News Moore Threads MTT S80 gets DX12 support in beta driver, two years after launch — but 3D performance is still lags the GTX 1650

[Admin]

Moore Threads MTT S80 gets its DX12 drivers, shows performance of GeForce GTX 1650.

Moore Threads MTT S80 gets DX12 support in beta driver, two years after launch — but 3D performance is still lags the GTX 1650 : Read more

 

[das_stig]

With China pushing towards its chips and Linux, whose to say what level of performance can be achieved in the next 10 years and the new x86S group have to invite Chinese companies to the table.

 

[evdjj3j]

I wonder why they have such a strange name, Moore Threads? Is it a reference to Moore's Law?

 

[King_V]

I wonder why they have such a strange name, Moore Threads? Is it a reference to Moore's Law?

I assumed it was meant to be a play off both Moore's Law, and More Threads.

 

[King_V]

With China pushing towards its chips and Linux, whose to say what level of performance can be achieved in the next 10 years and the new x86S group have to invite Chinese companies to the table.

Invite them to what table?

 

[razor512]

The performance is quite poor from what is described, and it overall seems like a company rushed to create a GPU in response to various trade embargos.

 

[mitch074]

The performance is quite poor from what is described, and it overall seems like a company rushed to create a GPU in response to various trade embargos.

Very likely. Still, not many company could even get that far - so, it is an accomplishment in itself.
I mean, look at Intel and its multi million dollars investments in GPUs - considering they already had integrated GPUs and poached head honchos from AMD, which company looks worse? The one starting with complete testing suites, cash and know-how, or the one starting from scratch?

 

[Joseph_138]

With China pushing towards its chips and Linux, whose to say what level of performance can be achieved in the next 10 years and the new x86S group have to invite Chinese companies to the table.

There's no way they get down to 4-5nm in the next 10 years, and 10 years from now the rest of the world will have moved on from that. The Russians are also trying because of the sanctions, but I think they can't do better than 28nm, and even at that, their yields aren't good. Realistically, 32nm, or somewhere around there, is probably the best they can do without outside help, without wasting tons of silicon.

 

[Kondamin]

Very likely. Still, not many company could even get that far - so, it is an accomplishment in itself.
I mean, look at Intel and its multi million dollars investments in GPUs - considering they already had integrated GPUs and poached head honchos from AMD, which company looks worse? The one starting with complete testing suites, cash and know-how, or the one starting from scratch?

They didn’t start from scratch they are use the power VR architecture, the same thing that apple is using under the hood of their “in house graphics”