Intel Launches Foundry Biz with $20B Fab Spend, Teams Up with IBM

By Tiffany Trader

March 24, 2021

Pat Gelsinger may have left Intel 11 years ago to secure his first CEO post elsewhere, but he returned in February to claim the company’s CEO spot — and he’s got a bold plan for the future. Gelsinger, who rejoined the company five weeks ago after eight years leading VMware, laid out his plan Tuesday in a webcast titled “Intel Unleashed – Engineering the Future.” The announcement had three thrusts: getting 7nm back on track with EUV technology, launching Intel Foundry Services with a $20 billion investment in two new fabs, and a new research collaboration with IBM. The company also announced it is reviving the spirit of the Intel Developer Forum event with the launch of Intel On, planned for October in San Francisco.

Out of the gate, Gelsinger said that the company’s 7nm process issue had been resolved, thanks in large part to Intel’s full embrace of EUV, where the company was previously on the “wrong side of the EUV maturity curve.” 

“We’ve rearchitected and simplified our 7nm process flow, increasing our use of EUV by more than 100 percent,” said Gelsinger. “We have a very strong partnership with ASML and our plans to now stay on the leading edge of EUV usage are well underway.”

Intel CEO Pat Gelsinger holds up “Ponte Vecchio” GPU at March 23, 2021, virtual event.

While timeline details were thin for 7nm datacenter CPU and GPU parts, Gelsinger said “confidence in 7nm health and competitiveness is accelerating; leveraging our 7nm process, we are advancing the development of lean datacenter and client CPUs, starting with Meteor Lake, our high volume 2023 client product.” Intel said it expects to tape out the 7nm Meteor Lake in the second quarter of this year.

On the GPU side, Gelsinger showed off Intel’s exascale-focused Ponte Vecchio datacenter GPU, but did not say when it would be ready for primetime. Ponte Vecchio will presumably take advantage of Intel 7nm EUV technology, but Intel has still not publicly confirmed the process node of the GPU die (Xe-HPC) or who will be making it. Despite Intel’s having the chip “in-hand,” the status and timing of the Aurora supercomputer engine (and thus the Aurora supercomputer) is still uncertain. Previously slated to arrive at Argonne National Lab in 2021, Aurora’s ETA is now TBA. Fortunately, Argonne has received samples of the smaller Xe-HP GPU to help with exascale application development.

Holding up the recently minted Ponte Vecchio chip, Gelsinger reported that it uses more than 40 different tiles — 47 “XPU” tiles to be exact — integrated into a single package. “This package is a diverse combination of more than 100 billion transistors,” said Gelsinger, “manufactured in multiple process technologies and packaged using both Foveros and EMIB.”

(Intel released a brief video on Ponte Vecchio. Watch it here.)

In a Q&A session at the end of the “Intel Unleashed” event, Gelsinger acknowledged the company’s setbacks with process technology, but said that when it comes to 3D packaging technology, Intel has “perfect, unquestioned leadership.”

Intel plans to officially launch its 10nm third-generation Xeon Scalable “Ice Lake” chips in two weeks, and says customers are currently testing the next-gen 10nm “Sapphire Rapids” CPUs. Gelsinger indicated Sapphire Rapids will be in production by the end of the year, ramping in the first half of 2022. Intel’s first 7nm server CPU, codenamed “Grand Rapids,” is on the roadmap for 2023.

Intel’s vision for IDM 2.0

Gelsinger revealed the company’s IDM 2.0 strategy as an ambitious plan that will restore its leadership manufacturing performance. Described as a major evolution of the chip company’s integrated device manufacturing (IDM) model, IDM 2.0 represents the combination of three factors: 1) Intel’s internal factory network; 2) the expansion of Intel’s use of third-party foundry capacity across its portfolio and 3) the launch of Intel Foundry Services. 

Intel is betting that this tri-part strategy will deliver, in Gelsinger’s words, “leadership products, leadership costs and leadership supply.”

