Intel Products CEO Michelle Johnston Holthaus attended a Bank of America Conference, where a strategic overview was presented. Intel’s longstanding financial woes have taken their toll on the once-leading manufacturer, leading to continual layoffs year after year for some time now. Former Intel CEO Pat Gelsinger was replaced by Lip-Bu Tan earlier this year, who has been tasked with getting the company back on track for profitability. Much of what Michelle Johnston has stated at the conference directly refers to Lip-Bu Tan’s plans for getting Intel back on track.
Jay Karya, an analyst from BofA and part of its semiconductor and semi cap equipment research team, was keen on getting details regarding exactly what Intel’s current strategy is for obtaining its goals. Michelle quickly began explaining how Intel is using its US-based foundries for its 18A node while leveraging TSMC’s foundries to shift volume when needed, while also having access to EUV manufacturing technologies. Both foundries will be used for the upcoming NovaLake product line, the flagship processor of which has been said to feature up to 52 cores made up of 4 low-power E-cores, 32 standard E-cores, and 16 P-cores and is rumored to combine Xe3 and Xe4 graphics onto the package.
Intel has also partnered with Samsung to use its foundries for production as well, and through this three-pronged approach has established not only a means for new nodes but also is able to have access to industry-available package IPs. Another advantage to these partnerships is that while there are added costs with them, Intel’s partners can shoulder some of manufacturing R&D costs thus offloading that burden from Intel. Manufacturing of different packages can then also be directed to the foundry who has the most optimal process for it as well.
Much of what Johnston covers indicates Intel is “laser-focused”, something mentioned multiple times on different topics, to its commitment in regaining its lead in the enterprise sector, but it is also changing its consumer strategy. Rather than launching a myriad of processors into the wild, hoping that customers will gravitate towards them per their needs, Intel is paying closer attention to market trends to identify potential product solutions. In essence, instead of throwing things at the wall to see what sticks, it is trying to identify where an actual demand exists for a product. Moving forward, the first goal is getting its gross margins back above 50%, but development teams for new products will now have to prove they can meet that same goal before moving into production.
It’s obvious that Lip-Bu has a longterm plan for Intel but also that things are in place for the short term. From ongoing dialogue with customers regarding AI integrations to its Granite Rapids successful launch it is already putting things in place for 2026 and beyond, and is looking to hire new talent.
“We have the right APs in CPU, we have the right APs in graphics, and interconnectivity, we know how to do rack scale, but we also need an infusion of talent, this is a huge focus for Lip Bu, of bringing in some new talent to help us here, and I think we also have to look at delivering products in these categories that fit in some seams in the marketplace, maybe non traditional places where we can differentiate ourselves against you know, the market leader Nvidia and so that will be a big focus for us as well. The one thing I can tell you, and I’m sure as you’ve had your meetings throughout the day, people want alternatives. People want choice.”
Michelle Johnston Holthaus – Intel Products CEO
Discussion (9 replies)
Join Discussion →They need to take a page out of nVidia's book....
[URL unfurl="true"]https://macrotrends.net/stocks/charts/NVDA/nvidia/gross-margin[/URL]
What's more interesting to see there is the net margin tab. Still over 50% profit NET. That's bonkers.
I don't know why I even try to have hope for them anymore.
[URL unfurl="true"]https://www.techpowerup.com/339638/intel-reportedly-struggles-with-panther-lake-manufacturing[/URL]
It's like watching someone shoot themself in the foot, only to drop the pistol and get shot in the knee, all the while trying to reach for an automatic to finish the job. Sigh.
That's a pretty grimdark assessment. I mean it isn't wrong per say.. but very grimdark. Really we need an outside company like a Tesla (no not tesla) to come in and just buy their local fab's to start producing chips. Sustain themselves on microcontrollers and device level SOC's like AMD did for so long.
Honestly I don't want Intel broken up... I think that would be a bad move and wish AMD never sold off or spun off global foundries. That was a mistake. But it was clear AMD didn't have the oomph to keep up with companies like TSMC for production and didn't want the lead weight of an aging tech production infrastructure weighing them down.
Ideally I want Intel to surprise everyone with some true next gen hardware on what they can make that exceeds expectations and blows people out of the water. Because brass tax... I don't care what nm rating it's made in as long as it does what I need it to do faster than or as fast as the next guy with greater reliability and lower TCO.
How many times has AMD been dead and buried b4 they came up with ryzen?
Seems ro me that a lot of people are keen to see intel fail as they used to be the big dog, that does not mean they will.
Intel is a strategic company, it will never fail, i dont know if AMD is considered strategic yet, probably not. That said Intel has done a lot of crap moves, specially stock buy backs, and forced ranking stupidity, which has rotted them pretty badly.
They came up with optane a great product I thought, instead of scaling it, and price it agressively, noooo, gotta have that huge margin, its plain stupid.
I wish they would bring Optane back.
What Intel needs to do is pull an IBM and just double down on the enterprise market. Make them their focus and pull the micro transactions OUT of the pipe line for the CPU market. (licensing features to enable them on your CPU.... no thank you!)
If they can reclaim the ground that AMD is and has made in the enterprise infrastructure market, and stay ahead of Nvidia/Microsoft/Amazon's custom high density low performance CPU's... then they have a chance.
Otherwise just shrink your market, dive in fully on secure processors, and cater to the TAA/DOD/High security minded market. Market your CPU's as the LEADER in compute security. OWN it. Nobody else is.... everyone else is being reactive to security and riding a razors edge of vulnerability to performance. BE the secure CPU to have in your servers. Sponsor active attacks against your hardware. PROVE it's the one to own.
Banks, DOD, TAA, Healthcare, and other high security minded companies will ABSOLUTELY purchase your proven hardened solutions over another 30 threads and faster throughput per socket.
They will even pay a PREMIUM for that added level of security. It will save them money on the back end insurance costs to protect against financial losses stemming from intrusions/data exfiltration.
That and every hypervisor/OS license in the world will soon be per thread after watching Broadcom do it so successfully. So people won't be making the MOST DENSE possible clusters.
Well, at least it's one, sort of, revenue stream. However, I think that what @Grimlakin has said makes a whole lot more sense.
[URL unfurl="true"]https://www.techpowerup.com/339728/intel-wins-tesla-dojo-3-packaging-contract-in-dual-supplier-strategy[/URL]