News on China's scientific and technological development.

emblem21

Major
Registered Member
be careful what you wish for ;)
Hmmmm, I am not sure of the likelihood of something like that happening. It would be a freak event but something that I don't think is very likely to happen. Still something should happen to finally shut Pompeo and Trump up permanently since I believe the world would be better off without them and their lies
 

latenlazy

Brigadier
It also only took about a year for TSMC to move from DUVL 7nm to EUVL 7nm
One big value of EUV is that it makes sub 14nm nodes much easier to make again. Having to rely on multipatterning significantly upped the scale of difficulty for fabrication, which is partially why so many fabs ended up falling behind. If Chinese firms can get to an EUV scanner quicker, even if they don’t have a design ready for production, they can probably start researching and developing smaller nodes out of sequence using prototype builds, and move directly to that smaller node once that EUV scanner is ready for actual production. Of course this is a much higher risk development path, but if the willingness to spend and eat risk is there, it’s a feasible one, though it’s hard to say whether such an approach would be effective.
 

machupicu

Junior Member
Registered Member
Huawei won't abandon high-end smartphones despite US chipset ban

Please, Log in or Register to view URLs content!
..He said that China should adopt an advanced system with coordination of hardware and software so as to reduce reliance on single chips. "With the special 'China system,' both domestically made CPUs with conventional processes, a large amount of old-fashioned Intel CPUs and even easily accessed CPU chips for common personal computers can be taken advantage of to ensure that the establishment and operation of key infrastructure sectors and general data centers won't be impacted."
 

machupicu

Junior Member
Registered Member
..He said that China should adopt an advanced system with coordination of hardware and software so as to reduce reliance on single chips. "With the special 'China system,' both domestically made CPUs with conventional processes, a large amount of old-fashioned Intel CPUs and even easily accessed CPU chips for common personal computers can be taken advantage of to ensure that the establishment and operation of key infrastructure sectors and general data centers won't be impacted."
"I believe that it's actually national insecurity that's the cause. The U.S. is no longer competing with China in this 21st century high-tech world. Instead of taking the high road of competition that made the country a world leader, the U.S. is taking the low road of fear-mongering to technologically fence off and contain China.

U.S. decision-makers, from Trump on down, must know this as the data are crystal clear. According to the U.S. National Science Board, China passed U.S. expenditures in R&D last year for the first time, having grown by 18 percent a year since 2000. In fact, China has contributed 32 percent of all R&D spending growth since 2000, compared with 20 percent for the U.S. and 17 percent for the EU. And for the future, the OECD projects that the share of students with STEM degrees in the OECD and G20 countries will be: China, 37 percent; India, 26.7 percent, Russia, 4.5 percent; the U.S., 4.2 percent and Indonesia, 3.7 percent by rank for the top five countries.

Please, Log in or Register to view URLs content!
 

Aperture05

New Member
Registered Member
It also only took about a year for TSMC to move from DUVL 7nm to EUVL 7nm
I'm not sure if SMIC have developed a competitive DUV "7nm class" node as of yet. TSMC's 7N and 7NP are almost 100% DUV, with the 7NPP being 30%? EUV.

They said that N+1 will begin trial production Q4 this year, which indicates volume production by 2021-2022, which is great, but SMIC have allegedly said that N+1 will not be directly comparable in terms of performance to other nodes such as Samsung 10/8nm or TSMC 7N.

Given that N+2 is going to be on EUV, and that ASML does not appear to be granted an export license any time soon, I still maintain that by the time SMIC hits 7nm class performance, the node would be somewhat outdated already.
 

Aperture05

New Member
Registered Member
The difference is unlike China they had access to US tech.

That's not the issue. SMIC isn't really restricted from US tech right now (AFAIK). The issue is that they haven't developed a comparable node even on existing equipment.

N+1 will be exciting, yes, but it would bring the performance parity that people are hoping for.
 

Aperture05

New Member
Registered Member
I'm not sure if SMIC have developed a competitive DUV "7nm class" node as of yet. TSMC's 7N and 7NP are almost 100% DUV, with the 7NPP being 30%? EUV.

They said that N+1 will begin trial production Q4 this year, which indicates volume production by 2021-2022, which is great, but SMIC have allegedly said that N+1 will not be directly comparable in terms of performance to other nodes such as Samsung 10/8nm or TSMC 7N.

Given that N+2 is going to be on EUV, and that ASML does not appear to be granted an export license any time soon, I still maintain that by the time SMIC hits 7nm class performance, the node would be somewhat outdated already.
Ran out of time to edit, so I'll add on some more things

To put this into perspective, Samsung is allegedly going to introduce GAAFET 3nm volume production in 2022, with Intel to follow "within the next 5 years" with either the 7nm or 5nm node. TSMC has announced that it will pursue FinFET up to 3nm.



SMIC really needs to step up their research efforts in order to reach node parity.
 
Top