No such message found
Aiming to become the global leader in chip-scale photonic solutions by deploying Optical Interposer technology to enable the seamless integration of electronics and photonics for a broad range of vertical market applications
Message: What's that lie?
Your Vote:
Did you know?
You can earn activity points by filling your profile with information about yourself (what city you live in, your favorite team, blogs etc.)
Discussion]-IBM unveils world’s first 7nm chip (self.buildapc)
soumis il y a 10 heures par AWildRedditorApeared
On POET:
''
[–]Shandlar 23 points il y a 6 heures
They stressed an indium dope over Gallium Arsenide substrate to create quantum wells that facilitate ridiculously high electron mobilities. They believe they can achieve over 10x the mobility of ultra-high end silicon. They are already selling switches for the mobile market that outperform the theoretical max of silicon (higher switching Hz, at a much much lower power) and far outstrip what was commercially available before them.
...
[–]Shandlar 4 points il y a 5 heures
Appears to be trade secret. Supposedly their 100 gHz switches run at 0.3v. It's a good ways beyond my technical knowledge, and technical information is scarce at best. But they are already selling products, with their 40nm process ramping to volume production to start shipping in November. They must have solved the problem to at least some degree.
...
[–]Shandlar 7 points il y a 7 heures
http://nextbigfuture.com/2014/09/will-gallium-arsenide-finally-be-ready.html
Sept 2014 article from when they had a working 100nm process and needed capital to commercialize. They obtained said capital, and are selling components currently on this process. They apparently also have prototypes of new components on the 40nm process now using a retired and retrofitted TMSC foundry to be on sale this year.
The company has not been trying to build CPUs though, since they don't have the billions needed to R&D such a huge undertaking. They provide very high HZ, extremely low power switches for the mobile market and elsewhere.
The tech is definitely usable for CPUs though. I'm hoping their 40nm process goes swimmingly and one of the big boys will buy up the company and get a team working on a process. After 7nm, there is literally no where to go with silicon. We need to get started on something new if we expect any significant increases in CPU performance after 2020.
None of their products for sale are CMOS transistor chips though. Theoretically its a highly suitable technology for it, however.
...
12 Recommendations
Loading...
Loading...
New Message
Please
login
to post a reply