AMD Says That K10 Problems Will Only Be Cured by K11

[Epsilon84]

If 45nm K10 tops out at 2.8GHz as this article seems to suggest, http://www.nordichardware.com/news,7688.html then I can see why AMD is willing to write K10 off, even at this early stage.

[savantu]

According to NVIDA Chief Scientist David Kirk , AMD is dead :

..."AMD has been declining because it hasn’t built a competitive graphics architecture for almost two years now—ever since the AMD/ATI merger. They’ve been pulling engineers [from the GPU teams] to Fusion, which integrates GPU technology onto the CPU. They have to do four things to survive, but I don’t think they have enough money to do one thing.

“The first thing they have to do to compete with Intel is the process technology – they have to build the new fabs. The second thing is the next-generation CPU technology. The third one is the next generation GPU technology—we’re going to invest one billion dollars in here this year and they need to invest on the same level to keep up with us. And then the fourth thing is they say the future is going to be this integrated CPU/GPU thing called Fusion, which there’s no evidence to suggest this is true but they just said it. They believe it and they’re now doing it.

“So they have to do these four multi-billion dollar projects, they’re currently losing half a billion dollars per quarter and they owe eight billion dollars. Their market cap is about three billion, so it’s hard to see where the future is in that picture. Really speaking, they’re going to have to pull not one, but several rabbits out of the hat.”

http://www.bit-tech.net/hardware/200...rk-interview/6

[AliG]

I agree with that article, that's not demeaning amd like many sites have, but rather putting it out straight what needs to be done and what they're issues are. Its quite simple, to catch up with intel in the next 4 or 5 years (yes 4 or 5, anyone that thinks they can do it in just a year or two should planning on saving up for a long time), they need to match their production, price, and offer something with about the same performance. And with larabee coming out, that means also they'll need to spend a lot of $$$ on gpus too, so yeah it's definitely going to be hard on them unless they have a breakthrough along the line from either a major investor or some new technology they develope

[perkam]

Translation:

"We have the money, but we still can't avoid being killed off if both AMD and Intel decide to integrate the Motherboard chipset and graphics into the CPU."

Seriously, I would love for Nvidia to kick Intel arse, if only because I don't like monopolies, but this guy is just being pompous. Not to mention, spending billions of dollars on a new architecture does not always guarantee success and if they get even a single hiccup, it can have severe consequences, as ATI's R500 fiasco showed to all of us.

Perkam

[nn_step]

According to NVIDA Chief Scientist David Kirk , AMD is dead :



http://www.bit-tech.net/hardware/200...rk-interview/6

1) AMD doesn't need new fabs, just needs to upgrade current
2) There has not been any revolutionary new features in CPUs in the past decade. OoO, uOPs, and hardware virtualization have been around for over a decade.
3) They may have a point here
4) Fusion is about combining 2 finished products, not exactly a hard task unless they plan on trying to optimize it

[Dimitriman]

I'm sorry AMD but if you don't sort at least your manufacturing and shipping before end of 2008, there will be no 2009-2011 for you.

Good luck with that.

[villa1n]

1) AMD doesn't need new fabs, just needs to upgrade current
2) There has not been any revolutionary new features in CPUs in the past decade. OoO, uOPs, and hardware virtualization have been around for over a decade.
3) They may have a point here
4) Fusion is about combining 2 finished products, not exactly a hard task unless they plan on trying to optimize it

3) the amount of money thrown at tech, doesnt necessarily equal the best tech....
if that were true, the highest capitalized company would always possess the best tech, which is obviously not true. albeit more cap, makes it easier to explore, it guarantee's nothing.

[NickS]

it's going to be a long 20 months for AMD.....but I hope they pull something together...I'm bored with the all Intel scene at current

Me too, it's gotten really slow lately because we haven't had anything new... at all.

[iddqd]

I am so sad by this that I am actually mad. AMD needs to shut their trap and disappear into the abyss if they're going to just continue being so damn piss-poor. Unless they are going to say "We just killed Intel with the processor we're releasing today", they can f' off. I loved buying AMD, supporting the little guy. But they are so pathetic now, I have no sympathy.

When they can't have the performance crown, they can always go back to being the bargain basement. There's a lot more money in the mid-range segment than there ever will be in the enthusiast segment.

Besides, less and less people care about how fast the CPU is. It's becoming more or less irrelevant at this point.

[savantu]

1) AMD doesn't need new fabs, just needs to upgrade current

Interesting; why do they bother with Albany New York ?

And what can you upgrade on FAB 30 ? It's an empty shell.

2) There has not been any revolutionary new features in CPUs in the past decade. OoO, uOPs, and hardware virtualization have been around for over a decade.

Run ahead execution , scout threads , integrated accelerators , etc.

Look at SUN Rock or Niagara ; Intel Larrabee , there's plenty of new work going on.

