Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/new-intel-ceo.22268/page-4
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021770
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

New Intel CEO

Cadence was obviously a point on the contract. Now you have Cadence as the only entry point for 2 tier-1 foundries
Expect Cadence licenses for software for latest node work to jump up in price soon.

Foundries are EDA neutral, they have to be. Cadence did not get favorable treatment when Lip-Bu joined the Intel board and that will not change now that he is CEO. Synopsys is favored by Intel design and Synopsys owns the foundry IP business so it would be bad to alienate them. Same with Siemens EDA. Siemens is an $80B+ force of nature. Alienating them would be a serious mistake. I work with all of these companies and know them well. The playing field here will be as level as possible.
 

What's Lip Bu Tan, Intel's next CEO Likely to do?​

Jeff Morrison

Jeff Morrison​


Financial Cultural Operational and Technical Consultant - Alpha Sense Financial Consulting


March 12, 2025
Intel's decision to appoint Lip-Bu Tan as CEO signals a strategic pivot toward strengthening its position in semiconductor design and foundry services. Tan’s leadership at Cadence Design Systems demonstrated his ability to drive innovation, expand product offerings, and build deep relationships with semiconductor clients. His appointment suggests Intel will double down on its Intel Foundry Services (IFS) strategy, aiming to establish itself as a formidable competitor to TSMC and Samsung in advanced node manufacturing. Given the challenges Intel faces in regaining process leadership, Tan's experience with EDA tools and ecosystem partnerships could be critical in refining Intel’s foundry roadmap.

For Intel Foundry Services, Tan will likely focus on increasing external customer adoption, addressing concerns about Intel’s historical execution gaps. His industry connections, especially with AI and hyperscaler firms, could help secure more long-term foundry contracts, reducing reliance on internal chip demand. Expect to see renewed investment in process technology innovation, particularly in sub-3nm nodes, where Intel must prove its High-NA EUV capabilities against established rivals. Additionally, Tan may look at improving IFS’s business model to offer more flexible, customer-centric solutions, mirroring TSMC’s successful approach.

High-Performance Computing (HPC) is another key battleground where Intel must strengthen its position. While it has made strides with Xeon, Gaudi AI accelerators, and Ponte Vecchio GPUs, competition from NVIDIA and AMD remains intense. Tan's expertise in silicon design and AI-driven workloads positions him well to guide Intel’s roadmap toward more specialized compute architectures. Expect a stronger push into AI-optimized chips, leveraging Intel’s strengths in packaging technologies like Foveros and EMIB.

One immediate challenge for Intel in HPC is aligning its product roadmap with evolving AI and supercomputing demands. Given Tan’s background in EDA and silicon IP investment, Intel may place greater emphasis on co-designing AI hardware with major cloud providers and research institutions. This could involve forming strategic alliances with key players like AWS, Google, and Meta to develop more custom silicon for data center and AI inference applications.
Another likely shift under Tan’s leadership will be Intel’s approach to software-hardware integration. Given his experience driving Cadence’s AI-driven design tools, Intel may accelerate investments in open software ecosystems to better support its chips. This could involve optimizing software stacks for AI workloads, improving developer tools for oneAPI, and enhancing support for industry-standard frameworks to drive adoption across AI and HPC markets.
Financially, Tan will need to balance Intel’s heavy capital expenditure with a return to profitabil5ity in IFS and its core processor business. Cost discipline, strategic asset partnerships, and possibly even divestitures of non-core assets could be on the table to streamline operations. Intel’s foundry business will likely require aggressive pricing strategies and government incentives to compete effectively against Taiwan and Korea’s incumbents.
Ultimately, Tan’s leadership will be measured by Intel’s ability to execute on its foundry and HPC ambitions. If he can accelerate process innovation, deepen customer relationships, and align Intel’s product roadmap with industry trends, the company has a real chance at regaining its competitive edge. However, execution will be key, as past leadership transitions have struggled with delivery timelines.

"Intel's decision to appoint Lip-Bu Tan as CEO signals a strategic pivot toward strengthening its position in semiconductor design and foundry services. "

No kidding 🤔

What else does Intel do?
 
Good article find.

This makes me increasingly skeptical that TSMC is going to see any competition after 14A :(. Samsung seems broken on node development at this point.

A spun-off Intel Foundry is going to require a lot of customers to justify further node development, and that window seems to be closing rapidly. After Pat, (completely uneducated guess), I think the board is going to set a deadline of "get as many customers as you can by [end of?] 2026, and then start shopping out the Foundry for sale". (I think the board will give him 2025 to get engineering process and company size in order, 2026 to prep hard for the pivot, then 2027 it's sale time).

The only scenario that works I think in continued competition is if TSMC stumbles hard in the next 2-3 years.
I think Intel Products will always favor Intel Foundry (no matter what they say in public). Even if spun off (say 2027+), Intel parent company would still own a majority stake in it (like Mobileye now or soon to be Altera spin off). Intel still ships about 35 to 40 Million units of PCs every quarter. On top of that Data Center, Network & Edge products. That is not going away anytime soon.

So Intel Products is a whale customer for Intel Foundry if 18A & 14A ends up good & competitive nodes. They just need some more customers to sustain node development further (they need them fast now to prove themselves). Intel Foundry just needs to become lean and efficient to make money. And we know Intel products team in the past took the fabs for granted (too many steppings & hot lots etc. screwing up fab economics).

imho, I think LBT is dealt way better hand than PG with respect to timing of leading Intel. I mean, Pat fixed mostly the process issue with 5N4Y. Most of the high investment that needed to be spent to get back to process leadership/parity & to build the shells are done. LBT is taking over right about when things are turning around (reduced capex going forward, enough govt grants <if not cancelled> & partner contribution to offset most of the capex portion) and only thing remains now is execution. Pat is not without blame for some things like being over optimistic about PC market after COVID, missing the AI boom, not managing expectations for External Foundry & could have been little cost conscious. But it is unfortunate Pat is not at the helm when things turn.

Even before yesterday, Wall st consensus was Intel was expected to be GAAP net income positive & adjusted free cash flow positive in 2025 mainly due to cost cuts and increased foundry utilization by Panther Lake & Granite rapids ramp. 2026 was forecasted to be even better (only ? was CHIPS act assumptions in the forecasts - this has not changed even now).

We can argue about these estimates but it is not rocket science really, Intel' CFO guided for $17.5B in OpEx, $20B in Capex and $10B in Govt grants & Partner contributions (SCIP) in 2025. Plug these numbers into adjusted 2024 financial numbers (adjusted to remove 1 time charges in Q3'24 results) and you will see it closely match with Wall St Analyst estimates for 2025. So it is a matter of hitting those guidance.

With that said, I hope LBT brings additional efficiency improvements to Intel operations as a whole. His resume is very illustrious, he basically capitalized on all the successful waves in semi at the right time as a VC. I hope he continues his success at Intel too.
 
Last edited:
"Intel's decision to appoint Lip-Bu Tan as CEO signals a strategic pivot toward strengthening its position in semiconductor design and foundry services. "
No kidding 🤔
What else does Intel do?

He also missed the trust and foundries should not compete with customers part. Lip-Bu is a Morris Chang disciple and customers must come first. This is not possible for an IDM foundry like Intel or Samsung. Jeff spent 23 years at Intel manufacturing ending in 2006 and I would bet has never even met Lip-Bu.
 
LBT disagreed with what Pat was doing and planning to do. He quit the board. Now he has committed to be CEO. I think we can summarize that they will not be following Pat's plan and may do the opposite (I hope so). And like Pat, I am sure he got the backing of the board to do what he wants.

Good Outstanding hire.... Can't wait to see the changes. I would expect the roadmap changes to be announced by June. He has to do something with that Foundry spending quick.
 
Foundries are EDA neutral, they have to be. Cadence did not get favorable treatment when Lip-Bu joined the Intel board and that will not change now that he is CEO. Synopsys is favored by Intel design and Synopsys owns the foundry IP business so it would be bad to alienate them. Same with Siemens EDA. Siemens is an $80B+ force of nature. Alienating them would be a serious mistake. I work with all of these companies and know them well. The playing field here will be as level as possible.

