News Intel's Jim Keller Resigns, Will Assist With Transition

atomicWAR

Glorious
Ambassador
I'm sure Jim Keller will always have a seat at AMD =D.

He only needs to ask Lisa Su nicely and he'll be welcomed back with open arms.

Yeah that may be true...though my guess is Intel made him sign a non-compete agreement. Won't last forever sure but I'd wager it is in effect for at least a couple years or more. So we won't see that move likely happen anytime soon.
 
Yeah that may be true...though my guess is Intel made him sign a non-compete agreement. Won't last forever sure but I'd wager it is in effect for at least a couple years or more. So we won't see that move likely happen anytime soon.

Last I checked, non-compete agreements are not legally binding.

That said I do hope he's okay physically/mentally. He always had an uncanny ability to look at requirements data and figure out the best architectures to solve those problems. It's not so much thinking up new solutions as it is balancing your existing silicon to get the most overall performance for the next couple years. For example: Would a 10% AVX IPC improvement makes sense if it take 30% more silicon, 40% more power when only 2% of instructions use it?

I have read papers about how certain people at Intel have a tendency to push you out if you don't play ball.

Either way it doesn't matter. This will be a loss for Intel no matter which way it unfolds. I bet their stock will take a hit today.
 
Last I checked, non-compete agreements are not legally binding.

That said I do hope he's okay physically/mentally. He always had an uncanny ability to look at requirements data and figure out the best architectures to solve those problems. It's not so much thinking up new solutions as it is balancing your existing silicon to get the most overall performance for the next couple years. For example: Would a 10% AVX IPC improvement makes sense if it take 30% more silicon, 40% more power when only 2% of instructions use it?

I have read papers about how certain people at Intel have a tendency to push you out if you don't play ball.

Either way it doesn't matter. This will be a loss for Intel no matter which way it unfolds. I bet their stock will take a hit today.
Why would you even think that?!?
From the article:"his depth of experience designing heterogeneous architectures" the sunny+tremont CPU was announced yesterday,you even said yourself that he has an uncanny ability to look at requirements data and figure out the best architectures to solve those problems.
"Keller is known for relatively short stints at companies, typically leading turnaround efforts for roughly two years before moving on"
Keller was hired 2 years ago,april 2018.
https://www.tomshardware.com/news/amd-intel-jim-keller-hire,36963.html
There is zero reason for any movement in the stock price for completely normal and expected things that happen.
Keller is known as a leader of transformational efforts, and his depth of experience designing heterogeneous architectures played well to Intel's move towards multi-chip processor designs. In tandem with Raja Keller and Murthy Renduchintala, Keller was responsible for designing and aligning Intel's silicon portfolio under a new six pillar strategy that plays to the strengths of the company's IP.



Keller is known for relatively short stints at companies, typically leading turnaround efforts for roughly two years before moving on to other challenges. He served as Intel's senior vice president in the Technology, Systems Architecture and Client Group (TSCG) and general manager of the Silicon Engineering Group (SEG),

Last I checked, non-compete agreements are not legally binding.
Yeah but other companies are going to think really hard about hiring you if it gets out that you do stuff like that.
 
Why would you even think that?!?

Why would I think "they would suffer"? Look at what happened when Steve Jobs left. And Job was just a brilliant marketer, not an engineer.

Keller is considered a very valuable chess piece to have on your board. His career is a testament as such.

Because transformations aren't a one and done. Zen 2 had a lot of architectural changes over Zen 1. Zen 3 will have even more with a unified cache and IPC improvements. You have to aggressively keep at it. AMD did do this with 50% (Pile driver->Zen) ->12% (Zen ->Zen 2)-> rumored 15%+ IPC improvements (Zen 2->Zen 3).

