Hammer is basically a RISC CPU that does CISCy x86 stuff. While this isn't new (I think as far back as 286 or 386) this has been what the chips looked like, translating the CISC code into mo-betta RISC code for the CPU to run through.
While the RISC architechture of Hammer is exciting (and hopefully a little easier to program for), I think the more exciting prospect is that of HyperTransport and what it will mean for bandwidth inside the system (as well as allowing for 'glueless' multiprocessing). Then again, when I mention that I find HyperTransport more interesting than Hammer, most people get a certain look on their face, one that says "You... suck." But I'm still excited about it.
Re:Why not buy Alpha from HP? (Score:2)
A lot of people would love to see Alpha continue, but this would be a issue for HPaq/Intel who have a lot of the rights at the moment and wish to drive high end customers in the direction of Itanium.
I guess developing a new Alpha isn't easy because of the diversion of resouces away from x86. It doesn't sell the numbers associated with x86 so it would take longer to pay for itself. AMD has already spent a bundle on Hammer, so I can't see them wanting to change their direction and although Alpha was an engineer's dream, it was a marketing disaster.
Re:Why not buy Alpha from HP? (Score:2)
Well, it has a higher margin than ia32, and how popular Hammer will be remains to be seen... It could well be that Alpha will be a better seller than Hammer. With Alpha, they would already have an installed base, and a guaranteed market. Not to mention that the Alpha has already shown that it can be several times faster than Intel, which would be a huge advantage.
Re:Why not buy Alpha from HP? (Score:2)
Alpha was great but unlike SPARC, it never had the numbers to get a low price for the high-end corporate workstation market. With better market saturation, it would have been a clear winner. At Digital, they had a saying "Digital make watches, don't they?" because of the poor efforts at corporate marketing. The Suns people were buying for the desktop were crap and badly engineered (lots of daughter boards, reducing reliability). However, people thought that the Suns were fast. Solaris was also not exactly easy to use (it has got better) and had a lot of problems whilst OpenVMS and Tru64 were boring (they worked). You bought 50 Sun workstations, then you went out and bought a Sun server or two to go with it. Sun didn't even have proper clustering in those days and failover was something you more or less had to do yourself. However, those Sun UltraSparcs were dirt cheap (low margin).
Alpha is nice, but I believe that Compaq and now HP have blown it through not understanding what they have. AMD believes that Hammer will be supported by Microsoft, so they have a market there.
Re:Why not buy Alpha from HP? (Score:2)
Linux, FreeBSD, NetBSD, OpenBSD... It sounds like AMD could get the rights to Tru64/DigitalUnix very cheap too. In addition, Windows 2000 is available for Alpha, so I wouldn't say they've dropped support.
Re:Why not buy Alpha from HP? (Score:2)
If HPaq are so dedicated towards Itanium, I can't see them wanting to let it go in the form where it may become a competitor to them. They may sell the hardware but would they sell Tru64? What commercial sense would there be for AMD to go for a chip that only will be supported by Open Source software. It would be nice, but very brave!
I would like it. I have an Alpha at home (running OpenVMS) and have written for Alphas in C, Fortran and COBOL and Bliss. My biggest client is still running Alphas for the server part of their application. Regrettably, they killed off the Alpha clients due to poor management of the development process and the low level of interest by their cutomers (who wanted Sparcs and Windoze, the latter because of the commodity cost of the platform).
Re:Why not buy Alpha from HP? (Score:2)
AMD is going a slightly safer route for their transaction, so let's look at someone like Intel since the issues are clear... Intel's switch to ia64 is an all-or-nothing proposition. They include no backwards compatibility, so they are already in the position of trying to convince customers to make the huge jump with no safety net.
So, since Intel's ia64 chips have no backwards compatibilty, they could just as easilly take an Alpha core, name it ``Itanium II" and sell it. The differences would be invisible to programmers anywhere above Bios-level... Since the Alpha has proven it's superiority to all other processors, Intel would get a great number of things from making a pure Alpha core.
AMD is a little more complex because they provided a bit of a safety net that eases trasition (backwards compatibility). They still face almost the same problems as Intel... They are treading on almost completely new territory, and their CPU could flop or skyrocket. However, to sell their 64-bit processor, they have to convince everyone to rewrite their programs. When rewriting programs, it doesn't matter if it is a Hammer you are programming for, or an Alpha. If machine code isn't involved, they wouldn't even realize there was a difference. I don't suspect backwards compatibility a feature that will help adoptation much, so Hammer really doesn't have any advantage IMHO, except it's mindshare as the followup to their current line of successful processors.
