Thursday, June 11th 2020

Jim Keller Resigns from Intel

Intel today announced that systems designer-extraordinaire Jim Keller has departed the company citing personal reasons. Whether or not this is a blow to Intel likely depends on how far Jim Keller brought their Technology, Systems Architecture and Client Group throughout his two-year tenure at the company whilst serving as its Vice President. The semiconductor and chip architecture world isn't being driven by Mr. Keller himself, obviously; there are a number of architects and designers that bring the industry forward through their concerted efforts. However, it's hard to look past Jim Keller's pedigree when it comes to doing his job - if anything, AMD's Zen architecture is a testament to that, and has put Intel in the place we now see it in the CPU world.

To fill in the void, Intel has announced a reshuffling inside their Technology, Systems Architecture and Client Group. Jim Keller will still be serving with Intel for the next six months as a consultant, thus easing the transition. Read the full press-release below.
Today, Intel announced that Jim Keller has resigned effective June 11, 2020, due to personal reasons. Intel appreciates Mr. Keller's work over the past two years helping them continue advancing Intel's product leadership and they wish him and his family all the best for the future. Intel is pleased to announce, however, that Mr. Keller has agreed to serve as a consultant for six months to assist with the transition.

Intel has a vastly experienced team of technical leaders within its Technology, Systems Architecture and Client Group (TSCG) under the leadership of Dr. Venkata (Murthy) Renduchintala, group president of TSCG and chief engineering officer. As part of this transition, the following leadership changes will be made, effective immediately:

Sundari Mitra, the former CEO and founder of NetSpeed Systems and the current leader of Intel's Configurable Intellectual Property and Chassis Group, will lead a newly created IP Engineering Group focused on developing best-in-class IP.
Gene Scuteri, an accomplished engineering leader in the semiconductor industry, will head the Xeon and Networking Engineering Group.
Daaman Hejmadi will return to leading the Client Engineering Group focused on system-on-chip (SoC) execution and designing next-generation client, device and chipset products. Hejmadi has over two decades of experience leading teams delivering advanced SoCs both inside and outside of Intel.
Navid Shahriari, an experienced Intel leader, will continue to lead the Manufacturing and Product Engineering Group, which is focused on delivering comprehensive pre-production test suites and component debug capabilities to enable high-quality, high-volume manufacturing.
Intel congratulates Sundari, Gene, Daaman and Navid as we begin the next phase of our world-class engineering organization and look forward to executing on our exciting roadmap of products.
Source: Intel
Add your own comment

114 Comments on Jim Keller Resigns from Intel

#101
londiste
Lindatje
yeeeemanTechnically, not true. 1065G7 has ~10% higher IPC than 3900x. Case re-opened and closed.
Uhmm the facts say`s no.
Do you have a good link where Ice Lake and Zen2 is compared apples-to-apples?
Posted on Reply
#102
DrCR
ARF, just FYI, attempting to find any understanding with regard to Microsoft Window's update process may be a fruitless endeavor. And it doesn't have anything to do with your CPU being x86.
Posted on Reply
#103
yeeeeman
londisteDo you have a good link where Ice Lake and Zen2 is compared apples-to-apples?
He doesn't but he's stating stuff he doesn't know about...
Look here: www.anandtech.com/show/14664/testing-intel-ice-lake-10nm/4
Last table you see IPC in INT/FP of Ice Lake is better than 3900X which is a best case scenario for Zen 2. Renoir will have lower IPC than 3900x because of smaller cache.
The fact that Intel has sh*t process in the 10nm and it doesn't allow them to make a good product out of Ice Lake for desktop it doesn't mean suddenly they lost their technical advantage or that Zen 2 is some miracle uArch that the world has never ever seen.
Zen 2 is a good uArch which is put in a very good light thanks to TSMC 7nm. If it were to be fabricated on 12nm, AMD couldn't fit this amount of cache and even if they could the power consumption would have been much worse. Ice Lake is the other way around. It inherits all the advantages of Skylake uArch (which btw is still decent even if fabbed on 14nm and you can check the power/performance number for the latest 10th gen parts here on techpowerup, you'll see that 10700/10500/10300 are pretty similar to what 3700x/3600/3300x achieve in efficiency and absolute performance) that is low latency between cores, low latency to memory but it is held on by the bad 10nm process.
I don't want to detract anything from AMD merits, because they do have the chiplet implementation and in general a very good/smart implementation all around. But I don't like this stupid, clueless hate speech towards Intel products only because they stalled progress, they sell old parts at higher prices than AMD or whatever.
We should discuss technicalities in an objective manner and analyze each product without bias. If you will look on what I said about 10700/10500/10300 you'll see it is true. They do keep up with latest AMD parts in power and performance. Worse pricing, but technically they are similar which is something given they are basically respinned 5 year old parts. I bet you that AMD couldn't do that with their Buldozer parts, push them and still compete.
You will also see that the table from anandtech shows clearly that Ice Lake with its Sunny Cove core is indeed better than Zen 2 all around. If you leave the hate behind for a second and just analyze you'll see the reality. But it is getting harder and harder these days for people to be objective and it is all just a childish fanboy talk with 0 knowledge about actual facts.
Posted on Reply
#104
Lindatje
@yeeeeman
An apple vs orange benchmark, as Anandtech itself indicates. Maybe read your own links. ;)