For any serious design tapeout on TSMC latest nodes, full Cadence shop is a prerequisite for non billion dollar companies.

Only the largest fabless have enough labour to fully design, and verify 100% own silicon. Even Intel buys 3rd party IP for less important stuff these days, and I heard they have finally started hiring external contractors too. 9 out of 10 South Asian VLSI sweatshops are Cadence shops.

Intel has tried to wean itself from Synopsys dependency, but went back to it. Just too much internal tools were written to work with Synopsys software, I was told.
 
This is certainly going to be very interesting. Whether Intel can absorb the cultural and behavioural changes Lip-Bu is talking about (and has always followed) can't be certain. It's a whole new type of leadership for Intel and Intel will likely feel like a quite different company if he can pull this off. Certainly impressed that Intel persuaded him to sign on. And admire Lip-Bu's guts for signing up for the toughest job in the industry right now.
 
This is certainly going to be very interesting. Whether Intel can absorb the cultural and behavioural changes Lip-Bu is talking about (and has always followed) can't be certain. It's a whole new type of leadership for Intel and Intel will likely feel like a quite different company if he can pull this off. Certainly impressed that Intel persuaded him to sign on. And admire Lip-Bu's guts for signing up for the toughest job in the industry right now.

I don't think so. If I were Chen:
1. Make normal PDKs
2. Bribe EDA vendors, and biggest 3rd party IP vendors to support your process flow as a priority
3. Make discount deals for EDA users. Give away IP blocks for use with your foundry service, especially every tricky analog circuit, and everything optimised specifically for your process.
4. Just play to Intel's product strength: brand, good support, good software, sane sales (AMD sales people are plainly terrible)
5. Crank Xeon prices again
6. Once you get enough load on foundry from client, start a bit of an expansion
 

Intel’s New CEO Called ‘Strong Choice’ to Respin Company​

By Alan Patterson 03.13.2025

Intel’s new CEO, Lip-Bu Tan, is a good choice to turn the struggling U.S. chipmaker around, according to analysts who spoke to EE Times.

Tan, an industry veteran, is rejoining the Intel board after departing in August 2024, the company said in an announcement yesterday. He succeeds interim Co-CEOs David Zinsner and Michelle Holthaus. Zinsner will keep his position as executive VP and CFO, while Holthaus will remain CEO of Intel Products.

Tan faces numerous challenges. Three months ago, Pat Gelsinger departed as Intel CEO. For years, the top U.S. chipmaker has lost market share in its core CPU business to AMD while failing to enter new businesses like smartphone and AI chips. At the same time, Intel’s effort to divest its unprofitable chipmaking unit, Intel Foundry, as a standalone business is likely to take years. Intel Foundry is still trying to catch up in process tech with foundry leader TSMC.

“I see significant opportunities to remake our business in ways that serve our customers better and create value for our shareholders,” Tan said in a prepared statement. “Intel has a powerful and differentiated computing platform, a vast customer installed base and a robust manufacturing footprint that is getting stronger by the day as we rebuild our process technology roadmap.”

Intel is rolling out its 18A process tech this year, which is likely to bring the company on a par with TSMC’s 2-nm process. Still, the company needs to switch from its focus on making Intel-branded chips only to making chips for a wide range of customers.

Tan is a “strong choice” to respin Intel Foundry into a customer-first organization, TechAnalysis analyst Jeff Koch told EE Times.

“Given that he left the board because of frustration with the slow-moving bureaucracy, he likely will give high priority to speeding up and slimming down the company,” Koch said. “We expect significant restructuring of the slow and ineffective middle management at Intel. Although his recent time was spent in management, Lip-Bu has technical bona fides and should approach the role with a technical eye. This contrasts with some of his predecessors who focused on financial engineering, to Intel’s long-term detriment.”