“That is the right strategy for Intel,” said the CEO, and one the company is making, it says, without as-yet-committed government support, although it hopes to see government investment from, e.g., the U.S. CHIPS Act, and said it is currently competing for a U.S. Department of Defense secure foundry contract.

With a vision to become a major provider of U.S.– and Europe-based foundry capacity, Intel has established a standalone foundry business (Intel Foundry Services) led by Dr. Randhir Thakur, who reports directly to Gelsinger.

“We will be differentiated from other foundry offerings with a combination of leading edge packaging and process technology, committed capacity in the U.S. and Europe — available for customers globally — and a world class IP portfolio that customers can choose from, including x86 cores, graphics, media, display, AI, interconnect, fabric and other critical foundational IP along with Arm and RISC-V ecosystem IPs,” said Gelsinger.

He added that the business unit will be completely dedicated to the success of its customers with full P&L responsibilities.

“We conservatively size the foundry opportunity as a $100 billion addressable market by 2025 with most of the growth coming from leading edge computing, which is our expertise. The majority of leading edge foundry capacity is concentrated in Asia, while the industry needs more geographically balanced manufacturing capacity,” said Gelsinger.

Intel said it will make the outlay expenditures necessary to be a world-class foundry with investments in industry-standard PDK models and simplified design rules. “We are building on our EDA partnerships with both Cadence and Synopsys to enable their design tools for our Intel foundry services,” said Gelsinger, adding the company will also make investments in its own competitive foundational IP offerings, leveraging standard interfaces as well as third party IP availability.

Intel is also expanding the relationships it has with TSMC, Samsung, UMC and Global Foundries and leveraging the foundry network, Gelsinger reported. “This will provide us with increased flexibility and scale we need to optimize our roadmaps for cost, performance, schedule and supply, giving us a unique competitive advantage,” the CEO said.

However, the CEO affirmed that Intel will continue to build the majority of its products in Intel fabs.

$20 billion for Two new Ocotillo fabs

To accelerate its IDM 2.0 strategy, Intel is spending $20 billion to build two new fabs at its Ocotillo campus in Chandler, Arizona. These fabs will support Intel’s needs, as well as provide committed capacity for foundry customers. 

Planning and construction will commence this year and the project is on-track to create more than 3,000 permanent high-tech, high-wage jobs, over 3,000 construction jobs and 15,000 long term jobs in Arizona, according to Intel.

Referencing very strong global demand for semiconductors, fueled by worldwide component shortages, Intel says it in a unique position to leverage this market opportunity. The company plans to announce additional factory locations in the U.S. and Europe within the next year.

Intel + IBM

In a move perhaps more surprising than Intel getting back into foundry business it had tried out and exited several years ago, Intel has announced a research partnership with long-time rival IBM, focused on creating next-generation logic and packaging technologies.

“For more than 50 years, the two companies have shared a deep commitment to scientific research, world-class engineering and a focus on bringing advanced semiconductor technologies to market,” Intel said in a statement. “These foundational technologies will help unleash the potential of data and advanced computation to create immense economic value focused on advanced silicon process and packaging technology.”

“By bringing together two of the best semiconductor research organizations in the world, we will greatly accelerate innovations in the semiconductor industry,” said IBM CEO Arvind Krisha, speaking at Intel Unleashed. “Let me also express our support for Intel’s decision to expand and bring a new foundry business that will bolster the United States competitiveness in semiconductors.”

Final takeaways

Gelsinger said on yesterday’s webcast that the theme of his young tenure as CEO of Intel is “execute, execute, execute.” The light irony of that statement aside, that is the core question. Can Intel execute? Can they build two fabs in two years? Can they streamline and modernize and not be distracted? Can they deliver promised and future parts?

The real milestone will be fab construction, realistic use of third party fabs and delivering parts on time. That’s a lot, but Intel has staged — and executed — comebacks before: when it switched focus from memory to microprocessors under the leadership of Andy Grove in the late 70s/early 80s, and when it took back the market share lost to AMD Opterons about a decade ago.

