Intel Announces New Leadership Roles, Business Unit Changes

[coder]

Well, since new architectures take years to develop and put into the shelves, is difficult to estimate exactly what cards he was responsible for. Personally I think the first ones were models launched 2 years after he joined AMD. Since they launched a couple of under-performing models while he was there, he ended taking the blame for other people's mistakes.

He was "vice president in Visual Computing" from 2013 and then given the title "senior vice president and chief architect" in 2015. So, Vega should be all him, if not also Polaris. He probably also had a big hand in RDNA, to be fair.

Is like what Jim Keller said in his recent interview: success has many names. Failure only has one.

First of all, I never said he was the only one deserving blame, or even the main one. As an outsider, I just look at the timing and mystery around his departure and speculate as I'm free to do. My speculation about such things is worth little more than the bits it's written in.

[M@GOid]

You might be thinking of Raja as the guy who went straight from AMD to Intel. The circumstances of his departure from AMD were never entirely clear. IMO, the only thing that saved Vega from being a flop was the crypto boom. Even Polaris wasn't very competitive, except for its price. In terms of perf/W and presumably profit margins, Nvidia's Pascal ran circles around it (most GTX 1060's used a 200 mm^2 die @ 16 nm, whereas RX 570 and 580 used a 232 mm^2 die @ 14 nm). With a track record like that, I could imagine Raja's departure might've been a "mutual decision".

Well, since new architectures take years to develop and put into the shelves, is difficult to estimate exactly what cards he was responsible for. Personally I think the first ones were models launched 2 years after he joined AMD. Since they launched a couple of under-performing models while he was there, he ended taking the blame for other people's mistakes.

Is like what Jim Keller said in his recent interview: success has many names. Failure only has one.

[coder]

I thought Jim left AMD way earlier than that, and then re-joined to work on Zen. Then he got poached by Intel.

He did the K7, then worked at a few other places (including Apple), then went back to AMD to do Zen. His second stint at AMD was just over 3 years, and he'd gone to Tesla before Zen even launched. His time at Intel came after Tesla, where he says they built all the HW & SW for autonomous self-driving in just 18 months. He's now back to building AI chips, at a startup helped fund (Tenstorrent).

You might be thinking of Raja as the guy who went straight from AMD to Intel. The circumstances of his departure from AMD were never entirely clear. IMO, the only thing that saved Vega from being a flop was the crypto boom. Even Polaris wasn't very competitive, except for its price. In terms of perf/W and presumably profit margins, Nvidia's Pascal ran circles around it (most GTX 1060's used a 200 mm^2 die @ 16 nm, whereas RX 570 and 580 used a 232 mm^2 die @ 14 nm). With a track record like that, I could imagine Raja's departure might've been a "mutual decision".

[edwaleni]

https://en.wikipedia.org/wiki/Jim_Keller_(engineer)

Jim Keller joined DEC in 1982, and worked there until 1998, where he was involved in designing a number of processors, including the VAX 8800,^[1] the Alpha 21164 and the Alpha 21264 processors.^[3][4] Prior to DEC, he had worked at Harris Corporation on microprocessor boards.^[1] In 1998 he moved to AMD, where he worked to launch the AMD Athlon (K7) processor and was the lead architect of the AMD K8 microarchitecture,^[17] which also included designing the x86-64 instruction set and HyperTransport interconnect mainly used for multiprocessor communications.^[3]

In 1999, he left AMD to work at SiByte to design MIPS-based processors for 1 Gbit/s network interfaces and other devices.^[4][12][18] In November 2000, SiByte was acquired by Broadcom,^[19] where he continued as chief architect^[9] until 2004.^[3]

In 2004 he moved to serve as the Vice President of Engineering at P.A. Semi,^[3][11] a company specializing in low-power mobile processors.^[4] In early 2008 Keller moved to Apple. P.A. Semi was acquired by Apple shortly afterwards, reuniting Keller with his prior team from P.A. Semi.^[6][17] The new team worked to design the Apple A4 and A5 system-on-a-chip mobile processors. These processors were used in several Apple products, including iPhone 4, 4S, iPad and iPad 2.

In August 2012, Jim Keller returned to AMD, where his primary task was to lead development of new generation of x86-64 and ARM microarchitectures called Zen and K12.^[15][14] After years of being unable to compete with Intel in the high-end CPU market, the new generation of Zen processors has restored AMD's ability to do just that.^[3][13] On September 18, 2015, Keller departed from AMD to pursue other opportunities, ending his three-year employment at AMD.^[20]

In January 2016, Keller joined Tesla, Inc. as Vice President of Autopilot Hardware Engineering.^[21]

In April 2018, Keller joined Intel, where he served as Senior Vice President.^[21][22][23] He resigned from Intel June 2020 officially citing personal reasons.^[24] Though later it was reported that he left over a dispute on whether the company should outsource more of its production.^[25]

Jim Keller joined Tenstorrent as CTO in December 2020.^[26]

[schmidtbag]

Also, I think Jim left AMD in like 2015, when AMD was still bleeding cash. He went to Intel in like 2018, after doing the self-driving ship at Tesla. The situation at AMD is certainly much better, today.

I thought Jim left AMD way earlier than that, and then re-joined to work on Zen. Then he got poached by Intel.

[andre30correia]

no big news intel is a much bigger company than amd

[coder]

I was watching a interview with Jim Keller. The team he had at the time he left AMD was 2,400 people. The Intel team he had was about 10,000 people. That shows how big Intel is.

The key is focus. If you don't overextend yourself, you can do a lot with a smaller team.

Intel has a lot more products, in a lot more markets and segments. Plus, Intel has their own fabs, and it's not clear if or how many of those folks (or at least the process research folks) were under him, but I think some.

Also, I think Jim left AMD in like 2015, when AMD was still bleeding cash. He went to Intel in like 2018, after doing the self-driving ship at Tesla. The situation at AMD is certainly much better, today.

[onlyLinuxLuvUBack]

Does it matter? Ten monkeys is not necessarily smarter than five monkeys and they usually create more mess as well. Linus law applies still, but 2400 may be more than enough already.

or twenty major ashholes are not as good as one dark helmet?

[waxhead]

I was watching a interview with Jim Keller. The team he had at the time he left AMD was 2,400 people. The Intel team he had was about 10,000 people. That shows how big Intel is.

Does it matter? Ten monkeys is not necessarily smarter than five monkeys and they usually create more mess as well. Linus law applies still, but 2400 may be more than enough already.

[M@GOid]

I was watching a interview with Jim Keller. The team he had at the time he left AMD was 2,400 people. The Intel team he had was about 10,000 people. That shows how big Intel is.