So, the only thing AMD and Intel's processors have going for them, is the one thing the Alpha never had going for it: marketing.
Re:Why not buy Alpha from HP? (Score:2)
Windows 2000 can be bought along with most new Alpha systems. It's not the latest version per say, but it's definately new enough for anyone wanting to run a Windows server.
I can't speak for Novell because I don't follow it very closely, but, at worse, an Alpha version of Novell would be just as easy to build as an Itanium/Hammer version.
Re:Hnuh? Can somebody clear this up? (Score:2)
I believe he was refering to the Windows NT series of Operating systems. NT3.1, NT3.5, NT4.0, Win2000, WinXP, etc.
No, I did not buy one... I did see that Wnidows 2000 came with a few of the workstation Alpha models. I was under the same impression as you until I saw that. Since they have chaged their layout, I can't find much of anything on their alpha page (http://compaq.com/alphaserver is the easiest way to get there).
**MORE INFO and PHOTO/VIDEO HERE** (Score:2, Informative)
http://www.hardwareanalysis.com/content/article/1
Re:AMD is dying! (Score:1)
Re:AMD is dying! (Score:1)
> How is that 4-5 months behind schedule?
Not that I want to help the troll to whom you respond, but the Hammer was initially projected for a 2001 release. Barton was initially slated for something like early to mid 2002. Of course they're late. But they're not so bad as the aforementioned troll mentioned. Their processors are far more competitive than, for example, when they had the K6 family. It is simply that it's been over a year since AMD's processors were definitively faster on the highest level for Microsoft Windows (they still have an advantage, iirc, in prominent Linux benchmarks). And AMD is only seen by the general geek public as "equal" to Intel if their processors are simultaneously unarguably faster and cheaper. Right now, they're merely almost as fast and generally cheaper, so people get confused as to whether or not the company still exists.
But I digress [fdisk.com]. The AMD processors in question, just like most AMD, Intel, Motorola, Sun, IBM and HP processors, are much delayed.
-JC
Re:I'd stay away from this one (Score:4, Funny)
Re:I'd stay away from this one (Score:1)
It's great to see all the amd fanboys yell, "MHz doesn't matter!" (which is totally correct) and then turn around and drool over a new 64 bit "extension" (I won't call it a full 64 bit chip). Show us a speed increase for desktop, server, or enterprise server applications.
Re:I'd stay away from this one (Score:1)
Regarding K8/Hammer performance features (Score:5, Interesting)
> chips, combined they'll give a speed increase of not inconsiderable proportions.
It is notable to, er, note that the former advantage helps (possibly considerably) towards both recoded AND legacy (eg, normal) programs, whereas the additional physical registers would require recompilation in order to show a benefit, which means that everybody but Windows users will get an immediate use outta that.
Other advantages of the Hammer? Well, not counting the 64-bit yunk (that WILL provide benefits, but I want to cover benefits that will help legacy programs, like Civilization III, The Sims, Unreal II and other antiproductivity applications):
Hyper Transport. That's not much on its own, but it essentially equates to a reduction in loss of bandwidth to the chipset and between processors when you add an additional processor. On the Intel setups, the processors share a set amount of bandwidth to the chipset, so putting eight chips on a 2.4GB/s bus means that you have each chip getting 0.3GB/s. The AMD setup theoretically lets each processor get that 2.4GB/s. Of course, that's in a perfect world, chip-level, but it probably amounts to some benefit. AMD's K7 family has similar advantages, which probably assists in explaining why they get higher performance at each given clock in mainstream applications (which at least somewhat depend on the memory subsystem) even though the AMD cpu to memory bandwidth was usually 2.13GB/s (now it's 2.67GB/s, unless you count stuff like the nForce, which has some extra memory bandwidth, but the extra benefit there is eaten up by the onboard video), whereas the Intel cpu to memory bandwidth was usually 3.2GB/s. Anyway, the idea is that HyperTransport will (on a hypothetical level) make it much easier to make n-way systems without either a tremendous performance impact or an expensive crossbar workaround setup thingy.