Gelsinger, a 30-year alum of Intel before leaving in 2009 to pursue a CEO path via EMC/VMware, was mentored under Grove.

“We’re bringing back the execution discipline of Intel,” Gelsinger said on yesterday’s webcast. “I call it the Grovian culture — that we do what we say we’re going to do, that we have that confidence in our execution, that our teams are fired up; we said we’re going to do X, we’re going to do 1.1X every time that we make a commitment.”

Jack Gold, principal analyst of J.Gold Associates, LLC., shared that Gelsinger is “known for being a ‘take action’ executive that doesn’t rest on past laurels.”

“Intel was for many years 2-3 years ahead of everyone else in the industry when it came to chip processing,” writes Gold in an analyst report he provided to HPCwire. “But about 3-4 years ago it began to stumble badly and now its biggest processing rivals, TSMC and Samsung, both of which supply chips to key “fabless” semiconductor companies like Qualcomm, Nvidia, and many others, are 2-3 years ahead of Intel. Gelsinger has stated publicly that Intel wants to regain the lead, but has not stated how soon that might happen. In our opinion, this could take 2-3 years at least to achieve, and of course the competition is not sitting still, so it will take a massive effort to accomplish. Intel does have enormous resources it can bring to bear on this task, so it’s certainly possible to achieve, but it won’t be easy.”

For the moment, despite Intel’s rocky track record of late with canceled and delayed products, the company is also facing a convergence of potentially favorable conditions: a new tech-focused CEO in Gelsinger, their not-to-be-underestimated IP (and in particular their packaging technology), the international rivalry that is catalyzing U.S. and EU investment, the partnerships the company is building and robust demand fueled by ongoing industry-wide component shortages. At the very least, Intel has done something big to shake up its image, appease investors and generate conversation.

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industry updates delivered to you every week!

Under The Wire: Nearly HPC News (June 13, 2024)

June 13, 2024

As managing editor of the major global HPC news source, the term "news fire hose" is often mentioned. The analogy is quite correct. In any given week, there are many interesting stories, and only a few ever become headli Read more…

Quantum Tech Sector Hiring Stays Soft

June 13, 2024

New job announcements in the quantum tech sector declined again last month, according to an Quantum Economic Development Consortium (QED-C) report issued last week. “Globally, the number of new, public postings for Qu Read more…

Labs Keep Supercomputers Alive for Ten Years as Vendors Pull Support Early

June 12, 2024

Laboratories are running supercomputers for much longer, beyond the typical lifespan, as vendors prematurely deprecate the hardware and stop providing support. A typical supercomputer lifecycle is about five to six years Read more…

MLPerf Training 4.0 – Nvidia Still King; Power and LLM Fine Tuning Added

June 12, 2024

There are really two stories packaged in the most recent MLPerf  Training 4.0 results, released today. The first, of course, is the results. Nvidia (currently king of accelerated computing) wins again, sweeping all nine Read more…

Highlights from GlobusWorld 2024: The Conference for Reimagining Research IT

June 11, 2024

The Globus user conference, now in its 22nd year, brought together over 180 researchers, system administrators, developers, and IT leaders from 55 top research computing centers, national labs, federal agencies, and univ Read more…

Nvidia Shipped 3.76 Million Data-center GPUs in 2023, According to Study

June 10, 2024

Nvidia had an explosive 2023 in data-center GPU shipments, which totaled roughly 3.76 million units, according to a study conducted by semiconductor analyst firm TechInsights. Nvidia's GPU shipments in 2023 grew by more Read more…

Under The Wire: Nearly HPC News (June 13, 2024)

June 13, 2024

As managing editor of the major global HPC news source, the term "news fire hose" is often mentioned. The analogy is quite correct. In any given week, there are Read more…