Handel Jones, CEO of International Business Strategies, who has known Tan for years, called the former Cadence CEO a “good choice” for Intel.

One of the priorities for the company is to become more competitive in AI accelerators for data centers, a business dominated by Nvidia, Jones said.

“Intel has the architectural expertise to accomplish this, but it will take 24 to 36 months,” Jones said. “It is likely that Intel will need to participate in the Arm processor ecosystem.”

Intel Foundry

Tan will need to accelerate efforts to spin off Intel Foundry as an independent entity or save the manufacturing unit from a predicament where Intel Products is the only major customer, according to analysts.

“The return of Tan as CEO suggests that spinning off or selling Intel Foundry will accelerate,” independent analyst Nicolas Baratte said in a report provided to EE Times. “It probably also means scaling down considerably Intel manufacturing plans and outsourcing more to TSMC. If you can’t beat them, join them.”

In February, Intel said that it plans to continue outsourcing some production to TSMC to help launch products more quickly. Intel outsources about a third of its production to TSMC, according to Baratte.

Others said Intel Foundry is essential to reviving the U.S. semiconductor industry.

“Intel Foundry’s survival is key to America’s national and economic security as the only domestic company with advanced logic R&D and manufacturing capability,” SemiAnalysis said.

Shareholders and customers would benefit if Intel were to split in two, but doing so will take $30 billion in funding support for Intel Foundry to become financially viable, Jones said.

“That money can be found because it is strategically important in the U.S. for Intel to survive and have advanced manufacturing capacity,” Jones said.

Opposite directions

TSMC and Intel have moved in opposite directions with expansion plans. TSMC earlier this month pledged to invest an additional $100 billion in chip production in the U.S. Intel will delay the opening of its new site in the state of Ohio to around 2030.

TSMC makes more than 90% of the world’s most advanced chips for customers like Apple, Nvidia and AMD. The Taiwanese company is growing while smaller rivals like Samsung and Intel slow manufacturing expansions in the U.S. The U.S. government may put pressure on TSMC to help Intel become a viable second source, according to Jones.

“The U.S. government wants a U.S. vendor that has competitive technology,” Jones said. “Intel will have trouble getting there without help. For TSMC, it will be a safety valve against U.S. government pressure. That would be much better for TSMC than sharing fab capacity.”

Press reports in February said that TSMC was considering the acquisition of a controlling stake in Intel Foundry as urged by U.S. President Donald Trump. Analysts said the idea was highly unlikely.

Tan’s effort to reshape Intel will probably face internal resistance from people who want to restore the company to its position 20 years ago as the world’s largest chipmaker and keep the company intact, Baratte said.

Another group, including Tan, understands that Intel Foundry cannot catch up to TSMC without burning up to $200 billion, he added.

 
imho, I think LBT is dealt way better hand than PG with respect to timing of leading Intel. I mean, Pat fixed mostly the process issue with 5N4Y. Most of the high investment that needed to be spent to get back to process leadership/parity & to build the shells are done. LBT is taking over right about when things are turning around (reduced capex going forward, enough govt grants <if not cancelled> & partner contribution to offset most of the capex portion) and only thing remains now is execution. Pat is not without blame for some things like being over optimistic about PC market after COVID, missing the AI boom, not managing expectations for External Foundry & could have been little cost conscious. But it is unfortunate Pat is not at the helm when things turn.
If Intel turns around bigly, Pat G will be remebered very positively in history, for sure. To me, his main issues as CEO are over-optimism in planning and over-confidence in communication.
 
Question: Why is Intel rehiring people they previously let go?

Pat Gelsinger – ex-Intel, rehired as CEO, ousted.
Lip-Bu Tan – ex board member, ousted due to disagreements on how to turn the company around.

If Intel removed him over disagreements on turnaround, what will be different this time?

These two CEO hires suggest that Intel’s board of directors is the problem.
 
This is certainly going to be very interesting. Whether Intel can absorb the cultural and behavioural changes Lip-Bu is talking about (and has always followed) can't be certain. It's a whole new type of leadership for Intel and Intel will likely feel like a quite different company if he can pull this off. Certainly impressed that Intel persuaded him to sign on. And admire Lip-Bu's guts for signing up for the toughest job in the industry right now.

