Intel Expects More Market Share Loss Throughout 2023, Will Likely Exit More Businesses
Intel CEO Pat Gelsinger expects more pain on the road to recovery.
Intel CEO Pat Gelsinger spoke at the Evercore ISI TMT conference yesterday, saying that the company expects to continue losing data center market share throughout at least 2023 and will only begin regaining in 2025 and 2026. Gelsinger also said that the company would likely exit other businesses, much like it recently began exiting Optane memory, as it continues to narrow its focus to its core competencies.
Intel's recently delayed its Sapphire Rapids launch again, with the chips now slated to arrive in 2023. Gelsinger remarked that while the new chips are 'better than the AMD alternatives" in power and performance and will win in some benchmarks, the advantages aren't dramatic enough to slow AMD's advance. As a result, Intel's data center business won't grow at the same rate as the market, meaning the company will continue to lose market share.
"We do expect that overall our data center business grows every year as we go forward. From where we are, as we said, Q2, Q3 [is] the bottom. But we believe that we're still losing share at least through next year," Gelsinger said.
"Competition just has too much momentum, and we haven't executed well enough. So we expect that bottoming. The business will be growing, but we do expect that there continues to be some share losses. We're not keeping up with the overall TAM growth until we get later into '25 and '26 when we start regaining share, material share gains," Gelsinger added. Notably, the statement isn't definitive about the company's performance in 2024 — Gelsinger specifically stated that the company wouldn't begin regaining market share until 2025.
"Now, obviously, in 2024, we think we're competitive. 2025, we think we're back to unquestioned leadership with our transistors and process technology," Gelsinger said.
AMD has already taken data center market share from Intel for 13 consecutive quarters, reaching 20.2% of the market, and Gelsinger's comments point to at least five more quarters of share losses — and perhaps more.
Gelsinger pointed to the company's Sierra Forest processors as a key innovation that will help the company address the other chip architecture steadily siphoning off market share — Arm. The Sierra Forest Xeon processors have efficiency cores optimized to provide the utmost power efficiency and performance density, so they'll have higher core counts.
"Well, when we deliver the Forest product line, we deliver power performance leadership versus all Arm alternatives, as well. So now you go to a cloud service provider, and you say, 'Well, why would I go through that butt ugly, heavy software lift to an ARM architecture versus continuing on the x86 family?'" Gelsinger said.
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.
AMD vs. Intel Roadmap | 2022 | 2023 | 2024 |
Intel | - | Sapphire Rapids / Emerald Rapids - Intel 7 | Granite Rapids / Sierra Forest - Intel 3 |
AMD | Milan-X - 7nm | Genoa - 5nm - 96 Cores | Bergamo - 5nm - 128 Cores | Turin - 4nm/3nm |
Intel's Sierra Forest processors, which the company designed at the behest of its largest customers, look promising. However, Sierra Forest isn't scheduled to arrive until 2024.
Meanwhile, AMD's 5nm Bergamo chips, which employ 128 simplified 'Zen 4C' cores in a similar density-improving arrangement to address the same market segments, arrive a year earlier in 2023.
The Long Road Ahead
While it's clear that Intel has several tough years ahead as it works to rebuild, Gelsinger did point to leadership changes that will help accelerate the turnaround.
"Seventy percent of the leaders, or the leaders minus one, are new to the company or new to the role that they're in. So it's been a pretty dramatic rebuilding of the leadership team." Gelsinger noted. He also announced that the company had promoted Shlomit Weiss to senior vice president and Co-GM of the Design Engineering Group.
Gelsinger also plans to continue focusing on the company's core competency: Logic chips. That means he will stay open to exiting more businesses, much like Intel recently decided to wind down its Optane business.
"Obviously, Optane. And man, I sort of joke that Intel exited the memory business 40 years ago, and they've just kept making that decision. Right? Well, I'm gonna close that frickin' door, and we're gonna stay out of the memory business and really get a cleanliness of our business strategy around logic," Gelsinger said. "You know, we have a few more that we'll likely exit as we continue to prune and get more focused."
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.
-
Neilbob I'm finding it hard to believe a single thing this guy says lately. A few things I raised an eyebrow at:Reply
'Gelsinger remarked that while the new chips are 'better than the AMD alternatives" in power and performance and will win in some benchmarks'
In which metrics? AMD seem to be demonstrably more power efficient in large multi-threaded tasks. Even when comparing their older architectures.
