Intel Anticipates Further Market Share Loses To AMD Through 2023, Aims To Re-Establish Growth By 2025

Intel's CEO Rumored To Visit TSMC Next Month As 14th Gen Meteor Lake Allegedly Pushed Back to End of 2023

Yesterday at the Evercore ISI TMT conference, Intel's CEO, Pat Gelsinger, stated that Chipzilla is anticipating further market share losses to AMD through 2023.

Intel Losing Market Share To AMD Through 2023, Rebound Expected In 2025

During the talk, Intel's CEO talked about how the dramatic changes in the PC market have not only affected their business but also the businesses of other companies in the tech market. He goes on to say that their own execution of products resulted in adverse earnings during the last quarter but the majority of the change in outlook was driven by market conditions.

Related StoryHassan Mujtaba
Intel Arc A770 & Arc A750 Limited Edition Graphics Card Benchmarked, Show Marked Improvement In OpenCL & Vulkan API

Since then, I think things have materialized pretty much as we expected, but even a little bit worse. And so we gave a range for our outlook on the call, something Intel never does. We always give you a number. This time, we gave a range, right, given the overall economic uncertainty at the time. And I'll say we're within the range for the quarter and for the ear, but trending toward the lower end, right?

And things have generally deteriorated a bit more than we would have forecast earlier in the year, but still within the range or the quarter and the year. But it's pretty rough out there. And a lot of OEMs shifting views and outlook, channel inventory adjustments, and still a lot of economics to work through.

at said, and as we said on the earnings call, we're now shipping below the consumption rate in the end market, right? o we are seeing those inventory bums. It does give us, I'll say, a view of growth in Q4. We said Q2 and Q3 was the bottom. That's still our view, growth in Q4 as some of those inventory burns start to reverse. And generally, Q4 is a better quarter. So all of that taken together, we'd say we're in the range, but a little bit more tepid than we would have indicated t the beginning — or at the Q2 call.

And essentially, every other announcement has been confirmatory of what we said so far in the industry. The few exceptions have a very unique proprietary position on short nodes and technologies and everything else is just a firm, that perspective.

Intel CEO, Pat Gelsinger at Evercore ISI TMT

Talking about how the company sees AMD moving forward, Pat says that the competition has done a good job and while Intel is still on a process node/technology deficit (10nm vs 5nm), the company is getting there with their future products with Emerald Rapids in 2023 and Granite Rapids / Sierra Forest in 2024. The CEO goes on to mention the advantages of Sapphire Rapids & how it is so good at AI. Sapphire Rapids is said to offer better performance and power than the AMD alternative but not in a significant way and Pat's goal is to bring out products fast that are not just better but more than better compared to the competition.

However, rolling out products takes time and Intel knows that they will be losing market share to AMD for a good while and throughout 2023 and even 2024. This is because the competition just has much more momentum in almost all segments. As such, Pat believes that Intel will start to regain its market share by 2025-2026. Till then, they'll bottom out and start trying to keep up with the TAM.

And if we step back from it, obviously, Ice Lake now with Sierra Forest, next year with Emerald Rapids, in '24 with Sierra Forest and Granite Rapids, the road map gets stronger, right? And per the earlier question on execution, our disciplines, quality, and volume deliveries are getting better. That said, our competition has done a good job, right? And we haven't for a number of years, and we're still on a process technology deficit. Sierra Sapphire Rapids, looking good. We said we'll go to production this year. That's looking good now. And customers are very excited about it.

Sapphire Rapids is the best product in areas like Al performance and security capabilities, highly differentiated from anything else in the marketplace. It's better than the AMD alternative in terms of power performance but not dramatically so. So we win the benchmarks, but it's not like it's a compelling crushing leadership position. As I tell our teams if the product is close to competition, it's not a good product, right? An Intel product will be dramatically better than the alternatives as we go forward.

And that's what we got to get back to, that if you're not running on Intel, you're just not making a good decision, right? And we just have to rebuild that confidence in the industry. We do expect that overall, our data center business grows every ear as we go forward from where we are. As we said, 02, 03, the bottom. But we believe that we're still losing share at least through next year, right?

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, 26 when we start regaining share—material share gains.

Intel CEO, Pat Gelsinger at Evercore ISI TMT

One interesting thing that Pat has mentioned is that Intel is fundamentally rebuilding its execution machine by putting all of its projects into a consistent development methodology under a new PLC model known as Palladius. This will play a key role in the next generation of products but Pat says that Sapphire Rapids' design was started pre-Palladius, almost 5 years ago. The products that are already in the design phase will continue to get better compared to their A0 tapeouts, products like Arrow Lake & Lunar Lake for clients and Emerald Rapids & Granite Rapids for servers but it is the 2025 & 2026 product lineup that will truly take benefit from Intel's new design methodologies.

Now the problem is products like Sapphire Rapids were started 5 years ago. So we're sort of injecting the new methodology into projects in flight. Right? And while we're making them better, they're not entirely born in the new methodology. So I'd just say execution gets better each year and things like getting full software emulation done before AO tapeouts, shift left of all of these types of things, just good, modern development practices. Hey, they're being incrementally put into every product on the road map.

So Lunar Lake and Arrow Lake and the client side are getting better. The server, Emerald Rapids, Granite Rapids, CRRs are getting better. But it's not until '25 until they were born entirely in the new methodology. But I have confidence that every aspect of the development program is getting better, right, as we go forward because we're inducing more of those engineering disciplines.