i rest my case.
Posted on Reply
#105
kapone32
ARFThe "up to" is wrong. It goes up to all the way up to 1.1 - 1.2 GB/s copying and pasting in the same folder.
For one SSD? What are you running that on? Surely not a SATA port as you have the fastest SSD I have ever heard of.
yeeeemanHe doesn't but he's stating stuff he doesn't know about...
Look here: www.anandtech.com/show/14664/testing-intel-ice-lake-10nm/4
Last table you see IPC in INT/FP of Ice Lake is better than 3900X which is a best case scenario for Zen 2. Renoir will have lower IPC than 3900x because of smaller cache.
The fact that Intel has sh*t process in the 10nm and it doesn't allow them to make a good product out of Ice Lake for desktop it doesn't mean suddenly they lost their technical advantage or that Zen 2 is some miracle uArch that the world has never ever seen.
Zen 2 is a good uArch which is put in a very good light thanks to TSMC 7nm. If it were to be fabricated on 12nm, AMD couldn't fit this amount of cache and even if they could the power consumption would have been much worse. Ice Lake is the other way around. It inherits all the advantages of Skylake uArch (which btw is still decent even if fabbed on 14nm and you can check the power/performance number for the latest 10th gen parts here on techpowerup, you'll see that 10700/10500/10300 are pretty similar to what 3700x/3600/3300x achieve in efficiency and absolute performance) that is low latency between cores, low latency to memory but it is held on by the bad 10nm process.
I don't want to detract anything from AMD merits, because they do have the chiplet implementation and in general a very good/smart implementation all around. But I don't like this stupid, clueless hate speech towards Intel products only because they stalled progress, they sell old parts at higher prices than AMD or whatever.
We should discuss technicalities in an objective manner and analyze each product without bias. If you will look on what I said about 10700/10500/10300 you'll see it is true. They do keep up with latest AMD parts in power and performance. Worse pricing, but technically they are similar which is something given they are basically respinned 5 year old parts. I bet you that AMD couldn't do that with their Buldozer parts, push them and still compete.
You will also see that the table from anandtech shows clearly that Ice Lake with its Sunny Cove core is indeed better than Zen 2 all around. If you leave the hate behind for a second and just analyze you'll see the reality. But it is getting harder and harder these days for people to be objective and it is all just a childish fanboy talk with 0 knowledge about actual facts.
Except for power draw or anything other than Gaming. Regardless of how you may want to put it there is no objective reason for a PC Enthusiast to buy into Intel at the moment. As we have no idea what Intel is really working on we may get a nice new series of chips between 2021-2023. The thing is though if you think AMD will remain stagnant in that same time frame you have been living under a rock. I believe even more now the the XT CPUs are just red herrings and this fall will have a chip that Intel will have no response to (for now).
Posted on Reply
#106
yeeeeman
Lindatje@yeeeeman
An apple vs orange benchmark, as Anandtech itself indicates. Maybe read your own links. ;)

i rest my case.
Whatever, I won't translate the information for you to understand. Believe what you want.
kapone32For one SSD? What are you running that on? Surely not a SATA port as you have the fastest SSD I have ever heard of.


