antiterror13
Brigadier
More bans are coming. Now even the unlicensed A800s may be banned from exported to China. Seems like a full on tech blockade
Go ahead, I think it is a great news for China's chip development
More bans are coming. Now even the unlicensed A800s may be banned from exported to China. Seems like a full on tech blockade
India can only afford to buy a 2 ply toilet paper not 3 ply.What if India decides to buy a lot more US treasuries? Does India have enough US$ for buying US treasuries?
India can only afford to buy a 2 ply toilet paper not 3 ply.
I really want to see an industry effort at standardising a programming model to rival CUDA instead of having AscendCL, SUPA, MUSA etc. all co-existing. You could allow for some vendor-specific APIs to account for quirky hardware features but should be de-duplicating the generic stuff as much as possible so that a developer doesn't need to learn three ways to code the same thingIn China instead it seems everybody wants to build their own model. There is a huge and widely spread duplication of efforts and energies, across a huge number of actors, big and small, old and new.
Please use the search function. I have mentioned this in multiple places in this threadHow do the latest Huawei chips match up with Nvidia products?
That would be idea, but in reality, most people have already gravitated toward huawei ecosystem. China already has 1.5 million Ascend developers. What can't you make with that many developers?Changing the Chinese AI ecosystem from CUDA to a more open source one using alternatives like OpenCL or SysCl to break the vendor lock-in nightmare that Nvidia has over the AI world will not only benefit China but the entire world.
i mean there are plenty of other domestic options like Biren, Tianshu, Moore Thread & MetaX that are being used by different smart computing projects. It does seem to me Huawei has the whole stack and is sanction proof at this point. So, it's an up hill struggle for everyone else.Big tech companies might want Huawei to spin off their AI chip business before they buy from them
What if India decides to buy a lot more US treasuries? Does India have enough US$ for buying US treasuries?
Bro, here's my thinking, just ask a Tesla customer which vehicle they prefer, the one build in Shanghai or in Fremont.But they are building more iPhones now.
Talk is cheap, is $600 billion international reserve enough to cover a trade deficit of nearly $150 billion in 2023 and rising.What if they are stupid enough and actually buy more US treasuries?
Sir from Remittances and BPO.India has been in deficit in current account balance and trade for years .. where the US$ come from?
whaaaaat??One difference I see in AI marketplace development in US and China is that in US companies are willing to make synergies and collaborate among them, for instance there are only few big players with big models and big GPU farms, and many companies that leverage them for specific applications and market niches.
In China instead it seems everybody wants to build their own model. There is a huge and widely spread duplication of efforts and energies, across a huge number of actors, big and small, old and new.