Intel's Skylake Visionary Returns to Lead Client Chip Development
Intel closes ranks
Intel announced that it is bringing back another one of its famed chip architects, 28-year veteran Shlomit Weiss, as CEO Pat Gelsinger continues to rebuild the company's engineering roster. Weiss will lead all of Intel's consumer chip development and design. Weiss returns to Intel after a four-year stint as Mellanox/Nvidia's Senior VP of Silicon Engineering, where she ran the company's networking chip design group, a team of more than 1,000. Still, her achievements during her prior service at Intel are just as impressive.
During her tenure at Intel, Weiss received Intel's Achievement Award, the company's highest award, for her work developing the company's dual-core architecture. She was then entrusted with leading the team that developed Intel's famed Sandy Bridge and Skylake processors.
Intel's Skylake is perhaps one of the company's most famous chip architectures simply because the company shipped multiple iterations of the same design for six long years as it dealt with the fallout of its 10nm delays. Perhaps a testament to the potency of the design, Skylake played a big role in helping Intel survive an incredibly long delay between process nodes — the architecture spanned from Intel's sixth-gen to tenth-gen Core processors and multiple generations of server chips. Intel has ceded roughly 10% of its overall market share to AMD (~8% for server, ~10% for desktop/laptop) during that time frame, but that's surprisingly limited damage given the severity of the company's delays.
Weiss rejoins Intel as the Senior VP and Co-General Manager of the company's Design Engineering Group (DEG) and will be responsible for all of the company's client chip development and design processes. It appears that Weiss will assume the same role previously held by Uri Frank, who recently left Intel to lead Google's SoC development.
Weiss will work with Sunil Shenoy, another long-term Intel veteran that recently rejoined the company. Shenoy will co-manage the Design Engineering Group with Weiss in Isreal. We aren't clear on the division of responsibilities yet (he will likely lead the data center design initiatives), but we're following up with Intel for more details. Both will also work with famed chip architect Glenn Hinton, who also recently returned to Intel.
Intel's most famed architectures have long come from its Israel design teams. Gelsinger's continued focus on rebuilding the design teams comes on the heels of his recent restructuring of the data center group, which resulted in the departure of long-time DPG lead Navin Shenoy.
Gelsinger is obviously relying upon experience from proven Intel veterans as he rebuilds the company after it has suffered from years of missteps and the exodus of much of its brain trust. Weiss holds an M.Sc. with honors in electrical engineering and a first degree in computer science from the Technion - Israel Institute of Technology. She has registered several patents on microprocessor development and also founded the Intel Israel Women Forum.
Weiss commented on her appointment, "I am thrilled to return to the place that had been my home for 28 years, where I grew and developed professionally, as a manager and as a person. I have been following Intel Corporation's Pat Gelsinger, charting a new, bold strategy for the company, which I believe will accelerate the 'company's leadership. I will devote my energy to ensuring Intel continues to lead in hardware and chips."
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Paul Alcorn is the Managing Editor: News and Emerging Tech for Tom's Hardware US. He also writes news and reviews on CPUs, storage, and enterprise hardware.
-
Howardohyea I'm delighted to have the competition from AMD getting Intel to truly innovate. I hope Alder lake would be competitive thoughReply -
VforV Ok, ok... more good news from Intel, but I actually expect them to be competitive in 2023.Reply
Alder Lake will not beat Zen3+ (3D) and Zen4 in late 2022 will still win, but after that maybe in 2023 Intel can come back to lead? Maybe...
The best scenario would be AMD and Intel both trading places in lead, alternatively.
Anyway, thanks AMD for these past year and for waking up Intel! (hopefully) -
gargoylenest cant wait to see what are the direction line from intel: invest lot of money long term to innovate, while loosing money short term but insuring return long term; or ,like so many business, make minimal investment and R&D, sell to people old shiny tech with minor innovation by making them believe its much better than it actually is...Intel has been doing the former for many years. Most investors usually feel safer doing this; to change things like AMD did, they would need a new CEO who have the balls to dont care that much about investors.Reply -
thGe17 There will not be a Zen3+ on the desktop. The two V-Cache-Chips in 1HY22 will be based on regular Zen3.Reply
As known so far, AMD has canceled Zen3+ on the desktop. Rembrand will use Zen3+ for the APUs in 2022.
