[SemiAccurate] AMD kills Wichita and Krishna

[STaRGaZeR]

SemiAccurate has been hearing that the rumors that AMD’s (NYSE:AMD) Witchita and Krishna are on the chopping block for a few weeks. Over the last few days, that has been confirmed by several different sources.

Yes, you read that right, the follow on to Brazos/Ontario/Zacate has been knifed, there will be no 28nm Wichita and Krishna. Why? No one is saying definitively, but we would guess that it is due to timing of GloFo’s 28nm SHP process. As you can see from the picture below, taken from a GTC 2011 slide, the process is set for production in mid-2012. That means there won’t be significant volumes until Q3/2012 for a chip that is due to be replaced at the end of 2012.



GloFo process roadmap

This means that Wichita and Krishna would have a shelf life of ~6 months, likely less. No OEM is going to pick that part up and design machines around it for such a short time, it isn’t financially viable. Brazos is already set for a mild tweaking to get a little more speed and little less power use, so that will have to hold the fort for an additional six months.

If things go well, the additional resources needed to bring Wichita/Krishna to market will be diverted to the successor, and that will in turn be pulled in. This may lead to the post Wichita/Krishna part in Q4, maybe a bit before, and all will be happy. If not, things go pear shaped, people leave, and OEMs send impolite emails to their sales contacts. We won’t know which way it goes for a bit longer, but either way, the successor to Brazos is now quite dead.S|A

Source: http://semiaccurate.com/2011/11/15/e...a-and-krishna/

[Mechromancer]

AMD is caught between TSMC and GloFo. They got problems...

[savantu]

There's more to the story :

One of problems ahead is the fact that a certain upcoming chip got cancelled as well, but not because it underperformed, but rather because the chief architect moved to a certain semiconductor vendor that is entangled in a patent battle with Apple. He was joined by fellow engineers and there was nobody left to head the program.

http://vr-zone.com/articles/amd-s-pr...#ixzz1dsVFyqPV

[EniGmA1987]

^^^^^^
"moved" aka was let go in the big layoffs the new CEO feels is necessary? lol

[informal]

I've just found out who is that chief architect.. It's a complete disaster. The guy who left AMD's CPU team in Austin is no other than Brad Burgess,former chief architect of successful Bobcat core... Complete failure on AMD's part if they let go that guy. He is now working for Samsung as a chief CPU architect (which is understandable).
http://www.linkedin.com/pub/brad-burgess/26/aa9/93

[savantu]

The CPU team in Austin is in disarray, lots of people left. Outside of Austin, what CPU design centers to they have ? Boston and India ?

Who exactly is left at AMD to fight the relentless tick-tock ?

[sergiojr]

The only possible explanation: Bobcat program is cancelled in favour of a Bulldozer-based low-power processor.

[Manicdan]

The only possible explanation: Bobcat program is cancelled in favour of a Bulldozer-based low-power processor.

bobcat was awesome, bulldozer was not. i hope thats not their reason. if anything id rather seem them scale bobcat up to have more clocks and higher IPC to cover most of the Llano area too, then work on something thats built for killer high end perf (even if it the power consumption kills too)

[sergiojr]

bobcat was awesome, bulldozer was not. i hope thats not their reason. if anything id rather seem them scale bobcat up to have more clocks and higher IPC to cover most of the Llano area too, then work on something thats built for killer high end perf (even if it the power consumption kills too)

Bobcat is awesome compared to Atom. Bulldozer is bad compared to SB. If you compare Bobcat to SB then Bobcat is a total failure. If you add to this Intel's intention to kill Atom and replace it with Haswell-derived core, then you'll see that there is no place for little Bobcat in the future of x86 computing.

[undone]

The only possible explanation: Bobcat program is cancelled in favour of a Bulldozer-based low-power processor.

Another explaination: board force those executives to stick to ARM. So all existing low power product got canceled, bobcat would still alive in 2012 but dead in 2013.

Bobcat give amd a comfortable year, but when ARM arrive, whole market would get bombed.

[sergiojr]

Another explaination: board force those executives to stick to ARM. So all existing low power product got canceled, bobcat would still alive in 2012 but dead in 2013.

Then Brad Burges won't leave AMD to develop ARM processors at Samsung if he can do it at AMD. And AMD decision to complete drop any architecture besides Bulldozer completely explains this departure and departure of his engineering team.