4) Fusion is about combining 2 finished products, not exactly a hard task unless they plan on trying to optimize it

They are as different as different can be.You need to share resources like caches and IMC ; optimize for the same process ( SOI in this case , GPUs are made on bulk )

[ubuntu83]

( SOI in this case , GPUs are made on bulk )

High-K metal gate.

[hecktic]

as funny as this line sounds... its dead on the spot!!!

"They have to do four things to survive, but I don’t think they have enough money to do one thing."

[safan80]

and C2D will only be cured by C4D

[ubuntu83]

and C2D will only be cured by C4D

Where's C3D? Nehalem?

[savantu]

High-K metal gate.

??

[ubuntu83]

??

Here you go :

High-k/metal Gates. As part of AMD’s Continuous Transistor Improvement (CTI)
approach, AMD has the option to introduce high-k/metal gates into 45nm production
to further enhance transistor performance. The “gate first” approach, developed with
IBM, is designed to provide a simpler, less time consuming way to migrate to high-k
metal gate technology and secure benefits that include improved performance and
reduced power consumption

AMD

[savantu]

Here you go :


AMD

AMD's 1st iteration of the 45nm process is going to be low-k SOI.I'd say there is 90% chances that Swift will come on SOI instead of High-k/metal gate.

[ubuntu83]

AMD's 1st iteration of the 45nm process is going to be low-k SOI.I'd say there is 90% chances that Swift will come on SOI instead of High-k/metal gate.

Let's wait and see.

[3NZ0]

If I have to buy Intel, I won't be buying computers. I hate my Intel q666 (no I didn't mistype), it moves like a bloated whale, but sure it benches great. Whoop-de-frickin' doo. AMD or ANYONE get your together.

Processors don't move.

What problems are you suffering with it?

If you went with a phenom you'd have a slower, hotter processor that uses more power.

[_Lone_Wolf_]

AMD's 1st iteration of the 45nm process is going to be low-k SOI.I'd say there is 90% chances that Swift will come on SOI instead of High-k/metal gate.

How much of a drop in replacement would metal gates be?
I assume new masks would be needed for the relevant layers but would the logic layout require a rework (ignoring any efforts at optimisations)?

[savantu]

How much of a drop in replacement would metal gates be?
I assume new masks would be needed for the relevant layers but would the logic layout require a rework (ignoring any efforts at optimisations)?

It's totally different.The change is fundamental ; Intel worked for 8 years on high-k and metal gates until they had it right.AMD waits for IBM to perfect the technology.Basically their know-how is extremely limited.I'm not in the position to say how much the logic layout is changed when you switch processes but by the comments I've seen at more technical forums , it's no walk in the park.Basically , today's GPUs and CPUs are marred to the design process ; even the EDA tools take into account the performance of the process.

Not only that , AMD/IBM are using immersion at 45nm , a new set of problems.

Thirdly , it's a matter of cost , which is non trivial to develop high-k/metal gate.

All of the above combined meant AMD missed the train for 45nm high-k and will probably update the process in late 2009 as a stop gap until 32nm.IMO , I'm not sure AMD will bother with high-k and metal at 45nm.

[[XC] riptide]

Processors don't move.

What problems are you suffering with it?

If you went with a phenom you'd have a slower, hotter processor that uses more power.

QFT^

[[XC] riptide]

Besides, less and less people care about how fast the CPU is. It's becoming more or less irrelevant at this point.

Thats rubbish.

[Aussie FX]

It's totally different.The change is fundamental ; Intel worked for 8 years on high-k and metal gates until they had it right.AMD waits for IBM to perfect the technology.Basically their know-how is extremely limited.I'm not in the position to say how much the logic layout is changed when you switch processes but by the comments I've seen at more technical forums , it's no walk in the park.Basically , today's GPUs and CPUs are marred to the design process ; even the EDA tools take into account the performance of the process.

Not only that , AMD/IBM are using immersion at 45nm , a new set of problems.

Thirdly , it's a matter of cost , which is non trivial to develop high-k/metal gate.

All of the above combined meant AMD missed the train for 45nm high-k and will probably update the process in late 2009 as a stop gap until 32nm.IMO , I'm not sure AMD will bother with high-k and metal at 45nm.

[_Lone_Wolf_]

Thanks for replying.

Not only that , AMD/IBM are using immersion at 45nm , a new set of problems.

Could you elaborate on what problems?

I'm just a hobbiest with a interest in the technology and my limited knowledge is gleaned mainly from picking out those who know what their actually talking about in various forums and firing off the odd question. From that what I gather is immersion lithography is essential by the 32nm node so in utilizing it with the current node AMD/IBM is if anything benefiting for the additional experience gained. While it can produce lower yeilds relative to dry lithography again, from what I've read, AMD are actually doing very well in this regard. In contrast there have been yield issues suggested at Intel's reliance of dry double patterning. I know in Australia at least availablity for Penryns have until very recently been light on the ground and generally above list price.