SoI. Silicon-on-Insulator. This is one of those things that'll help with the process technology. In the end, it'll probably offer a little bit of a frequency boost by making the chip a little cooler or something like that. I forgot precisely what SoI's primary benefits were. It's been months since I've even thought about it.
Stages: As detailed here [chip-architect.com], the K8 adds two stages to the decoding part of the instruction pipe. The decoding part of the pipe is probably rather complex, so you may see a pretty neat frequency boost over the K7 family without the problem of a huge branch mispredict penalty. The number of cycles that a cpu wastes when it makes an incorrect guess on a low level "if/then" statement is somewhat proportionate to the number of pipeline stages. The AMD K6 and (iirc) Cyrix 6x86 were the mack daddies of branch prediction, since their pipes had only five stages or so, so they only had to wait a few cycles when they zigged instead of zagged. The PIII and K7 had over ten stages, so they had to wait a lot longer, but other advantages (such as the larger and sometimes faster caches and more accurate predictors) in those processors over their predecessors did their best to overcome this disadvantage. The Pentium 4 has a crippling 20 to 28 (depending on the situation, and depending how the trace cache handles the situation, and whether or not you want to count it) stages. This means that it can hit amazing clock frequencies, but it'll get cranky and drowsy for twice as long when it makes a predictive mistake. How does it get away with this? Well, the trace cache does its best to assist, but it didn't really help as much as I think the designers were hoping. But for multithreaded programs and OSes, the SMT implementation on the more recent members of the P4 family, an implementation known as "Hyper-Threading", probably pretty neatly alleviates much of this problem by putting operations from other threads into the cpu whenever the currently running thread stalls on a branch mispredict. The K8/Hammer approach is just to add stages where they hopefully will have the most balanced, beneficial effect to frequency boosting while only minimally increasing the branch penalty. SMT would be nice, but it isn't nearly as critical a need as it is on the P4.
Wider memory access. On the Sledge Hammer, if AMD's plans are still the same as when I wrote this [jc-news.com], the memory controller (which is embedded onto the cpu) will access PC2700 memory in a 128-bit configuration (ignore the "126-bit" typo on the linked page -- I can't believe I didn't notice that when I typed it nearly a year ago!), which leads to a 5.3GB/s path to memory. That's damned good, though I really think AMD should have focused on 366/183MHz (equiv to "PC2933") or 400/200MHz (equiv to "PC3200") memory instead of the 333/166MHz PC2700 that came out over a year ago. Still, servers often use memory that's lower than bleeding edge clock in order to maintain reliability, so bleh. Still, 5.3GB/s isn't bad for a setup that isn't based on a shared bus.
Enhanced branch predictor. Well, that's if my notes from a year ago are accurate. If true, this'll probably overcome any mispredict penaly performance disadvantage from those abovementioned added stages.
Larger TLBs, TLB flush filter, etc.. This stuff will have itty bitty advantages on a per-clock performance basis, but every little bit helps.
Larger caches. Hey, I should look this up to see what they're planning on. Is it just 512KB on-die L2, or is AMD planning on bringing it up to 1MB L2? The interesting thing about AMD's designs is that the die is really small on each processor. Remember how AMD has gotten occasional fire for processors overheating? Well, aside from a stupid lack of shutdown diodes in the past, the real cause wasn't that the AMD processors used more heat than the Intel processors. They usually generated about the same amount of heat, often less, but their processor surface area was substantially smaller, which made the chips less expensive to produce and less likely to have defects. But when you try to push an equal amount of heat through half the surface area, you end up with a higher amount of heat per area, which equates to a higher running temperature. The funky thing about this is that you could just added a whopping huge amount of on-die cache. That'd increase performance while also increasing the surface area. But the heat production would not be substantially affected. So you'd end up with a lower temperature processor. So the Hammer will have a higher ratio of cache to processor units in the cpu, so it won't be as much of a fire hazard. Frankly, they should have put 1MB on-die L2 onto the Thoroughbred/AthlonXP.
Crap. I need to research more on the K8. It probably changed a lot since I went into hibernation. The interesting thing is that in the last half year, I've largely moved from being a Windows 2000 power user to a Linux coder (I still use both operating systems for several different purposes, but I'm talking primary usage). I stand to be in the group that benefits most from the Hammer when it comes out, since I'll be able to './configure && make' or 'qmake -project && qmake && make' most of the programs I use and/or develop. So I'll instantly see the benefit of those extra registers. ^_^
-JC
http://www.jc-news.com/
Re:Regarding K8/Hammer performance features (Score:1)
Yes, but... wafer isn't cheap. Double the size and you use twice as much silicon, but it's not just a linear cost relationship because of defects.