Intel has a history of making like 3->5% IPC improvements with each generation from Sandy Bridge. That's 11 years of taking no risk. They didn't like investing in new architectures that much. It will be interesting to see if they can keep the ball rolling. (Or if they can get past their foundry problems. They are starting to look like GloFo with their string of failures compared to TSMC)


Yeah but other companies are going to think really hard about hiring you if it gets out that you do stuff like that.

Well then AMD would likely have a hay day of lawsuits with their execs leaving for Intel.
 

atomicWAR

Glorious
Ambassador
Last I checked, non-compete agreements are not legally binding.

That said I do hope he's okay physically/mentally. He always had an uncanny ability to look at requirements data and figure out the best architectures to solve those problems. It's not so much thinking up new solutions as it is balancing your existing silicon to get the most overall performance for the next couple years. For example: Would a 10% AVX IPC improvement makes sense if it take 30% more silicon, 40% more power when only 2% of instructions use it?

I have read papers about how certain people at Intel have a tendency to push you out if you don't play ball.

Either way it doesn't matter. This will be a loss for Intel no matter which way it unfolds. I bet their stock will take a hit today.

In the US....non compete agreements are legally binding as long as their duration and terms are not excessive. States vary but that is the general gist of them. Source below

 
Last edited:

Co BIY

Splendid
Sometimes someone is too good (and too valuable, sometimes too expensive ) to work at any one place too long.

Better to spread the genius around.

I think this is more true if you are a visionary, or process expert. Lay the vision, build the process and then "my work is done here".
 
In the US....non compete agreements are legally binding as long as their duration and terms are not excessive. States vary but that is the general gist of them. Source below


https://www.huffpost.com/entry/unde...compete-agreements_b_58af1626e4b0e5fdf6196f04
 
Why would I think "they would suffer"? Look at what happened when Steve Jobs left. And Job was just a brilliant marketer, not an engineer.
https://www.macrotrends.net/stocks/charts/AAPL/apple/stock-price-history
Their stock went from $50 to $350?!?!

Keller is considered a very valuable chess piece to have on your board. His career is a testament as such.

Because transformations aren't a one and done. Zen 2 had a lot of architectural changes over Zen 1. Zen 3 will have even more with a unified cache and IPC improvements. You have to aggressively keep at it.
Keller left AMD before ZEN 1 even released,keller left intel after the CPU he put together for them released,what's your point here?
 

atomicWAR

Glorious
Ambassador

Thanks for that. Went down quite the rabbit hole because of the information you supplied me with. So you are right save a handful of situations which would not apply here...like signing a non compete in a state where they are legal and job hoping from said state to California. And even then its not iron clad depending on the circumstances. Which as stated would not apply as his tenor was in California. Live and learn thanks for sourcing made the rabbit hole easier to climb into!
 
Last edited:
https://www.macrotrends.net/stocks/charts/AAPL/apple/stock-price-history
Their stock went from $50 to $350?!?!


Keller left AMD before ZEN 1 even released,keller left intel after the CPU he put together for them released,what's your point here?
That's the long term trend you cited on stock price. The key to Intel's long term health is if they actually keep the momentum going that he created. No one can say if that will happen. Hence there will be some doubt. Intel is not really seeing huge improvements yet because 10nm isn't offering them the power saving /speed they need, and 14nm obviously is at a dead end. Intel needs to go well beyond Sunny Cove and their current process nodes to retake the crown undisputed.

You see AMD's stock value kept going up after Zen 1 because they kept the momentum going. Combined with their node advantage and you can see why AMD is doing well.

https://duckduckgo.com/l/?kh=-1&uddg=https://www.ibtimes.com/apple-stock-plunges-steve-jobs-retirement-most-analysts-unfazed-304544
 
Last edited:

spongiemaster

Honorable
Dec 12, 2019
2,364
1,350
13,560
Keller left AMD before ZEN 1 even released,keller left intel after the CPU he put together for them released,what's your point here?
What CPU has Intel released that Jim Keller played a roll in developing? He was there for only 2 years. Anything he was involved with in the early stages should still be in the development pipeline.