[WOW] Huawei's HiSilicon Develops First RISC-V Design to Overcome Arm Restrictions

[Han Patriot]

And Huawei is free to make an Android fork as long as it doesn't use GMS.

What US origin? RISC-V is legally Swiss.

Its no use explaining to the kebab Kings, they only know how to bbq stuff.

Once the tech is open source, nobody can legally stop someone from copying it, even android is open source, that's why android and harmony is compatible.

 

[Arulmozhi Varman]

Is open source so is not subject to US export laws, because nothing is being sell, is a non profit foundation compose by multiple companies including Chinese ones. The US government control over RISC-V is almost the same as the Chinese control over RISC-V, the foundation can kick out every Chinese company but nothing will stop the Chinese from making and selling their own RISC-V processor.
That say, It could be illegal for a third company sell risc-v designed processors to Huawei but is not illegal for Huawei to design their owns.
I am not an expert but i don't think the processor ISA define the semiconductor manufacturing process like the microarchitecture does.

Making RISC V based micro processor isn't the issue or the big deal here. Lots of Indian universities have made micro processor based on RISC V arc cos it's open source. But when you take it commercially, there are only a few fabs who can do it. Let's say if Taiwan Semiconductor has 5 factories, only one among 5 would be capable of manufacturing and testing any design based on RISC V arc making the manufacturing cost really high. A real reason why this hasn't taken off from research and educational institution.
And yes, the microprocessor manufacturing is hugely dependent on architecture. It's not really difficult to come up with an architecture. But it has to work with existing manufacturing process or it has to be path breaking to be adopted by other companies.

What the other Chinese CCP bots don't understand is I m not against Huawei making or doing research on new arch. But the commerical difficulty in taking them to production. US control over the MC / semi industry is absolute.

 

[Han Patriot]

Making RISC V based micro processor isn't the issue or the big deal here. Lots of Indian universities have made micro processor based on RISC V arc cos it's open source. But when you take it commercially, there are only a few fabs who can do it. Let's say if Taiwan Semiconductor has 5 factories, only one among 5 would be capable of manufacturing and testing any design based on RISC V arc making the manufacturing cost really high. A real reason why this hasn't taken off from research and educational institution.
And yes, the microprocessor manufacturing is hugely dependent on architecture. It's not really difficult to come up with an architecture. But it has to work with existing manufacturing process or it has to be path breaking to be adopted by other companies.

What the other Chinese CCP bots don't understand is I m not against Huawei making or doing research on new arch. But the commerical difficulty in taking them to production. US control over the MC / semi industry is absolute.

Lotsa of Indians made it? Tell me which one is being used? Doing a uni project versus producing a commercially competitive chip requires billions in investment and research. You lil indics just talk and wallah, a chip supa powa.......yet you bozos need to import billions of stuff a year. LOL, are you guys born delusional or just plain naïve?

 

[CAPRICORN-88]

Sony and TSMC will build a chips factory in Kumamoto Perfecture under the auspices of Japan Ministry of Economy, Trade & Industry.

The joint investment will exceeds 1 trillion yen and it will be 40nm process.

 

[FairAndUnbiased]

Making RISC V based micro processor isn't the issue or the big deal here. Lots of Indian universities have made micro processor based on RISC V arc cos it's open source. But when you take it commercially, there are only a few fabs who can do it. Let's say if Taiwan Semiconductor has 5 factories, only one among 5 would be capable of manufacturing and testing any design based on RISC V arc making the manufacturing cost really high. A real reason why this hasn't taken off from research and educational institution.
And yes, the microprocessor manufacturing is hugely dependent on architecture. It's not really difficult to come up with an architecture. But it has to work with existing manufacturing process or it has to be path breaking to be adopted by other companies.

What the other Chinese CCP bots don't understand is I m not against Huawei making or doing research on new arch. But the commerical difficulty in taking them to production. US control over the MC / semi industry is absolute.

in general, tapeout costs depend more on the process type than on the architecture. and you can make any architecture you want as long as you have your own photomasks and a fab willing to take on a long term project (i.e. most big ones for big customers).

 

[Arulmozhi Varman]

in general, tapeout costs depend more on the process type than on the architecture. and you can make any architecture you want as long as you have your own photomasks and a fab willing to take on a long term project (i.e. most big ones for big customers).

Thank you for proving what exactly I was trying to say. It's what I have been saying. Finding a fab to manufacture this as a long-term project where the initial chips would cost high.

 

[FairAndUnbiased]

Thank you for proving what exactly I was trying to say. It's what I have been saying. Finding a fab to manufacture this as a long-term project where the initial chips would cost high.

The initial chips can be part of a multiproject wafer so the risks in the prototyping stage is low. For big companies the tapeout costs to production are not so bad, because they're going to make it back by volume and most fabs will be glad to take on a large project because that's revenue stability.

 

[JustAnotherPerson]

Making RISC V based micro processor isn't the issue or the big deal here. Lots of Indian universities have made micro processor based on RISC V arc cos it's open source. But when you take it commercially, there are only a few fabs who can do it. Let's say if Taiwan Semiconductor has 5 factories, only one among 5 would be capable of manufacturing and testing any design based on RISC V arc making the manufacturing cost really high. A real reason why this hasn't taken off from research and educational institution.
And yes, the microprocessor manufacturing is hugely dependent on architecture. It's not really difficult to come up with an architecture. But it has to work with existing manufacturing process or it has to be path breaking to be adopted by other companies.

What the other Chinese CCP bots don't understand is I m not against Huawei making or doing research on new arch. But the commerical difficulty in taking them to production. US control over the MC / semi industry is absolute.

Like i said, i am not an expert but as far i can understand the ISA doesn't define the manufacturing process only define the microprocessor design, as long you are able to make transistors, gates and IP blocks you will be able to make a RISC-V processor. So whatever is a FPGA, a fab or individual transistors really doesn't matter, SMIC or even Huawei future fabs can do it, i guess they will have their own IP blocks, testing tools, masks and others for that task.
You haven't done your research on RISC-V very well because commercially the ISA has already take off and is growing really fast, there is a big momentum in the industry with this ISA and a lot of companies selling and developing RISC-V IP, in fact i am using it right now as micro-controllers for agricultural sensors.