News [intel] Jim Keller resigns from Intel

[podspi]

There are people that excel in life without relying on others to motivate them. And there are those that only advance because competition requires them to do better.

So, no its not required.

Excess competition also causes social issues. People are under tremendous pressure to lie and cheat to get ahead. What about damage to the environment or to our health? Did you read about the pressure Apple's Chinese factories put on workers to meet deadlines? Employees are scrutinized by fractions of a second for their productivity.

I agree with most of the things you've said, except your conclusion. Competition is necessary because not all (probably most) people AREN'T self-motivated. When you have a big company like Intel, the chances of most people being self-directed is pretty low. In my experience, most of the people who are intensely self-motivated end up in academia.

As far as stress, the environment, that's our culture. We can still have business competition without all that if we (society) decided to.

And you were right, I completely agree and fully admit I was very much wrong and misguided by that line of thinking back then.

It is great to see you! Although we didn't always agree, our conversations were always interesting, and your knowledge and experience really added to the board. I know that due to time zone differences you're not around much anymore, I hope you are doing well and enjoying Taiwan!

 

[Magic Carpet]

Sure as hell, Intel will be fine without Jim, albeit their retail sales figures haven't been exciting in some regions as of late. At any rate, it would be interesting to re-visit this thread in ~2 years. Also, welcome back IDC.

 

[DrMrLordX]

It has all those things keller talked about though.

Ehhhh

Lakefield's layout was teased to the press last year. I have no idea why it was delayed as long as it was, but the design/layout seemed mostly finalized at that point. Keller came onboard in 2018. Allegeldy, Keller's assignment was the product formerly known as Ocean Cove, or . . . whatever was meant to replace that.

Intel will (eventually) have a lot of products that have "all those things keller talked about". Doesn't mean he worked on all of them.

 

[Kenmitch]

And you were right, I completely agree and fully admit I was very much wrong and misguided by that line of thinking back then.

Holly molly! You are the real Idontcare and not some ghost of Intel past?

Welcome back! The forum did miss you!

 

[Kenmitch]

I'm guessing he recently watched the movie Titanic and put 2 and 2 together.

Wish him the best of luck on his future adventures!.Hopefully all is well on the home front and it's just one of those time to move on scenarios.

 

[IntelUser2000]

I agree with most of the things you've said, except your conclusion. Competition is necessary because not all (probably most) people AREN'T self-motivated. When you have a big company like Intel, the chances of most people being self-directed is pretty low. In my experience, most of the people who are intensely self-motivated end up in academia.

(This opinion of mine isn't going to be popular in a tech forum)

I'm saying, this push for the latest technology could be curtailed for the greater good of humanity.

In the long term, the market is quite logical. A monopoly that does not deliver an improvement will lose customers quickly, and it'll be forced to self correct, or die.

Yes most of us need that push, I agree. If we can start over, it should be about maintaining harmony among people as priority number one. Yea, I know its a utopian vision.

As far as stress, the environment, that's our culture. We can still have business competition without all that if we (society) decided to.

I don't see this as being possible. I think that extreme push that Apple is giving to its employees are needed to get a product out that good in that short of time. Few years ago they were touting some mere few dozen micron level variations for its latest product, and I immediately connected it to the terrible conditions at their factories.

CDPR, a company that's lauded by many gamers as being very favorable to them and making super high quality games are also known to have a very high turnover rate for employees. Some articles were talking about how the work is extremely strenous and the pay is below average for the industry.

It seems not only modern technology requires that you be best-of-the-best to create them, you gotta be absolutely be pushed to your limits. On his interview about his work at Intel, Pat Gelsinger, the former CEO of Intel stated that he had to work 80 hours+ per week! Imagine the impact on his family! The joke among CEOs/managers is that they work 24/7.

What if that decision that society makes is telling everyone your gadgets can't be as high quality? How would you mandate business owners to follow that, especially when running a company is probably the hardest thing you can do as a human being? Too many regulations and you also run into risks of killing businesses. Your product sales will naturally suffer as well.

 

[VirtualLarry]

On his interview about his work at Intel, Pat Gelsinger, the former CEO of Intel stated that he had to work 80 hours+ per week! Imagine the impact on his family! The joke among CEOs/managers is that they work 24/7.

