AMD CPUs, SoC Rumors and Speculations Temp. thread 2

Page 19 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
Status
Not open for further replies.
Maybe his work is done and another company with more budget want him to take on a new exciting project? I'm more interested in where he goes than him leaving amd tbh
 
AMDs lost a few APU engineers recently as well. This means either Zen is basically finished design wise, and AMD is trying to cut costs, or the ship is sinking and everyone is trying to bail before performance numbers leak.
 


This is the same team that believed that found a bug on the ARM ISA, but further research showed that the bug was on the tools used by the AMD CPU team. This left the x86 engineers a bit "daunted" according to Keller.

 


His work isn't done. Only Zen was finished. Skybridge was canceled, Zen+ is not finished and K12 is not finished. The fact Keller leaves projects unfinished clearly points to serious trouble at AMD.
 


I hate to say it but I tend to agree... I think this more than anything screams Zen isn't going to perform anywhere near where it needs to.

Shame- hopefully another company will get involved, we really don't want Intel as the only desktop cpu supplier...
 
I dont think we can take the departure of Jim Keller, as a worring sign in regards to zen performance.

Basically zen core design is done, and the next iteration are probably also done at a high-level. That was what he was brought in for, and he probably gonna keep pursuing things in his careers.
 


Anyone can make mistakes, especially when your tools have issues. I don't know what you mean by "daunted", but from what I can interpret, Zen is coming along just fine and the Team behind it has confidence.

In any case, if you make such a blunder, of course that is going to resonate for a good while in your head. You're only unworthy if you can't get your act together and keep on moving.

In regards to the shake this announcement has made, a lot of people in comments that remembers/knows AMDs history say Jim Keller did the same with K8's dev cycle. He left for Apple before the actual CPU was in production. I think it's worthy or not such a long shot to think that Zen will be good. How good, it's about to be seen, but at least not a bust.

Cheers!
 
Yuka, I like your optimism, although the signs aren't looking good (the ati split for example looks to segregate that off so graphics can be saved in the event the rest of amd collapses).

I'm very open to Zen being good, agreed nothing is forgone conclusion until we see the results. Hopefully they'll survive intact long enough to actually release it though (recent financials are not pretty).
 
I wonder how the re-structure of the GPU division prepares it for a sell or spin off if that were the case. We don't have the particular details on how that is being done and what that implies internally for AMD.

Case in point, big companies love shuffling divisions up and down, changing directors, area managers and even VPs. I know the first reading is "oh, a sell is imminent because of economic woes", but in reality a shake-up is not uncommon when things are not working as planned. It is fairly common. And like I said, we don't have the particular details of what was done in terms of management shake. All we know is Mr. Koduri got the helm and will be calling the shots for the whole division (or so was meant to be understood?). Plus, selling the GPU division would be very very very near sighted. A cheap "quick buck" if you ask me that would make no monetary sense in order to turn a profit out of the stock value of AMD.

Cheers!
 
Yuka, I don't think the idea of the gpu division split is lining it up for a sale to raise money exactly. I think it's intended to protect the gpu division from the rest of AMD if it gets to the stage it's beyond saving.

It makes sense if you look at it that way- AMD isn't viable without Radeon, Radeon would most definitely be viable without the rest of AMD though.

I envisage that if AMD as a whole gets into the red too deeply, they would split off Radeon into an independent company, then fold AMD taking all it's debts with it...

Not an ideal situation, but better than letting the whole group collapse.
 
As a re-hire at the top of the CPU chain, Keller's latest project at AMD was to develop the next generation of high performance processors for AMD and to build a team around the concept of PC performance.
Click to expand...
Jim's work on K8 was done over 3 years before K8 shipped in 2003.
Click to expand...
Given his history of switching jobs to work on new CPU projects and his high level of skill which has allowed him to so freely move between companies
Click to expand...
He probably left because his job in the company is done, he wants new challenges, and AMD can't afford to keep him.

I think he was hired to get them a new CPU, train the team that will handle it, and that's it. The fact that K12 was his pet project could only mean that they thought they had the resources to give him the challenge he needed, especially using the opportunity of having such a good (and expensive) talent on the company.

I don't think it speaks doom about AMD's future, it's probably just routine for Jim Keller and an expected scenario for AMD.
 


But the work for which he was hired is not made. Only Zen is finished.



"Daunted" is the word used by Keller.



K8 was finished. His work at Apple was finished. But he is now leaving AMD without finishing the work for which he was fired.
 

I think we have different views in regards to what Jim Keller was brought in for. I don't think he was supposed to be to work on the next couple of zen iterations, but rather lay out a solid foundation, for his now former colleagues to improve upon.
 


Years ago, AMD's official hire announcement listed the projects Keller would develop at AMD. Two of the projects were canceled, but where is K12 for instance?
 


I agree it looks bad, but on the other hand K12 may be finished as well from the design side of things (it may simply be a case due to their limited funds AMD can only afford to tool up and manufacture one core at a time- Zen being more important to their existing market which they MUST stabilize in order to survive).

If that is the case (as it was said Zen and K12 were 'sister' cores sharing many common design elements), then K12 is waiting for it's slot in the manufacturing pipeline. Alternatively it could have been canned (they've gone very quiet on it I agree), either way I guess Keller wouldn't be required.
 

From my understanding, K12 and Zen were been develop in conjunction.
It is probably waiting. AMD did delay it to 2017, as of the last I have heard.


 


Both were developed in conjunction because required pin compatibility and inside for the Skybridge project. But the Skybridge project was canceled.
 

...
... so.... what? you're not saying anything about k12 here, just that project skybridge got canned. please finish what you really want to claim instead of leaving the rest up to people's imagination. :)
 


And cutting down development expenses.
And what does that say about K12? Nothing in my ears.
 
AMD Confident Zen Will Deliver on Performance
http://www.eteknix.com/amd-confident-zen-will-deliver-performance/
excuse the 40% ipc boost bait. :)

GlobalFoundries develops 7nm and 10nm technologies in-house
http://www.kitguru.net/components/anton-shilov/globalfoundries-designs-7nm-and-10nm-process-technologies-in-house/
there goes the neighborhood...😛
 


Still I think people are focusing on the wrong market. Zen needs to be competitive in the server market and while we on the consumer desktop might not see the performance advancements in Intels chips the server market can as they normally take advantage of new extensions and such,

And I think GF is going to have a lot of fun doing in house nodes. I am sure they will have a new found respect for the companies like Intel and Samsung who spend billions in R&D trying to get new nodes working.
 
this is all very interesting but I tend to agree with yuka I don't think jim killer leaving will have a huge impact with amd as they have been in development with zen since steamroller FX chips got canned. that's a solid 3 years with keller and they have to had made some good progress with him there.

on a side note... keller is not the gabe newell of microarchitecture. while yes he is good but he isn't better than an entire team of engineers working closely together! one man can only do so much and to get some help from an experienced guy for 2 years of training would make me a hell of a cpu engineer. have faith! the amd zen team will certantly finish the job and in my mind it will hit close to or right on its IPC goals of 40%. I mean look at what they are coming from. ipc of piledriver is on line with core 2 duo so it cant be that hard to boost it up to 40%
 
Status
Not open for further replies.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom