Oh well, there goes AMDs last hope of getting back into the desktop CPU market!

Question is did AMD let him go, or did he go of his own accord?

If it's the former then more fool AMD, if it's the later then I guess there's not much they could've done about it, at least he finished ZEN core before he left.

I wonder if this has anything to do with the rumor that Microsoft will buy out AMD

Probably he was too expensive for amd at the moment, so he joined to make zen and now that it is as far as I know ready (design wise) he is leaving again.

Maybe he will join again if amd manage to get some profit with zen.

I have high expectations from zen and it happens to come out around the time I plan to upgrade my current system, I hope it will be good because I don't really want to jump ship to intel, but I will do it if I can't get a cpu that with good per core performance (in a good price ofc).

Either he rejoined to design and see Zen through to the final stages, or AMD are now in a very bad position.

Looks like those share prices that went up last week may well be headed down again then

shaithis
Oh well, there goes AMDs last hope of getting back into the desktop CPU market!

Well it aint all bad, zen will already be finished and taped out by now, and zen 2 will likely be just be an improved arch so the meat of the work is done.

No-one lasts for ever, so if AMD have any business acumen, they will have a team of designers (he probably didn't design them on his own!) under his leadership, and they will have another team leader in waiting. A change of leadership and/or key members is not always a bad thing. There is the danger of being stuck in a groove, and a change can inject new ideas, or allow other ideas to flourish.

Whether that is the case here, only time will tell.

shaithis
Oh well, there goes AMDs last hope of getting back into the desktop CPU market!

How so? Zen's design was finished a long time ago, the thing already taped out months ago in fact. Zenv2 and maybe even Zenv3 are also finished in the design phase already. That's how these design processes work. His role is finished as an architect, and this is probably the last uarch he will ever make. He's in his late 50's and he's very well-off financially. I'd wager a guess and say there's a 99% chance he left of his own accord.

To wit, his original run with AMD back in the late 90's was much briefer than his latest run.

There are so many possibilities as to why this may have happened it is probably unwise to read anything into this yet.

If he was fired it would have been bad news for AMD as it would have indicated a strong possibility that Zen was a dud, but leaving of his own accord could be because he disagreed with AMD's future direction or simply because he felt he has completed his project and has other challenges or offers on the table that appeal more.

People can look at his work history:
https://en.m.wikipedia.org/wiki/Jim_Keller_(engineer)

He spent only 1 year at AMD before he left and that was years before the Athlon 64 was released,around the time the original Athlon was out.

His second stint at AMD was three times longer.

It only took a few seconds on Google to find it out.

Yeah you're right CAT, I originally overreacted at this article, but I should have Googled… guess I fell for the Hexus clickbait :p

It makes sense - if the meat of his work is done, he probably wants to move on to other projects and/or AMD can't afford to keep him employed and idle.

Really hope K12 and Zen are home runs for AMD…

Here's to Zen hitting the ground running next year! It does make me wonder if Zen will be i7 beating in a few areas.. Definitely a Zen based system for me if so. I suppose there is nothing stopping Keller coming back to AMD again at some point in the future also.

Sumanji
It makes sense - if the meat of his work is done, he probably wants to move on to other projects and/or AMD can't afford to keep him employed and idle.

Is it done? The article mentions Zen, but nothing of the upcoming ARM core. That is probably the interesting bit of work, and it is interesting work that keeps top engineers at a company.

I think x86 has peaked, if you have a Haswell or later I doubt you will ever have a reason to upgrade other than device failure. ARM is interesting, V8 is a modern arch and has the possibility to go faster than x86 (heck it is already up to i3 levels). I shall be very interested to see where he has jumped ship to.

nitro912gr
Probably he was too expensive for amd at the moment, so he joined to make zen and now that it is as far as I know ready (design wise) he is leaving again.

Maybe he will join again if amd manage to get some profit with zen.

I have high expectations from zen and it happens to come out around the time I plan to upgrade my current system, I hope it will be good because I don't really want to jump ship to intel, but I will do it if I can't get a cpu that with good per core performance (in a good price ofc).

That makes sense. I doubt this would affect zen in any way as it takes years to design these processors and he was there for most of it from the beginning.

Damn it! I know his work will already be done on Zen, but AMD will definitely feel his loss.

Can't AMD get rid of Roy Taylor instead of Jim Keller? (I know it's not comparing apples to oranges, but Roy Taylor needs to go given the recent fiasco with cherry-picking review sites to send preview R9 Nano units to and accusing multiple review sites of giving unfair reviews.)

I don't know much about the guy but he strikes me as being someone who gets very interested in the creative challenge but gets bored thereafter. As Ken Mishima points out, Zen is no longer a project, it's a roadmap. Uncharted terrain is exciting for an explorer, maps are the kiss of tedium.

I don't think that this is necessarily the last micro-architecture that Jim Keller will work on. The 50's is way too young for a creative mind to stop thinking, dreaming and planning. Keller's financial strength isn't a reason to stop, it's the power to be highly selective.

Assuming that the leaving is an amicable and reasonably mutual end of needs, my reckoning is that he'd be quite open to returning to AMD if they have something sufficiently new and juicy in due course, and if something else hasn't already piqued his interest at that time.

devBunny
Assuming that the leaving is an amicable and reasonably mutual end of needs, my reckoning is that he'd be quite open to returning to AMD if they have something sufficiently new and juicy in due course, and if something else hasn't already piqued his interest at that time.

A high end multi threaded ARM V8 design is probably about as good as it gets out there in CPU land. If that didn't keep him on board, then either AMD ARM is canned or he has found something *really* interesting and I want to know what that could be.

“Legendary” in the title seems a bit strong as well, though I'm sure he wouldn't mind lol.

I expect Nvidia could do with some help on Denver, which is well into the “interesting design” territory. Wonder how deep Nvidia's pockets are ATM.