@HPC Podcast Archives - OrionX.net - HPC News Bytes – 20250317

Episode Date: March 17, 2025

- What to expect from Intel's new CEO - Did D-Wave achieve quantum computational supremacy on a useful problem? - "AI Woodstock" GTC25 is in San Jose, CA this week [audio mp3="https://orionx.net/wp-c...ontent/uploads/2025/03/HPCNB_20250317.mp3"][/audio] The post HPC News Bytes – 20250317 appeared first on OrionX.net.

Transcript
Discussion (0)
Starting point is 00:00:00 Welcome to HPC News Bites, a weekly show about important news in the world of supercomputing, AI, and other advanced technologies. Hi, everyone. Welcome to HPC News Bites. I'm Doug Black of InsideHPC, and with me is Shaheen Khan of OrionX.net. We'll start with the naming last week of new Intel CEO, Lipu Tan. To succeed, Pat Gelsinger left the company suddenly in December. Tan is 65 and was an Intel board member from 2022 until last August.
Starting point is 00:00:37 He's got impressive credentials, an advanced degree in nuclear engineering from MIT, a successful technology venture investor, and he led the turnaround of Cadence, which he became CEO of in 2009 when that company was in trouble, until 2022. The Wall Street Journal calls him a titan in the semiconductor business, and he's a winner of the Robert Noyce Award, the top honor from the S industry association. But he certainly faces a tall task at Intel, which is trying to catch up with TSMC on advanced chip manufacturing and with NVIDIA and AMD on chip design. GPUs in particular have been an ongoing problem for Intel, with NVIDIA continually
Starting point is 00:01:20 extending its lead in AI compute. I can't help noting Shaheen, the contrast between this appointment with the excitement generated three years ago when Gelsinger returned to Intel from VMware. He was soon making pronouncements about Intel recapturing chip industry leadership. Now he might argue that his plan needed more than three years to happen,
Starting point is 00:01:41 but as of now, Intel recapturing industry leadership is not in sight. Well, the market seemed to like the appointment. And for a board that seems to enjoy roller coaster rides, this appointment was as fast as and as good as it could be. So kudos to them. He's an excellent choice with great experience in all aspects of the business, financial, technology, software and hardware, global, including China, and
Starting point is 00:02:05 go to market. But Gelsinger was also an excellent, maybe even ideal choice to run and transform the company. So the question is, why would the board go through all this trouble and risk to replace an excellent choice with, thankfully as it turned out, another? And the answer must be that with a new CEO, they would get a new strategy that is more aligned with what they wanted. So we can safely assume strong alignment between the board and the CEO. And maybe that was never quite the case last time. While it is too early to tell how things
Starting point is 00:02:36 will be different, various guesses include deep and fast cuts to people and products, splitting the company, finding ways to disrupt the GPU part of the business, and strengthen work with the US government. All of that would also represent a recalibration of what the company is. But Fab 18A and 14A continue to produce good news and look close to completion. So it will not be surprising to see, let's say in a year's time, the new regime highlighting that as the big achievement that it is. A holy grail of quantum computing is quantum superiority, the ability of a quantum system to perform workloads faster
Starting point is 00:03:15 than the most powerful classical supercomputers. D-Wave created a stir last week reporting that its annealing quantum computer outperforms the Frontier exascale system at Oak Ridge National Lab, solving complex materials simulation problems. D-Wave's peer-reviewed paper in the journal Science, quote, validates this achievement as the world's first and only demonstration of quantum computational supremacy on a useful problem, D-Wave said. But at least two doubters emerged to refute D-Wave's news, and one industry analyst told me last week he is frustrated with unverified reports by quantum vendors. I agree quantum is such a rarefied technology that it's hard to get a grasp on what vendors
Starting point is 00:04:00 are telling us. Yeah, the main counterpoints came from the Flatiron Institute and New York University and EPFL, one of the top universities in the world in Lausanne, Switzerland, who all basically said D-Wave compared their work to a suitably slow algorithm and that there are in fact other classical algorithms that are faster and more accurate than what the quantum computer did. D-Wave countered that by saying the groups were solving a different, less demanding problem. So as usual, the progress is real, but the claims are being disputed.
Starting point is 00:04:33 The whole field is still evolving in unpredictable ways. Quantum annealers are analog quantum computers that are especially suitable for optimization problems. They start by creating a superposition of all possible states and then evolve the system towards its so-called ground state, a low-energy state where it is in strong equilibrium. If you can formulate a problem as a system that is seeking its ground state, you can see why quantum annealing is good for optimization. The name is drawn from metal annealing, where a metal is heated to a
Starting point is 00:05:05 high temperature and then slowly cooled and its crystal structure is allowed to arrange itself optimally. Quantum annealers are viewed as too specialized compared to gate-based systems. We should say D-Wave is the original quantum computing company, having been formed in 1999, spun out of the University of British Columbia in Canada and pursuing superconducting quantum annealing. In 2021, they also started the project to build a gate-based quantum computer. D-Wave went public via this PAC process, Special Purpose Acquisition Company, when it merged with DPCM Capital in August of 2022. Okay, we'll sign off by noting that the advanced computing world will be at GTC this week. 25,000 attendees are expected in person, including a sellout throng at San Jose's Professional
Starting point is 00:05:57 Sports Arena where Jensen Wong will deliver his keynote. For thoughts on what to expect at the conference, we suggest you check out our two most recent at HPC podcast conversations with Handel Jones and Ian Cutress, both of whom follow NVIDIA closely. All right, that's it for this episode. Thanks so much for being with us. HPC News Bytes is a production of OrionX
Starting point is 00:06:20 in association with Inside HPC. Shaheen Khan and Doug Black host the show. Every episode is featured on insidehpc.com and posted on orionx.net. Thank you for listening. ["Inside HPC"]

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.