(waxing a little political now)

Perhaps THIS is the reason that CEO pay is so high, because the opportunity cost in today's hyper-competitive world, leaves little time for family, and then, what then?

 

[IntelUser2000]

Perhaps THIS is the reason that CEO pay is so high, because the opportunity cost in today's hyper-competitive world, leaves little time for family, and then, what then?

I don't know how others do it but Intel revealed their reason for justifying CEO pay. Someone asked bluntly why the CEO is getting paid several hundred times of a regular employee.

They said it made sense to pay him such an amount after calculating out all the responsibilities and impact he made for the company.

Bringing it directly on topic,

Exclusive: Internal Memo Shows Murthy's Remarks Over Jim Keller's Departure, Details About New Structuring, Raja Koduri To Head Architectural Roadmap

You might have seen Intel's declaration that Jim Keller has left the company, effective immediately, and we have more details for you. Jim Keller was one of the masterminds behind Intel's new architectural efforts and had been with the company for two years when he announced his resignation due...

wccftech.com

Internal memo says its a personal issue.

I know its WCCFTech and the reputation it has. But its a different viewpoint to most articles which are merely copies of each other with few paragraphs changed.

 

[CakeMonster]

Intel used to do tick-tock with a new chip every year and there were always architectural changes even on the tock (or tick, I can't remember which was which). I was always looking forward to the yearly update and see how they tweaked the design to slightly increase the IPC. This went on up to like 2015 with Skylake and them boom it stopped. The steady yearly architectural changes in the 10 years up to that were larger than what we've seen between 2015 Skylake and 2020 10xxx chips. True, they have added more cores, which was certainly more effective than say Sandy-->Ivy, but why did they stop with the tweaking? If they had announced that "Hey guys, we are gonna focus on squeezing out more cores and efficiency and keeping the exact architecture until we have a new completely overhauled architecture in 3 years", then fine, I'd understand it, but since they never did it does not appear that they planned on ending up like this at all.

So, why did they stop tweaking the architecture on a yearly basis, or what made them stop?

 

[jpiniero]

Intel used to do tick-tock with a new chip every year and there were always architectural changes even on the tock (or tick, I can't remember which was which). I was always looking forward to the yearly update and see how they tweaked the design to slightly increase the IPC. This went on up to like 2015 with Skylake and them boom it stopped. The steady yearly architectural changes in the 10 years up to that were larger than what we've seen between 2015 Skylake and 2020 10xxx chips. True, they have added more cores, which was certainly more effective than say Sandy-->Ivy, but why did they stop with the tweaking? If they had announced that "Hey guys, we are gonna focus on squeezing out more cores and efficiency and keeping the exact architecture until we have a new completely overhauled architecture in 3 years", then fine, I'd understand it, but since they never did it does not appear that they planned on ending up like this at all.

So, why did they stop tweaking the architecture on a yearly basis, or what made them stop?

10 nm.

 

[IntelUser2000]

So, why did they stop tweaking the architecture on a yearly basis, or what made them stop?

Yes process.

They could have made an updated uarch on the same 14nm process but they were unwilling to admit that 10nm is going to be badly delayed.

 

[Ajay]

Intel used to do tick-tock with a new chip every year and there were always architectural changes even on the tock (or tick, I can't remember which was which). I was always looking forward to the yearly update and see how they tweaked the design to slightly increase the IPC. This went on up to like 2015 with Skylake and them boom it stopped. The steady yearly architectural changes in the 10 years up to that were larger than what we've seen between 2015 Skylake and 2020 10xxx chips. True, they have added more cores, which was certainly more effective than say Sandy-->Ivy, but why did they stop with the tweaking? If they had announced that "Hey guys, we are gonna focus on squeezing out more cores and efficiency and keeping the exact architecture until we have a new completely overhauled architecture in 3 years", then fine, I'd understand it, but since they never did it does not appear that they planned on ending up like this at all.

So, why did they stop tweaking the architecture on a yearly basis, or what made them stop?

