The long term solution is RISC-V which is free as it is open source.The first step to innovation is technology transfer.
The long term solution is RISC-V which is free as it is open source.The first step to innovation is technology transfer.
Im not sure how IP works in semi conductor but could a design that accelerates Arm processor be used to also accelerates RISC-V processors or is the same acceleration design in ARM, x86/x64 and RISC-V three separate Intellectual property.The long term solution is RISC-V which is free as it is open source.
Im not sure how IP works in semi conductor but could a design that accelerates Arm processor be used to also accelerates RISC-V processors or is the same acceleration design in ARM, x86/x64 and RISC-V three separate Intellectual property.
The very nature of RISC-V (open source and royalty free) make it impossible to restrict, in practice is just a book. a set of instructions that you have to follow to make a compiler or a CPU. From here is just developing the software tools, like simulators and EDA tools like IP cores to make the process simpler. Chinese RiSC-V developers could develop their own tools from zero if the want, -personally I don't know why anyone will want that given the vast disponibility of RISC-V tools out there, but as long they follow the ISA, Chinese processors will be able to run European code by example.
RISC-V is developed independently and did receive funding from USA (and obviously other countries as members in the foundation).
It is open-source, and the foundation had moved from USA to Switzerland to try to make it sanction proof.
But again, Switzerland is now known not to be a neutral country anymore.
As Android has never been subjected to US restriction (which not clear if there is any legal path to make RISC-V sanctionable.
Im not sure how IP works in semi conductor but could a design that accelerates Arm processor be used to also accelerates RISC-V processors or is the same acceleration design in ARM, x86/x64 and RISC-V three separate Intellectual property.
At contrary, general purpose computing is what this ISA is good for you can make from a server CPU to a tiny microcontroller, that is why so many companies had joined the foundation. Companies like Nvidia has made general purpose controller for their products other like Alibaba are making CPUs to power server applications to IOT devices. The problem is for specific applications like GPUs or media accelerators, there is not ready to go instructions for those. The RISC-V Vector and SIMD instructions are probably good enough but a GPU or a media accelerator is much more than general computing, is not impossible but difficult.The problem with RISC-V is that it is still quite far from usable for general purpose.
It is mostly something you would use for, say, a hard disk controller or something like that.
It's called the loongson. Apparently it's quite disliked on chinese socmed and a lot of people say it's uselessI think there was a Chinese MIPS processor that did that, it had some blocks dedicated to x86 emulation. I believe the problem with these kinds of designs is that you are dedicating hardware to running software at 75% speed at best, when it could be 100+% performance by porting the software properly (since these emulation blocks are not being utilized with native applications). Basically a dead end