You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
OAKLAND, Calif./HONG KONG, Feb 16 (Reuters) - Chip technology firm Arm China suffered a 90% drop in profit last year despite revenue rising more than 30% during the first year management appointed by SoftBank Group Corp. took over, according to a financial document reviewed by Reuters.
The company, set up in 2018 as a joint venture of British chip technology firm Arm Ltd, laid off nearly 100 employees last week, most of them engineers, Reuters reported exclusively on Friday.
Chip technology firm Arm China suffered a 90% drop in profit last year despite revenue rising more than 30% during the first year management appointed by SoftBank Group Corp took over, according to a financial document reviewed by Reuters.
Just like how x86 is a spent ISA that can’t scale any further and will get eclipsed by ARM in all categories any day now? My opinion all three ISAs aren’t going anywhere any time soon and will all found their niches/use cases. There is so much money/software/IP in the ARM and x86 ecosystems that people won’t just drop them for the sake of change. ARM and x86 are moving targets that will continue to increase performance, features, and efficiency every year.
Long story short ARM will never get good enough to run x86 better than x86, and risc V will never be good enough to run ARM code better than ARM. Without this occurring none of these ISAs will kill any of the others.
No one I'm aware of has proposed a plan for keeping China (or any other out of favor country) from using or even contributing to open source projects. People I've talked to active in the open source community thinks China has downloaded every project in Github, and is known to be developing its own Github alternative. It's probably based on Github. Android, for example, is still the most popular mobile OS in China, even after the boxing match with Google. But open source software is easier. Open source chip designs still need pretty good chip fabrication for use in many interesting HPC applications. I think part of the challenge is that most senior people in government are not deeply aware of how open source projects work, or the potential of a project like RISC-V, so they don't even know to ask the right questions.
Long story short ARM will never get good enough to run x86 better than x86, and risc V will never be good enough to run ARM code better than ARM. Without this occurring none of these ISAs will kill any of the others.
And anyway, most customers want an off the shelf implementation. A few of those are open source, but the quality ones, and likely the one optimized for the niche the customer cares about, will be proprietary and subject to licensing. So the correct question is whether the market for licensed IP in the RISC-V ecosystem will be more respectful and mutually beneficial than with prior hardware (and software) usage in China.
Not much point applauding the "market" if it has no rewards for the innovators.
And anyway, most customers want an off the shelf implementation. A few of those are open source, but the quality ones, and likely the one optimized for the niche the customer cares about, will be proprietary and subject to licensing. So the correct question is whether the market for licensed IP in the RISC-V ecosystem will be more respectful and mutually beneficial than with prior hardware (and software) usage in China.
Not much point applauding the "market" if it has no rewards for the innovators.
Most of the innovators I know in open source are working for companies which want to drive an open source project for their economic benefit. MongoDB is a good example. There is a free open source version, but many features are withheld for fee-only versions, and Mongo and other software companies have profitable support and industry-specific application divisions.
RISC-V is a bit unusual because it began as an academic project in UC Berkeley. The most interesting RISC-V company I've seen yet is Ventana Micro Systems:
/cloudfront-us-east-2.images.arcpublishing.com/reuters/7MONPV56URLGTAEYMCHQSQLYXI.jpg)
www.reuters.com
Android, for example, is still the most popular mobile OS in China, even after the boxing match with Google. But open source software is easier. Open source chip designs still need pretty good chip fabrication for use in many interesting HPC applications. I think part of the challenge is that most senior people in government are not deeply aware of how open source projects work, or the potential of a project like RISC-V, so they don't even know to ask the right questions.