Intel's corporate culture is important, but its IDM business model is the fundamental problem. I believe the right business model can shape an appropriate corporate culture and ultimately determine success or failure. A flawed business model will lead to countless poor strategies and misguided decisions that cannot be salvaged, even by the most talented CEOs or board of directors.

Yes, there are still some successful IDMs, such as Texas Instruments, but they operate in the specialty or analog/mixed-signal business. In the leading-edge digital logic semiconductor industry, there are only two IDMs - Intel and Samsung - both of which are struggling.
 

Analysis: Intel's new CEO should merge Intel Foundry with GF to challenge TSMC's reign​

M. S. Lin

M. S. Lin, a former Fab director at TSMC. Credit: Joseph Chen

Lin believes that Lip-Bu Tan's best course of action to revitalize Intel's foundry business is to emulate the successful strategy of AMD CEO Lisa Su. "Tan should consider merging Intel Foundry Services (IFS) with GlobalFoundries," Lin emphasized. "This would allow Intel to focus on designing and developing its own chip products while fostering a strong, US-based foundry to directly compete with TSMC in the US"

According to Lin, this move would be a game-changer for the industry. "The biggest benefit is that the US would have its own homegrown foundry powerhouse to rival TSMC, which is critical for both industry competition and national security," he explained. "This is also the best strategy for the U.S. to rebuild its semiconductor manufacturing capabilities."

 
Besides Intel, Samsung is the other company doing business with its own chip products besides its foundry for external customers. Interestingly, Samsung insisted to keep them together instead of splitting them. There is more to navigate in this situation. Samsung Foundry is arguably doing better than Intel Foundry right now.
Samsung is in the middle of a prolonged commodity memory downturn, a crumbling embedded business, and image sensors and Exynos got booted out of both Chinese and many of Samsung's own phones. So that obscures the situation, since Samsung's financials are far more opaque. They also missed the margin boon that is the HBM boom, and their inferior process technologies lead to margin pressure on both foundry and memory wafers sold. Even when Samsung's internal business was healthier, they had all Qualcomm SOCs, and NVIDIA's client parts they had worse scale on the 2 most leading logic nodes than Intel internal only. From the third party estimates from analyst firms who seem to have at least reasonable first order estimates, it is still estimated even when you cut Intel's product volumes in half that Intel still runs a significant volume advantage even if Samsung didn't lose any scale. And let's be honest; they probably did lose a lot of scale on account of smartphone sluggishness and many phones not using Samsung foundry anymore. I would make the argument that Samsung's failures speak more to Samsung failing at the industry's fundamentals rather than the IDM model.

Korean media seems to talk about Samsung capacity plans like it is alot. But the numbers I have seen them quote are even worse than I would have ever expected (by like 2x). I wish it was more clear what was said from Samsung and what is a number pulled from their rears because they mix rumor and fact in such a way without citations that it is hard to parse. So I wouldn't plan on taking these numbers to the bank. I saw multiple outlets claim the SF3 line in Hwaseong would be converted to a 2nm line with combined capacity at 15k WSPM. BK claimed that the line would be 7k WSPM in early 2025 when SF2 supposedly starts HVM this year. TSMC has said that 3 N5 fab phase is would do just over 1M wafers per year (around 28K WSPM) back when Fab18 was first ramping, and in 2024 TSMC said that Fab21 phase 1 would reach 30K WSPM. If we assume a 15% complexity increase for N3 and then again for N2 that would mean that an N2 fab module should be around 21-23K WSPM. We are talking SF3 and SF2 capacity being less than 2/3 of a single TSMC fab. Considering that is SF2 is likely a simpler process than N2 (as SF2 is basically a heavily performance enhanced derivate of SF3 from Samsung's disclosures), we could be talking a little more than 1/2 of a fab. Meanwhile, TSMC is building 4 fabs for Fab 20, and 2 more fab modules in Kaohsiung. Even compared to Intel, 15K is anemic. For 18A, even if we reasonably assume Ohio will no longer be needed to fit the demand requirements for 18A, we are talking minimum Fab 42, Fab 52, Fab 62, and the D1 pilot line until space is needed for 14A or 10A ramps.