'Well, when we deliver the Forest product line, we deliver power performance leadership versus all Arm alternatives, as well. So now you go to a cloud service provider, and you say, 'Well, why would I go through that butt ugly, heavy software lift to an ARM architecture versus continuing on the x86 family?'" Gelsinger said'
There have been similar claims made before. I'll believe it when I see it?
'Seventy percent of the leaders, or the leaders minus one, are new to the company or new to the role that they're in. So it's been a pretty dramatic rebuilding of the leadership team'
I'm not certain how having even more Managers and Vice Presidents will really help matters.
'Gelsinger's comments point to at least five more quarters of share losses — and perhaps more'
At least and perhaps more. Don't these two things basically mean the same thing? :p
<Sigh> Hopefully my own cynicism is blinding me to anything that could be positive in terms of competition. -
-Fran- Looks like AMD's not the only thing in the rear-view mirror, uh? Business Units are also gonna be left on the road. Sadge.Reply
Then again, losing the "fat" could be helpful to re-focus on what Intel does best.
And as for the market share... Hm... Until AMD can figure out ML accelerators and add them into Zen, I think Intel will still have an edge. This being said, Zen5 is posed to include ML accelerators, if not Zen6. So I don't think they'll stop the bleeding. Maybe slow it down a bit more until they reach balance with AMD.
Regards. -
InvalidError
You don't want to cut too much 'fat' since you need it to immunize yourself against significant dips in any one specific market segment. Like right now, the client compute market is getting hit hard by the post-COVID sales lull while graphics is getting assaulted by supply chain over-stock thanks to the trifecta of imminent next-gen launches, the likely end of Ethereum PoS and waves of heavily discounted retired mining cards hitting the used market.-Fran- said:Then again, losing the "fat" could be helpful to re-focus on what Intel does best. -
Jimbojan
Leadership changes the business direction, like Gelsinger himself, will make a different. Actually, Intel's Meteor Lake will have better power efficiency than AMD's product; Raptor Lake is now already slightly better than AMD power efficiency with better performance.Neilbob said:I'm finding it hard to believe a single thing this guy says lately. A few things I raised an eyebrow at:
'Gelsinger remarked that while the new chips are 'better than the AMD alternatives" in power and performance and will win in some benchmarks'
In which metrics? AMD seem to be demonstrably more power efficient in large multi-threaded tasks. Even when comparing their older architectures.
'Well, when we deliver the Forest product line, we deliver power performance leadership versus all Arm alternatives, as well. So now you go to a cloud service provider, and you say, 'Well, why would I go through that butt ugly, heavy software lift to an ARM architecture versus continuing on the x86 family?'" Gelsinger said'
There have been similar claims made before. I'll believe it when I see it?
'Seventy percent of the leaders, or the leaders minus one, are new to the company or new to the role that they're in. So it's been a pretty dramatic rebuilding of the leadership team'
I'm not certain how having even more Managers and Vice Presidents will really help matters.
'Gelsinger's comments point to at least five more quarters of share losses — and perhaps more'
At least and perhaps more. Don't these two things basically mean the same thing? :p
<Sigh> Hopefully my own cynicism is blinding me to anything that could be positive in terms of competition.
Leadership changes the business direction, like Gelsinger himself, will make a different. Actually, Intel's Meteor Lake will have better power efficiency than AMD's product; Raptor Lake is now already slightly better than AMD power efficiency with better performance.Neilbob said:I'm finding it hard to believe a single thing this guy says lately. A few things I raised an eyebrow at:
'Gelsinger remarked that while the new chips are 'better than the AMD alternatives" in power and performance and will win in some benchmarks'
In which metrics? AMD seem to be demonstrably more power efficient in large multi-threaded tasks. Even when comparing their older architectures.
'Well, when we deliver the Forest product line, we deliver power performance leadership versus all Arm alternatives, as well. So now you go to a cloud service provider, and you say, 'Well, why would I go through that butt ugly, heavy software lift to an ARM architecture versus continuing on the x86 family?'" Gelsinger said'
There have been similar claims made before. I'll believe it when I see it?
'Seventy percent of the leaders, or the leaders minus one, are new to the company or new to the role that they're in. So it's been a pretty dramatic rebuilding of the leadership team'
I'm not certain how having even more Managers and Vice Presidents will really help matters.
'Gelsinger's comments point to at least five more quarters of share losses — and perhaps more'
At least and perhaps more. Don't these two things basically mean the same thing? :p
<Sigh> Hopefully my own cynicism is blinding me to anything that could be positive in terms of competition. -
Neilbob Jimbojan said:Leadership changes the business direction, like Gelsinger himself, will make a different. Actually, Intel's Meteor Lake will have better power efficiency than AMD's product; Raptor Lake is now already slightly better than AMD power efficiency with better performance.