And for the rest, its all speculation and workload-specific. The top ALD CPU may still beat the 12-Core-V-Cache-Zen and the rest simply depends on the price (the same principles, which allowd AMD to survive in the last decade ;-)).
Zen4 again is a different story, because it has to compete with Raptor Lake and here again it remains to be seen what is relevant at the end for common workloads. Fans often tend to ignore that for consumers/gamers AMD has only managed to get "the performance crown" with Zen3, which is only a few month ago. Up to Zen2 Intel had no problem, even with their older Skylake-based architecture in 14nm, to compete against AMD and that is most likely the point at which fans eagerly throw the unbeatable 16-core-Ryzen on the table, but its relevance is small for the majority of the consumer market.
Just wait an see. We now have e healthy competition, Intel should be in the final steps of solving its manufacturing problem and we can expect even more powerful CPUs in the near future ... albeit this may not be of big importance for the over all market. ;-)
"The best scenario would be AMD and Intel both trading places in lead, alternatively."
The best for whom? For consumers, for AMD, for you? Who decides what is best? And why should the firm with the more limited resources take over? Such whishes make no sense.
And you seem to forget: The more AMD manages to establish itself, the more expensive their products will get. The only reason why e.g. Ryzen's have been "so cheap" in the past ist due to competion and gaining market share (as the underdog) and not because AMD is soo nice and friendly to its customers, almost altruistic. ;-) AMD wants exactly the same as Intel or Nvidia, there's no difference regarding to this point ... ok, to be exact, the difference is in the head of some customers, but this is more of a psychological topic and has almost nothing to do with technical details or market mechanics. -
watzupken With the new CEO, it seems like Intel is finally waking up at 100% after getting hit left right and center by their competitors. It also seems like Pat is throwing a lot of money to hire back ex employees whether from competitors or in retirement in addition to the push for TSMC 3nm which certainly will not be cheap. Let's see what they can come up with in the next 3 to 5 years.Reply -
Howardohyea using TSMC is a great call from Intel in my opinion. Sure the price will be extremely high but it will allow Intel to achieve competitive performance against AMD until their own fabs are up and running.Reply
From what I can see they rather loose profit than market share. -
watzupken
I doubt Intel would like to lose profit, market share and their face. But whether this direction makes sense, I think only time will tell. The problem is, while you can hire back the finest ex-employees or external hires, it does not always translate to a great product every time.Howardohyea said:using TSMC is a great call from Intel in my opinion. Sure the price will be extremely high but it will allow Intel to achieve competitive performance against AMD until their own fabs are up and running.
From what I can see they rather loose profit than market share.
My take on TSMC's 3nm is that they will probably reserve it for high margin products, so don't expect to see it on your everyday system. The problem is the cost is prohibitively high and usually only Apple have the deep pockets to tap on the most cutting edge node year after year. And they are selling these SOCs based on cutting edge nodes to retail, but of course they also charge a very steep premium for them. In addition, Intel probably have to compete with Apple to get some allocation for the node, and being a lesser priority customer (since they are also a competing fab), they likely have to pay the full fat price, as compared to Apple which is more like a strategic partner for TSMC. -
Howardohyea
have to agree with you right there, only time will tell what will happen to Intel's latest products.watzupken said:I doubt Intel would like to lose profit, market share and their face. But whether this direction makes sense, I think only time will tell. The problem is, while you can hire back the finest ex-employees or external hires, it does not always translate to a great product every time.
My take on TSMC's 3nm is that they will probably reserve it for high margin products, so don't expect to see it on your everyday system. The problem is the cost is prohibitively high and usually only Apple have the deep pockets to tap on the most cutting edge node year after year. And they are selling these SOCs based on cutting edge nodes to retail, but of course they also charge a very steep premium for them. In addition, Intel probably have to compete with Apple to get some allocation for the node, and being a lesser priority customer (since they are also a competing fab), they likely have to pay the full fat price, as compared to Apple which is more like a strategic partner for TSMC.