And hey, this is what I did for a living when I was an early chip designer, so I have a lot of passion in this area. The other thing that we're reinstituting is what we've called the TikTok methodology, which is the alignment of design with process technology. And with that, it's rebuilding that alignment because fundamentally, if we have the best transistors, Intel will be fine, right? Because even a mediocre design with the best transistors will still be better than the alternatives, right?

If you have a great design with the best transistors, now you have a killer product, right? So here, that 5 nodes in 4 years, rebuilding the process technology machine is going well. And it is essentially a rising tide for product development machines. Now obviously, '24, we think we're competitive. '25, we think we're back to unquestioned leadership with our transistors and process technology. But then we have to rebuild the cadence of products with process technology that is a risk management methodology so that you never miss, right?

You always have redundancy. You always have a new process with a proven design or a new design with a proven process. It's a risk management methodology. And so we're getting that tick-tock methodology back in place as well. So I feel every year is going to get better as we go through the saddle, right, if you would, right? We're still not done with the addle. We have another challenging year in front of us yet.

But as we get into '24 and '25 in execution, process technology„ leadership, all these pieces start to really come to fruition. And on the client space, which are simpler designs, we've already seen that. We delivered Alder Lake on schedule. Raptor Lake is on schedule and Lake is looking good for next year. So we're starting to see more of those execution disciplines emerging more rapidly on the simpler designs on servers. It's a little bit harder, more complex design. So it will be a little bit longer to come to fruition there.

Intel CEO, Pat Gelsinger at Evercore ISI TMT

Just last month, AMD achieved its biggest market share milestone with EPYC surpassing the historical share of Opteron and crunching away more share from Intel in the server segment. Surely, AMD has come a long way and with solid execution and product launches year after year, they have continued to see a major gain in the PC market.

Sandra Rivera, executive vice president and general manager of the Datacenter and AI Group at Intel Corporation, displays a wafer holding 4th Gen Intel Xeon Scalable processors (code-named Sapphire Rapids) before the opening of Intel Vision 2022 on May 10 in Dallas. During the hybrid event, Intel’s leaders will announce advancements across silicon, software and services, showcasing how Intel brings together technologies and the ecosystem to unlock business value for customers today and in the future. (Credit: Walden Kirsch/Intel Corporation)

Intel could definitely learn from this & it looks like they are by acknowledging the wins and leadership position of their rival. We definitely want to see more competition in the CPU segment and Intel has a long way to go to catch up to the red team but hopefully, if they execute their product lineups properly, they will come close to the goal of being the market leader once again.

Intel Xeon SP Families (Preliminary):

Family BrandingSkylake-SPCascade Lake-SP/APCooper Lake-SPIce Lake-SPSapphire RapidsEmerald RapidsGranite RapidsDiamond Rapids
Process Node14nm+14nm++14nm++10nm+Intel 7Intel 7Intel 3Intel 3?
Platform NameIntel PurleyIntel PurleyIntel Cedar IslandIntel WhitleyIntel Eagle StreamIntel Eagle StreamIntel Mountain Stream
Intel Birch Stream
Intel Mountain Stream
Intel Birch Stream
Core ArchitectureSkylakeCascade LakeCascade LakeSunny CoveGolden CoveRaptor CoveRedwood Cove?Lion Cove?
IPC Improvement (Vs Prev Gen)10%0%0%20%19%8%?35%?39%?
MCP (Multi-Chip Package) SKUsNoYesNoNoYesYesTBD (Possibly Yes)TBD (Possibly Yes)
SocketLGA 3647LGA 3647LGA 4189LGA 4189LGA 4677LGA 4677TBDTBD
Max Core CountUp To 28Up To 28Up To 28Up To 40Up To 56Up To 64?Up To 120?Up To 144?
Max Thread CountUp To 56Up To 56Up To 56Up To 80Up To 112Up To 128?Up To 240?Up To 288?
Max L3 Cache38.5 MB L338.5 MB L338.5 MB L360 MB L3105 MB L3120 MB L3?240 MB L3?288 MB L3?
Vector EnginesAVX-512/FMA2AVX-512/FMA2AVX-512/FMA2AVX-512/FMA2AVX-512/FMA2AVX-512/FMA2AVX-1024/FMA3?AVX-1024/FMA3?
Memory SupportDDR4-2666 6-ChannelDDR4-2933 6-ChannelUp To 6-Channel DDR4-3200Up To 8-Channel DDR4-3200Up To 8-Channel DDR5-4800Up To 8-Channel DDR5-5600?Up To 12-Channel DDR5-6400?Up To 12-Channel DDR6-7200?
PCIe Gen SupportPCIe 3.0 (48 Lanes)PCIe 3.0 (48 Lanes)PCIe 3.0 (48 Lanes)PCIe 4.0 (64 Lanes)PCIe 5.0 (80 lanes)PCIe 5.0 (80 Lanes)PCIe 6.0 (128 Lanes)?PCIe 6.0 (128 Lanes)?
TDP Range (PL1)140W-205W165W-205W150W-250W105-270WUp To 350WUp To 375W?Up To 400W?Up To 425W?
3D Xpoint Optane DIMMN/AApache PassBarlow PassBarlow PassCrow PassCrow Pass?Donahue Pass?Donahue Pass?
CompetitionAMD EPYC Naples 14nmAMD EPYC Rome 7nmAMD EPYC Rome 7nmAMD EPYC Milan 7nm+AMD EPYC Genoa ~5nmAMD EPYC BergamoAMD EPYC TurinAMD EPYC Venice
Launch201720182020202120222023?2024?2025?

News Sources: Eric Jhonsa , Retired Engineer , Tomshardware

WccfTech Tv
Subscribe
Filter videos by
Order