The tick-tock program was directly tied to process development. The new architectures Intel was working on were directly tied to the expected electrostatic and density expectations of those new process nodes. That was their product development model. When 10nm failed to deliver, there was no plan B. Tick-tock became tick-tick-tick (Intel did manage tock like improvement on the client side by adding more cores for heavily threaded workloads - but server was stuck in the mud because of power use). Ice Lake was supposed to be the first CPU to get to 8 cores (a little bit b/4 Zen1, but on a better process than Zen1, IIRC). When 10nm failed to deliver and was so broken that it didn't respond well to normal process improvement techniques - that product development cycle completely fell apart . It appears that the shock of that failure caused some amount of paralysis and denial at first - something that was purposely withheld from public view. Intel has been on their back foot ever since then, though, being the 800 lb gorilla in the biz - they still own a huge proportion of the the x86 CPU market.

 

[RetroZombie]

I don't think Intel would ask him to serve a consultant role for 6 months for transition

Maybe it's a pause (some excuse) before he can go somewhere else like Raja Koduri?

 

[moinmoin]

A monopoly that does not deliver an improvement will lose customers quickly, and it'll be forced to self correct, or die.

Current Intel proves the opposite: 5 years of meager improvements and market and stock by the way of record revenue and profit told Intel that it did perfectly fine. Customers got too used to the monopoly to be prepared to move away.

 

[ondma]

Current Intel proves the opposite: 5 years of meager improvements and market and stock by the way of record revenue and profit told Intel that it did perfectly fine. Customers got too used to the monopoly to be prepared to move away.

Part of it is marketing and name recognition, granted. However, one has to admit (well not necessarily in these forums) that Intel had a huge performance lead, so it took AMD several years to catch up. Also by concentrating on HEDT and server, they essentially ceded the mobile market to intel until the 4000 series mobile chips came out.

 

[uzzi38]

https://twitter.com/i/web/status/1271691206192279553

Seems like it's actually personal reasons.

 

[ClockHound]

And you were right, I completely agree and fully admit I was very much wrong and misguided by that line of thinking back then.

It's true! IDC is Jim Keller! Around the time Keller is sucked into the 14++++++ nm void, IDC goes radio silent here. Jim shakes off the blue shackles and IDC posts again. A coincidence? I don't think so.

Coincidence or not, great to hear from you, Jim IDC. We missed you. Not that you care.

 

[IntelUser2000]

Current Intel proves the opposite: 5 years of meager improvements and market and stock by the way of record revenue and profit told Intel that it did perfectly fine. Customers got too used to the monopoly to be prepared to move away.

When someone dies, how long would they have been suffering before they finally expire?

Likely its many years and steadily deteriorating conditions before it happens.

Companies are similar. In case of Intel, they had record revenue because they found ways to increase the price of products sold. RCP for Xeons is 3-4x it was in the Broadwell days. For client, they said they have record Core(not Pentium, not Celeron) sales. They focused on markets such as gaming. Because they are aware volumes have decreased substantially and have to make it up somehow.

According to revenues alone, Brian Kraznich's team would have been the most successful management in their history.

 

[StrangerGuy]

Well, who needs pesky things like good products when the corps can just print free money on Wall St.

 

[Exist50]

https://twitter.com/i/web/status/1271691206192279553

Seems like it's actually personal reasons.

While that would be the best case for Intel, those particular "sources" aren't exactly known for their accuracy.

 

[jpiniero]

Yeah not buying it. Even if there was someone in the family sick he doesn't have to quit over it.

 

[mikk]

Yeah not buying it. Even if there was someone in the family sick he doesn't have to quit over it.

Strange statement. He would be not the first person to quit his job because one's wife got seriously sick.

 

[uzzi38]

While that would be the best case for Intel, those particular "sources" aren't exactly known for their accuracy.

I only quoted Francois because he said it publically. Someone else mentionned it first, and here's someone else that can verify as well.

 

[itsmydamnation]

I only quoted Francois because he said it publically. Someone else mentionned it first, and here's someone else that can verify as well.

very sad , its got to be proper bad if taking extended leave/ leave without pay etc isn't a viable option.

 

[Kenmitch]

Yeah not buying it. Even if there was someone in the family sick he doesn't have to quit over it.

With all the corporate arse kissing lately there would be zero reasons to literally quit your job.