Huawei CEO: Our AI chip is faster than the Nvidia A100 GPU in some tasks – Computer – News

Almost all of your posts in your psot log repeat the same thing, and I think I’ve had this discussion with you before. I came up with the same list of nonsense.

China is very active in the world of electronic espionage. The vast majority of cases according to parties like Crowdstrike, Secureworks, and Symantec employees come from China. Please note that these parties are commercially active and have no relations with the government.

What news did the Chinese release? I have a laundry list here: FooBarWidget in “Huawei CEO: Our AI chip is faster than Nvidia A100 GPU on some tasks”

Phew, a processor that can finally, after 4 years, surpass the A100 in some tasks. We already have devices that are much faster and more economical at all tasks.

Related to that, this was the list you’d been repeating for a year:

• Wind Tunnel: This is the newest, but NASA had a hypersonic wind tunnel in 1969. Of course the new ones are faster.

• The bridge builder, again, is larger, but not new.

• Salt resistant rice is simply grown in India i.e. Pukkali rice varieties.

And also a lot of advertising.

In terms of satellite telephony, you’re really underestimating how great a technical achievement this is. Satellite phones usually have a huge antenna. Huawei phones don’t have that: they have managed to make the antenna smaller.

I actually owned shares in a company that pioneered this technology in the US 10 years ago. The iPhone 14 was the first commercial smartphone to support this. So the technology has been around for a while. It doesn’t have a lot of use cases outside of specialized tasks.

Moreover, producing a 5G chip is not at all intuitive. Such a chip is analog technology and is a solid electrical architecture. It’s not something a typical CPU designer can do. In fact, Apple invested a lot of money in manufacturing its 5G chip, and after many years it completely failed.

Apple did not succeed, but neither did many other Western companies. Suddenly a hardware company is being asked to produce network infrastructure products themselves. They can’t produce the screens themselves, but they buy them. Your argument is very strange.

Finally, the idea or concept is completely useless. Especially with more advanced technologies, all the value is in whether you can produce it, how well and cheaply. New ideas that are difficult or expensive to make are of no use to the world. The power of China, which produces increasingly difficult things on a large scale, reliably and cheaply, is largely underestimated by those who worship modernity.

The world has no use for new ideas? Did you say that before the first smartphone, before the first electric car, before the first jet, etc.?

It costs more money to manufacture more complex devices, and China’s power to be able to produce more cheaply is now starting to become less likely. In addition, it is a country suffering from massive youth unemployment, a real estate bubble, and trying to recruit massive amounts of foreign talent to make things like chips.

Their 7nm chip at Huawei is a prestigious product, because it is very expensive, consumes much more power and has low throughput. So much for cheap.