India reveals plan to become major RISC-V design and production player by 2023 India's government has announced a plan and roadmap for local semiconductor design and production, based on the open source RISC-V architecture, and set a goal of delivering world-class silicon by the end of next year. The Digital India RISC-V Microprocessor Program (DIR-V) will see Indian industry and academia team to develop …
India and Canada have a lot in common. Both our governments are prone to grand announcements that fizzle into nothing before the next election, or as soon as there is a change of primary leading party in government.
I wonder how much the Indian government wastes on its boondoggles? A lot of ours never get past the planning stage, but man do they plan! Reams and reams of documentation and diagrams that are to be reviewed and edited and gone over with a fine-toothed comb. If you think ISO9000 is overhead, you've clearly never worked for the military, the government, or a large corporation. *LOL*
It is a bold statement - "delivering world-class silicon by the end of next year." - but, it really is just loud words and is not achievable.
Similarly, the grand EU "save the IC world" plan will just suck in lots of Euros and, by the time it is producing anything saleable, the IC world (and TSMC) will have caught up to the point where there will be oversupply again.
And, I wonder, are India still receiving UK Foreign Aid? If they are then this RSC-V plan, along with their Space Program should be enough to convince the UK Government to turn the aid off. There are far more deserving countries out there.
TFA specifically say 2023 but I did wonder if the actual words were "next year". That would just be an acceleration of "in the next ten years" or "in the next five years". It's always ten years, or five years, every year.
If they're going to achieve anything their employee churn is going to have to be less than Infosys's.
... much derided my neoliberals...
... but government leadership can do things where The Market, or Industry alone, would not go.
I see that Qualcomm are interested in supporting the idea: https://auto.economictimes.indiatimes.com/news/auto-components/qualcomm-meity-partner-to-support-indian-semiconductor-startups/91187476
And, in another moment of contrariness : liking the microcontroller - Shakti - to a top flight Pentium processor of a generation ago seems wrong... Maybe IIT are interested in doing what CAS are doing, but everyone has to start somewhere.
Feels like this will go about as well as their moon landing.
Knowing the current power structure, they'll churn out a single ~30% performance per clock vs reference, low IPC chip that uses 4x the power of reference with an ("industry leading," of course) 15% production yield expectation. They'll then claim it was created entirely without outside influence or technology (despite being an open spec), but instead using plans found in some veda or other for political reasons.
Any poor benchmarks will, of course, be western or Chinese conspiracy/fake news depending on who is the big bad at the time.
The Indian government's recently developed tendency to make ridiculous statements about achieving world leading [insert thing here] from scratch at record low expense has me wondering whether the leadership is high on their own supply or just reaching the end of some massive grift.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