Assuming the defect rate of SRAM is as high as the rest of the core*, then you also double the defect rate when you double the size of the core. It actually costs 4 times as much to make a chip that has double the surface area under these assumptions.
*That may not be true, but surely the defect rate is not zero.
Re:Regarding K8/Hammer performance features (Score:1)
I don't think this is entirely true. You can't spread out SRAM throughout the chip as localized heatsinks - they usually (for the purpose of less area / easier-to-trace address / data lines - all of the on-die-cache/memory/storage are always grouped together on their own little corner. So while it would provide *some* benefit, I wouldn't think it solves all the problems, as the heat is locally produced (on the high-frequency switched parts) anyway.
besides; people misunderstand how small cache-transistors really are - or more precisely, how tightly they can be packed. While regular transistors are about the same size (shy ones that needs to drive many gates are bigger to provide more current), the ones on cache memory are packed together while the "logic" portion are always fudged because that's the only way you can run the interconnects (without painstaking human intervention - this is mostly spewed out by computer).
so... I'd say that adding cache is more for the purpose of adding cache, with the 10% cooling as a fringe benefit, instead of the other way around as you seem to believe it to be.
Re:I'd stay away from this one (Score:1)
Out of curiosity (Score:1)
Hammer Time! (Score:1)
Makes me say "Oh my Lord"
Thank you for blessing me
With all those cycles and two quicks gigz
It feels good when you know you're down
A super dope CPU from the D-town
And I'm known as such
And this is a CPU uh you can't touch
With apologies to everyone with any taste whatsoever.
Re:Another reason why AMD is better (Score:1)
Re:Another reason why AMD is better (Score:1)
Re:Good to see it built at last! (Score:1, Insightful)
Then there was talk of reviving it in 2000, only for the entire thing to be delayed once again until 2002, when the need for general expansion made the development of a new plant in Europe critical anyway. The old plans were dusted off, and now AMD has the plant it deserves.
The article says:
AMD was founded in 1969 and the Dresden Fab began mass production in 1999 thus covering 30 years, which is where the '30' in Fab 30 comes from.
The fact that the Dresden plant has been producing chips since 1999 indicates that you are just making up your facts in the hopes of getting some suckers' mod points.
Further your claims of 16384 48 bit registers are patently absurd. The x86 chipset has eight registers. Nowhere near the thousands you claim for this alleged Tombstone project, which as others have already indicated Google cannot substantiate.
Moderators, please think before you press the moderate button. This poster is clearly whoring for karma.
Re: (Score:3, Interesting)
Re:Good to see it built at last! (Score:2)
As an example of that in action, the AMD Hammer chips have an extra 16(?) general purpose registers accessible in x86-64 mode.
Score -6 Really Badly Done But Successful Troll (Score:3, Informative)
And Dresden has been running for quite some time, producing tons of 30nm copper interconnect (y'all do remember copper, right?) chips.
Still, Tiger's post got modded up to 5, so I guess that it's a phenomenally successful troll. It just irks me. It's like having an article on Microsoft's Windows code and having a comment about Microsoft switching over to gcc modded up!
--
-JC
http://www.jc-news.com/
Grove: So, what do you want on your TombStone, Jerry?
Sanders: Ummm
Grove: Aye, [shouting] TOMBSTONE PIZZA, CRAIG!
Re:Score -6 Really Badly Done But Successful Troll (Score:1)
That's twice this week I've seen a troll with +5... wtf is going on here?! LOL
Mod above to "-2 full of shit" (Score:2)
Re:Good to see it built at last! (Score:1)
Ok, in order:
Re:Good to see it built at last! (Score:1)
Also check out AMD Rumors [geocities.com] which covered the topic pretty well in 1999. Also Ars Technica [arstechnica.com] carried a good story about it in 2000.
Really great CPU. I gather the fact that Hammer was more capable and that the best performance they could manage out of a 333MHz TwoStone running in Intel mode was equivalent to a 266MHz Pentium II, sank the chip.
Re:I'm not interested if the chip supports DRM! (Score:1)