[Manicdan]

Bobcat is awesome compared to Atom. Bulldozer is bad compared to SB. If you compare Bobcat to SB then Bobcat is a total failure. If you add to this Intel's intention to kill Atom and replace it with Haswell-derived core, then you'll see that there is no place for little Bobcat in the future of x86 computing.

how is it even comparable to SB? different goals and markets and they never really competed. if SB somehow was competing with bobcat, then atom would not need to exist. and we dont have to add in haswell at all, its necessary for chips to be replaced with either an optimized core or a new arch that fits current/future needs. to believe that bobcat will never evolve means its about to be replaced with something better, but what does amd have thats better than bobcat for sub 20w platforms? its nothing right now.

[SKYMTL]

Really bad news if true. Low cost APUs could have (and some believe SHOULD HAVE) become AMD's bread and butter in a post Bulldozer world. It also reflects quite poorly upon GF's ability to roll out new manufacturing processes in good order.

Remember, Burgess was also instrumental in other APU development like Llano so naturally this raises questions about the integration of Trinity and Kaveri as well.

[god_43]

There's more to the story :



http://vr-zone.com/articles/amd-s-pr...#ixzz1dsVFyqPV

....ugh huh.

[Oliverda]

I think that Wichita and Krishna designs should be already finalized.

I've just found out who is that chief architect.. It's a complete disaster. The guy who left AMD's CPU team in Austin is no other than Brad Burgess,former chief architect of successful Bobcat core... Complete failure on AMD's part if they let go that guy. He is now working for Samsung as a chief CPU architect (which is understandable).
http://www.linkedin.com/pub/brad-burgess/26/aa9/93

According his profile he left the company in August, before the layoffs.

Anyway I didn't know that he was working on Tejas which was cancelled at the end.

Really bad news if true. Low cost APUs could have (and some believe SHOULD HAVE) become AMD's bread and butter in a post Bulldozer world. It also reflects quite poorly upon GF's ability to roll out new manufacturing processes in good order.

Remember, Burgess was also instrumental in other APU development like Llano so naturally this raises questions about the integration of Trinity and Kaveri as well.

He has never worked on Llano or Trinity.

[bamtan2]

we know that if amd switches to ARM, they will go from #2 in x86 to #5 or #6 in ARM. and that isn't a good trade. we know that, but do they? or are they blinded, thinking they can become #1 in ARM as long as intel isn't there? I think AMD might not survive long enough to find out...

[Sparky]

Oh, can it with the "AMD's gonna die" crap (not just you, to everyone). It gets old.

[JohnZS]

I am certainly not trolling here (or at least I am trying not to), but where it not for Fusion (which by the way I think is an amazing platform), AMD would be in a serious PR mess. A lot of people had their hopes riding on the BullDozer (which turned out to be the Faildozer) and it seems that ever since that fateful and long put back release nothing but bad news has been coming out of the AMD camp.
Even though I am an Intel Legion Hammer Member I feel sorry for those AMD enthusiasts as they have not really had a decent High End chip since the days of the AMD-FX A64 and Opteron 939.
I truly hope that the next revision of Bulldozer is competitive with Intel's HIGH/ENTHUSIAST end otherwise something will have to give.

I am not saying it is the end of AMD Sparky, but I am saying it is the start of bad times for AMD. Let us all hope that Fusion continues to flourish and make huge strides and that the 7000 series graphics cards are at least half as good as the hype.

Sadly with AMD (as of late) if their products were even half as good as the hype then Intel would have NOTHING to offer enthusiasts, but sadly as we have found out with BullDozer, AMD are great at hype, but fail to deliver at all, yet alone on time.

John

[Aberration]

Oh, can it with the "AMD's gonna die" crap (not just you, to everyone). It gets old.

It is the best thing that could happen. If AMD finally failed, then the government would have to finally to take a serious look at anti-trust with Intel.

[JohnZS]

and nVidia, by that statement I meant you cannot solely blame intel here, nVidia's TWIMTBP programme and what not means that a lot of developers will only do a lot of testing and development on nVidia hardware. So to solely blame Intel is a bit harsh (but I understand what you are saying, especially when it comes to those dodgy deals with Dell back in the days of the Presshot)
John

[SKYMTL]

He has never worked on Llano or Trinity.

Wrong. He and his team were directly responsible for several key technological implementations within the design of modern APUs.

[Mechromancer]

Wrong. He and his team were directly responsible for several key technological implementations within the design of modern APUs.

Your post made me think of this scene:
Attachment 122395

[Mechromancer]

Wrong. He and his team were directly responsible for several key technological implementations within the design of modern APUs.

Your post made me think of this scene:

[SKYMTL]

Your post made me think of this scene:
Attachment 122395

Let me be completely transparent here. I wasn't trying to be confrontational but I guess I could have worded my initial response better. Nonetheless, the team in question helped with many key differentiating architectural advances built into ALL of AMD 's APUs.

[Darakian]

Why would you axe the only line that is making money?