Labs Keep Supercomputers Alive for Ten Years as Vendors Pull Support Early

June 12, 2024

Laboratories are running supercomputers for much longer, beyond the typical lifespan, as vendors prematurely deprecate the hardware and stop providing support. Read more…

MLPerf Training 4.0 – Nvidia Still King; Power and LLM Fine Tuning Added

June 12, 2024

There are really two stories packaged in the most recent MLPerf  Training 4.0 results, released today. The first, of course, is the results. Nvidia (currently Read more…

Highlights from GlobusWorld 2024: The Conference for Reimagining Research IT

June 11, 2024

The Globus user conference, now in its 22nd year, brought together over 180 researchers, system administrators, developers, and IT leaders from 55 top research Read more…

Nvidia Shipped 3.76 Million Data-center GPUs in 2023, According to Study

June 10, 2024

Nvidia had an explosive 2023 in data-center GPU shipments, which totaled roughly 3.76 million units, according to a study conducted by semiconductor analyst fir Read more…

ASC24 Expert Perspective: Dongarra, Hoefler, Yong Lin

June 7, 2024

One of the great things about being at an ASC (Asia Supercomputer Community) cluster competition is getting the chance to interview various industry experts and Read more…

HPC and Climate: Coastal Hurricanes Around the World Are Intensifying Faster

June 6, 2024

Hurricanes are among the world's most destructive natural hazards. Their environment shapes their ability to deliver damage; conditions like warm ocean waters, Read more…

ASC24: The Battle, The Apps, and The Competitors

June 5, 2024

The ASC24 (Asia Supercomputer Community) Student Cluster Competition was one for the ages. More than 350 university teams worked for months in the preliminary competition to earn one of the 25 final competition slots. The winning teams... Read more…

Atos Outlines Plans to Get Acquired, and a Path Forward

May 21, 2024

Atos – via its subsidiary Eviden – is the second major supercomputer maker outside of HPE, while others have largely dropped out. The lack of integrators and Atos' financial turmoil have the HPC market worried. If Atos goes under, HPE will be the only major option for building large-scale systems. Read more…

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Everyone Except Nvidia Forms Ultra Accelerator Link (UALink) Consortium

May 30, 2024

Consider the GPU. An island of SIMD greatness that makes light work of matrix math. Originally designed to rapidly paint dots on a computer monitor, it was then Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Some Reasons Why Aurora Didn’t Take First Place in the Top500 List

May 15, 2024

The makers of the Aurora supercomputer, which is housed at the Argonne National Laboratory, gave some reasons why the system didn't make the top spot on the Top Read more…

Leading Solution Providers

Contributors

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

The NASA Black Hole Plunge

May 7, 2024

We have all thought about it. No one has done it, but now, thanks to HPC, we see what it looks like. Hold on to your feet because NASA has released videos of wh Read more…

Google Announces Sixth-generation AI Chip, a TPU Called Trillium

May 17, 2024

On Tuesday May 14th, Google announced its sixth-generation TPU (tensor processing unit) called Trillium.  The chip, essentially a TPU v6, is the company's l Read more…

Intel’s Next-gen Falcon Shores Coming Out in Late 2025 

April 30, 2024

It's a long wait for customers hanging on for Intel's next-generation GPU, Falcon Shores, which will be released in late 2025.  "Then we have a rich, a very Read more…

GenAI Having Major Impact on Data Culture, Survey Says

February 21, 2024

While 2023 was the year of GenAI, the adoption rates for GenAI did not match expectations. Most organizations are continuing to invest in GenAI but are yet to Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

Intel Plans Falcon Shores 2 GPU Supercomputing Chip for 2026  

August 8, 2023

Intel is planning to onboard a new version of the Falcon Shores chip in 2026, which is code-named Falcon Shores 2. The new product was announced by CEO Pat Gel Read more…

How the Chip Industry is Helping a Battery Company

May 8, 2024

Chip companies, once seen as engineering pure plays, are now at the center of geopolitical intrigue. Chip manufacturing firms, especially TSMC and Intel, have b Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire