.png/BROK/thumbnail/1600x900/format/jpg/quality/80/final-fantasy-14-buried-memory-(9).png)
.jpg/BROK/thumbnail/1600x900/format/jpg/quality/80/gothic-remake-(2).jpg)
AMD doesn’t want to use Intel fabs, but it might be tempted to use
TSMC is the leading foundry, helping AMD retake the lead in PC processing in its CPU and graphics divisions. It also provides advanced manufacturing processes for Apple, Nvidia, and now even Intel, and underpins a range of next-generation products, including the latest game consoles from Sony and Microsoft.
It also announced possible delays that could put it far behind the competition in the silicon node race, encouraging those customers to start considering whether Intel’s own foundry is worth a shot.
So yes, AMD chips made in Intel fabs?it is possible But it sounds unlikely. I mean, Nvidia is reportedly in talks to use Intel’s foundry services in the future.
Report on Seeking Alpha, titled “TSMC may be doomed‘ is definitely too strong, and probably tainted by the fact that the author is an Intel bull with a long-term stake in the company. But if you’re relying on TSMC for your future silicon dreams, its view of a year-long delay in the future N2 node is certainly worrying.
It’s worth noting that, unlike Intel, TSMC typically doesn’t share its roadmap and has been mum on its N2 (2nm) node. But we expect the N3 node (the last one to use the classic FinFET transistor technology so widely used in all current PC hardware) to be available in 2022, although it is now only expected to be in production in 2023.
However, it still depends on TSMC getting the tools to make chips for companies that have signed up to the N3 process. There have been reported issues with the delivery of the tools, although TSMC CEO CC Wei has said “we hope there won’t be any major issues”. Latest Earnings Conference Callhe also said it wasn’t “ready to share…how many problems we can solve.”
Still, all is well, if the usual two-year rhythm is followed, the N2 should be available in 2025, but while TSMC recently stated that “the N2 is on track”, CC Wei also noted that “in 2025, it will be in production, Probably towards the second half or the end of 2025. That’s our schedule.”
It also suggests that the product won’t appear until 2026.
Intel, on the other hand, has promised to release a large number of nodes in just a few years, almost inexplicably.recently Intel Investor Conference Keyvan Esfarjani explained in his IDM 2.0 presentation: “Our technology development organization, led by my good friend Dr. Ann Kelleher, is doing something unprecedented in Intel’s history, delivering five process nodes in four years. “
It’s almost unbelievable, and indeed many have questioned how likely Intel is to keep this going, but it has pushed back its schedule for a competing 20A process, nominally a similar 2nm design, using a similar new gate A pole-all-around (GAA) designed transistor is called a RibbonFET. And plans to launch the 18A node in the second half of 2024.
TSMC and AMD continue to deliver on their promises, and many more believe it will continue to deliver in the future
It has shown an 18A test wafer (although just an SRAM wafer, so not a super-complex chip at CPU or GPU scale), which shows that it has overcome the punitive problems faced by the 10nm node – which required years of latency sunk A few successive generations of chips and getting back on track in the node development game.
We spoke to Intel engineers and they noticed that Intel CEO Pat Gelsinger was very concerned about timing. Completion on schedule is paramount, and this regular, reliable cadence of new technology is critical to regaining the trust of the industry and building a sustainable foundry service.
This is arguably Intel’s biggest challenge.
Despite all the positive comments on the timeline and progress, everyone remembers the problems the previous leadership had with pushing the technology forward and how TSMC has dominated the frontier ever since. From there, AMD also began to re-emerge in the consumer market and now the server market.
TSMC, and even AMD, has been delivering on its promises, and compared to Intel’s promises, people still believe it will continue to deliver in the future. But if it manages to stay on track for the next few generations of Intel’s 4, 3, 20A, and 18A nodes, and TSMC fails to manufacture as expected, Intel’s expected technological lead could turn into dominance.
However, AMD has played the foundry game deftly and has never relied on TSMC’s leading nodes to provide technology. The upcoming Zen 4 CPUs just moved to TSMC’s N5. This somewhat insulates the company from potential delays, so unless N2 slips further, I doubt it will be necessary to jump ship to Intel’s fabs in the near future.
We all want the latest and greatest chips, right?It sucks not being able to have them when we want them
However, Apple has been at the forefront of TSMC’s new process technology, and delays could affect its product launches. While there is certainly zero chance that Apple will consider using Intel fabs for production, the apparent animosity between the two companies arose after Apple ditched the Core chip and developed its own.
However, while TSMC’s N2 and N3 nodes are expected to be delayed, digital times Today reports that Apple will be one of the first customers to adopt the two new production processes. Although there is another name mentioned in terms of early node adoption, that is Intel.
Regardless, Intel is betting on its own and TSMC’s nodes to produce products. Whether TSMC’s N2 problem is as bad as Intel’s 10nm nightmare, as some believe — and they probably aren’t — having both companies producing cutting-edge silicon for the semiconductor industry can only be a good thing for us consumers.
I mean, we all want the latest and greatest chips, right? It sucks not being able to have them when we want them. So no matter what you think of Intel as a company, we should expect it to deliver on its promise to make chip production available to everyone.