https://www.trendforce.com/news/202...pansion-in-hwaseong-and-pyeongtaek-next-year/ (Business Korea via Trendforce)
https://www.chosun.com/english/industry-en/2024/12/16/SMAOH7NISJAZXBL7LFX3BX2YVU/ (Chosun Daily)
 

Analysis: Intel's new CEO should merge Intel Foundry with GF to challenge TSMC's reign​

M. S. Lin

M. S. Lin, a former Fab director at TSMC. Credit: Joseph Chen

Lin believes that Lip-Bu Tan's best course of action to revitalize Intel's foundry business is to emulate the successful strategy of AMD CEO Lisa Su. "Tan should consider merging Intel Foundry Services (IFS) with GlobalFoundries," Lin emphasized. "This would allow Intel to focus on designing and developing its own chip products while fostering a strong, US-based foundry to directly compete with TSMC in the US"

According to Lin, this move would be a game-changer for the industry. "The biggest benefit is that the US would have its own homegrown foundry powerhouse to rival TSMC, which is critical for both industry competition and national security," he explained. "This is also the best strategy for the U.S. to rebuild its semiconductor manufacturing capabilities."


It's much more likely they will spin off Intel's CPU and peripheral IC business, and leave shareholders with fabs, not the other way around.

Money-wise, that makes more sense, especially given that whomever will be the buyer, will have to use Intel foundry.
 
Question: Why is Intel rehiring people they previously let go?

Pat Gelsinger – ex-Intel, rehired as CEO, ousted.
Lip-Bu Tan – ex board member, ousted due to disagreements on how to turn the company around.

If Intel removed him over disagreements on turnaround, what will be different this time?

These two CEO hires suggest that Intel’s board of directors is the problem.

Three things:

Lip-Bu Tan was not ousted.

Lip-Bu would not have accepted the CEO job unless the Intel board accepted his turn-around plan.

This will be a defining moment in Lip-Bu's storied career. He does not need a giant shit stain to finish it off.
 
Intel's corporate culture is important, but its IDM business model is the fundamental problem. I believe the right business model can shape an appropriate corporate culture and ultimately determine success or failure. A flawed business model will lead to countless poor strategies and misguided decisions that cannot be salvaged, even by the most talented CEOs or board of directors.

Yes, there are still some successful IDMs, such as Texas Instruments, but they operate in the specialty or analog/mixed-signal business. In the leading-edge digital logic semiconductor industry, there are only two IDMs - Intel and Samsung - both of which are struggling.
Great reminder. While it's great to see Lip-Bu Tan back, the business issues Intel faces today aren't entirely different than when Pat took the helm.

Customers for Intel Product vs. Intel Foundry are fundamentally different; many are even competitors. Will be interesting to see how Lip-Bu threads the needle for his "customer-centric" approach (i.e., whether Product or Foundry gets prioritized); we definitely saw Pat struggle here.

With Product/Foundry now split (at least financially), it's unlikely we'll see more of the storied product/foundry integration that makes the IDM model so powerful.

Paul, who is Chen?
He probably means Lip-Bu Tan. "Chen" and "Tan" are different romanizations of the same surname in Chinese (陳).
 
It's much more likely they will spin off Intel's CPU and peripheral IC business, and leave shareholders with fabs, not the other way around.

Money-wise, that makes more sense, especially given that whomever will be the buyer, will have to use Intel foundry.
He obviously wrote that to try to lessen the pressure on TSM...

For Lip-Bu, the direction is simple: maximise shareholder's value to what it is possible and not the other way around. Most of the value are in products. Why would you want to spin off design?

They already said they need to put more emphasis on products. Furthermore, to drive the valuation up, they need to have a meaningful share in the AI market, again that is in products.

 
Back
Top