How do you know this, when neither Raptor nor Zen 4, let alone Meteor Lake, are released yet?
Also, we're talking servers and not consumers here. -
JayNor Intel taped in Sierra Forest, Intel-3 process, prior to q1 earnings call. They expected to have it back in the lab sometime this quarter.Reply
Meteor Lake, Intel-4 process, is "now broadly sampled to customers", according to the q2 earnings call. -
rluker5 AMD needs to find a solution to SQUIP.Reply
I don't know how many companies are allowed to process their customers data on insecure processors. Google has an unofficial fix that locks cores to processes which will cost some efficiency and performance depending on the use, and the processors can be made secure by turning off SMT and becoming very close to e-cores. But really I would think the rules corporate has to follow matter more. You can promise security to your customers or not. You need a secure processor (has fixes for the flaws they have) to do that. And most businesses don't have the resources to have an IT team good enough to fix this themselves other than turning off SMT.
I don't want my bank account info on a publicly accessible or shared server that is proven insecure. All AMD has to do is come up with a fix that works. Maybe test and certify Google's approach. Intel had their own fix for a while and I think the OS's went with Google's fix eventually so we might almost be there already. I know my info is almost certainly on some AMD servers, they should do their job and fix their problem before the owners of their server chips get sued for the consequences of knowingly putting their customers at risk. -
Wisecracker Times are tough for Chipzillah __ abandoning 10+++++ for 7nm . . .Reply
in 2023 o_O (after 8 years?) Can you say, "Tick--Tock--Splat ??"
I suspect Rory Read did a number on Intel by swooping-in to capture the "Freedom Fabric" from SeaMicro. The process difficulties were tough enough, but EMIB (and Foveros) was a shot night in the ruts to Chipzillah's future server compute density and bandwidth . . . -
shady28 rluker5 said:AMD needs to find a solution to SQUIP.
I don't know how many companies are allowed to process their customers data on insecure processors. Google has an unofficial fix that locks cores to processes which will cost some efficiency and performance depending on the use, and the processors can be made secure by turning off SMT and becoming very close to e-cores. But really I would think the rules corporate has to follow matter more. You can promise security to your customers or not. You need a secure processor (has fixes for the flaws they have) to do that. And most businesses don't have the resources to have an IT team good enough to fix this themselves other than turning off SMT.
I don't want my bank account info on a publicly accessible or shared server that is proven insecure. All AMD has to do is come up with a fix that works. Maybe test and certify Google's approach. Intel had their own fix for a while and I think the OS's went with Google's fix eventually so we might almost be there already. I know my info is almost certainly on some AMD servers, they should do their job and fix their problem before the owners of their server chips get sued for the consequences of knowingly putting their customers at risk.
This has been in the back of my mind since I saw that, and saw AMDs weak response. Testers were able to steal a 4096 bit encryption key from one VM using another VM on the same hardware (running on same CPU).
This makes any store or corporation using a cloud provider who uses AMD is a ticking time bomb for massive liability until a fix is implemented. Many companies use insurance for this as well, and insurance companies could jack up rates or refuse coverage if a fix is not implemented.
The first storefront that gets hacked this way & sued into oblivion will be the sign. With disabling SMT being the only known fix right now, AMD could quite literally be crippled by this overnight. -
waltc3 Intel is once again reduced to talking about the future--because even though Intel has products Intel says--or, rather Pat says, "are better than the AMD alternatives" they unfortunately aren't good enough to bring to market against those AMD alternatives. Too bad, Pat. (Sure thing there, Pat--keeping up with that Alzheimer's medicine? Maybe you should hit up Biden for some of his antipsychotics...;)) Intel marketing is fairly sickening these days--Gelsinger has no clue, and he's just lost, frankly.Reply
The biggest problem Intel has at present is the fact that AMD is a perpetually moving target. They're not sitting still for a few years waiting on Intel to catch up and move ahead. Which is very off-putting for Intel. Because just when Intel thinks it has an AMD beater--oooops, AMD is releasing newer CPUs that move it ahead again. Rinse and repeat. One of the things I found very interesting about the August Zen4 reveal was how much Papermaster and Su, both, if memory serves, emphasized the fact that AMD would continue in this huge R&D vein post Zen5. One or both of them said it more than once, which I thought was interesting, as it was letting competitors and investors know that AMD has plenty more up its sleeves. They've said it before, of course, but this time it seemed like they really wanted to stress it more than they have.