Intel's MESO Transistor Project Could See Results in Two to Five Years

AlphaAtlas

[H]ard|Gawd
Staff member
Joined
Mar 3, 2018
Messages
1,713
Late last year, Intel announced that they were working on a new type of transistor that could offer a massive performance leap over current CMOS chips. "MESO" transistors, as they call them, could operate at voltages as low as 100mV, but at the time, Intel said the technology was at least a decade away from commercialization. But Today, in an interview with VentureBeat, an Intel researcher said he is "excited about spin-off results MESO is likely to produce within the next two to five years." AI accelerators are supposedly less complicated an more fault tolerant that traditional chip designs, and MESO's characteristics are "coincidentally' well suited to neural network architectures, meaning they could hit the market sooner rather than later.

Khosrowshahi: CPUs, which are the most commonplace when you're building silicon, are oddly enough the hardest thing to build. But in AI, it's a simpler architecture. AI has regular patterns, it's mostly compute and interconnect, and memories. Also, neural networks are very tolerant to inhomogeneities in the substrate itself. So I feel this type of technology will be adopted sooner than expected in the AI space. By 2025, it’s going to be biggest thing... Young: If we can get these improvements in power-performance - MESO will be a 10 to 30 times better power-performance or energy-delay product - but let's say we only get a 2X improvement. That gives us, for a given power into the device, a 2X performance benefit, so it's a huge leg up on the competition. That's what drives this. Not only is this good for my company but it's an opportunity for the industry. The research is open, because we have so much heavy lifting to do with these materials. But if this is a thing that we as an industry can get a hold of, this could be a game changer for the semiconductor industry. It will take it through this curve that has been flattening. We may accelerate again. And that would be really neat.
 
  • Like
Reactions: PaulP
like this
Immediate first thought:

researcher_translation.png
 
Back
Top