Except for power draw or anything other than Gaming. Regardless of how you may want to put it there is no objective reason for a PC Enthusiast to buy into Intel at the moment. As we have no idea what Intel is really working on we may get a nice new series of chips between 2021-2023. The thing is though if you think AMD will remain stagnant in that same time frame you have been living under a rock. I believe even more now the the XT CPUs are just red herrings and this fall will have a chip that Intel will have no response to (for now).
What I say and what you understand.....
I said everything from a TECHNICALLY point of view, not from a price point of view.
I actually said a few times that Intel is not worth it because of higher prices but whatever.
Posted on Reply
#107
kapone32
yeeeemanWhatever, I won't translate the information for you to understand. Believe what you want.


What I say and what you understand.....
I said everything from a TECHNICALLY point of view, not from a price point of view.
I actually said a few times that Intel is not worth it because of higher prices but whatever.
Ok but you made the statement that if AMD had used 12nm they would not have had the ability to use that much cache (as if AMD didn't know) and you reference a site that has already been guilty of publishing false information. I am not saying that Intel is bad but they are not as good right now but let's keep in mind that their CEO just stated last week that "benchmarks don't matter".
Posted on Reply
#108
Imsochobo
R-T-BIf this years Ryzen releases have anything to say about it, no.



Mesh honestly works fine, it's just that it has higher latency, kinda like Ryzen chips. That's not the end all but it does hurt their "best gaming" claims because well, Ryzen already has them beat on every other front.
I used hot mess specifically cause mesh uses a whole lot of power.
Also it's not uniform dram latency access, nor is it uniform very uniform latency core to core, it has all of the disadvantages of zen and none of the advantages.
On zen you can atleast tell a process to stick to a ccx first and foremost, on skylake you're going in blind, far or near to dram (Controllers) as they are also multiple.
It also contributes to the crazy high power consumption, if you want it fast and not bottlenecking skylake cores it's chugging power.

Atleast it's more scalable than ringbus and can maybe be fixed and won't say it cannot work, just that current singular implementation isn't looking good.
Posted on Reply
#109
R-T-B
ImsochoboI used hot mess specifically cause mesh uses a whole lot of power.
I'm not sure you can blame the interconnect for the power part. That would I think, fall to the cores. But unsure.
Posted on Reply
#111
Flanker
R0H1TYou know what, get a better SSD. How about the upcoming true gen4 980 pro from Sammy?
I'd like to know wtf is going on with his PC. I did that update on my office PC with a 5400rpm WD green and it didn't take that long
Posted on Reply
#112
Juan Kerr
Quote: "Keller describes Intel's next big CPU core as being "significantly bigger" than "Sunny Cove," with its 800-wide instruction window, and "massive" data- and branch-predictors, to put Intel back on a linear performance growth trajectory between generations. Keller also commented on this being a "mindset change" at Intel, which over the past decade, only delivered minor IPC increments between generations, and focused on other areas, such as efficiency"

Mindset change? Good luck with that one. Pshhht. When your poop don't stink to begin with, there's no reason to believe that changing anything you do will result in anything positive. The cafeteria 1:1's in Oregon are testament to that.

His tenure was doomed from the get-go. My guess is he was in it for the almighty dollar, but when even that wasn't enough to keep him there, he left--- and assuming something else was going on in his personal life, why stay?
Posted on Reply
Add your own comment
Apr 26th, 2024 15:12 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts