6 ms·

Momentum builds for open-source processors (2001)

56 points by ashitlerferad 9y ago

lkcl 9y ago
the cost of doing say a 180nm chip is somewhere around... USD $10,000. that's $10,000 per shot... and you end up with something around... maybe... 200mhz maximum? and consuming vast amounts of power, relative to today's technology? here's the really interesting thing: i've been in touch with the head of the shakti team, madhu, and he tells me that the Indian Govt has basically given him CARTE BLANCH to design processors for primarily the indian market but also world-wide distribution. the "catch"? due to Intel getting into bed with the NSA, the entire chip and its associated source code MUST be entirely libre, right down to the bedrock. funnily enough, i don't have a problem with that... do you? :) so in discussions with him last week (he is very busy on a tape-out deadline at the moment), he said that he would be happy to offer the open hardware community free access to the university's 180nm fab, to be able to do proof-of-concept testing, and that he also wants to make available his team and expertise to create an entirely libre RISC-V multi-core SoC. i have done a preliminary design here: http://rhombus-tech.net/riscv/shakti/m_class/ http://rhombus-tech.net/riscv/shakti/m_class/ various foundries have offered free MVP samples, and various other companies have offered free access to tools that normally cost USD $80m to license. this on the basis that the Indian local market is so... INSANELY LARGE that they recognise that they will make literally billions of dollars over the next few years. so it's getting very interesting.
- Shoothe 9y ago
  Wow, that's really interesting. Is there a web page where one could watch the progress or read the news about this?
monocasa 9y ago
Does anyone here know why OpenCores atrophied so much between now and then? It had so much going for it.
- mlinksva 9y ago
  I don't really know, but maybe for one thing 2001 was still when x86 was destroying its RISC competition and getting much faster every year, and maybe before it was obvious ARM would be huge; there wasn't appetite for alternatives, let alone room for bottom-up adoption of an open architecture? OpenCores seems to still exist as does a relatively recent (2015) splinter LibreCores/Free and Open Source Silicon Foundation (FOSSi). Wikipedia says OpenCores was always a commercial organization; maybe it was not entirely clear how other commercial entities could relate? OpenCore's OpenRISC is LGPL/GPL which could have further put off commercial entities (whether it should have or not)? There's an associated annual conference https://orconf.org/ https://orconf.org/ which seems to be about any open chips topic now, unsurprisingly including several RISC-V talks last year.
- Analemma_ 9y ago
  I don’t think it was ever viable to begin with. Remember that this was when “Linux” was what “Bitcoin” is now: a word you stuck in your company name to boost your share price based on nothing but bubble hype. That looks to be what happened here. Also, it takes stupefying amounts of capital to be in the CPU business, especially when starting from scratch, and if that looked possible at one point, it certainly didn’t when the dotcom bubble ended.
  - BenoitP 9y ago
    I think the timing is pretty perfect for a RISC-V open eco-system to flourish in a Cambrian manner. With silicon scaling laws getting walls, Intel's ability to have an efficient general purpose CPU for every workload is fading. The only way to provide better performance is to provide specialization, now. Processing will specialize in the following: * databases: SQL projections in special hardware data slicing CPU instructions. Hardware-accelerated hashing, sorting, filtering. * dynamic languages: Special code caching for JITted code. Memory models that won't have to be emulated. Memory barriers and copy on write to help the GCs do the work. * AI: We're already seeing this with Google's TPUs: Low precision tensor operations. We're going to see more specializations inside this segment too. Tensorflow today has CPU, GPU and TPU backends. In a stable high volume configuration such as vision for an autonomous car, why not provide a fourth backend for a Tensorflow pipeline: VHDL/Scala? * IoT/Sensors: Extremely low power modes, on-chip LoRa. * Network: this is already happening with microsoft using FPGAs to help do the routing. Once settled, some use cases will have their silicon edition. ---- This will lead to a thining of the margins of Intel, NVidia and ARM, to the fragmentation of the monopolies, and less money will go to the lawyers.
olfactory 9y ago
Ironically now that it is known that the major vulnerabilities found have to do with processor features that have probably been the most protected IP in the entire design. I'd argue that for most applications fab isn't really all that relevant as a secret sauce. For any consumer device the benefit of an open core and a custom SoC that includes the CPU and other key parts will likely be acceptable even with 5 year old fab technology.

56 points by ashitlerferad 9y ago

lkcl 9y ago

the cost of doing say a 180nm chip is somewhere around... USD $10,000. that's $10,000 per shot... and you end up with something around... maybe... 200mhz maximum? and consuming vast amounts of power, relative to today's technology? here's the really interesting thing: i've been in touch with the head of the shakti team, madhu, and he tells me that the Indian Govt has basically given him CARTE BLANCH to design processors for primarily the indian market but also world-wide distribution. the "catch"? due to Intel getting into bed with the NSA, the entire chip and its associated source code MUST be entirely libre, right down to the bedrock. funnily enough, i don't have a problem with that... do you? :) so in discussions with him last week (he is very busy on a tape-out deadline at the moment), he said that he would be happy to offer the open hardware community free access to the university's 180nm fab, to be able to do proof-of-concept testing, and that he also wants to make available his team and expertise to create an entirely libre RISC-V multi-core SoC. i have done a preliminary design here: http://rhombus-tech.net/riscv/shakti/m_class/ http://rhombus-tech.net/riscv/shakti/m_class/ various foundries have offered free MVP samples, and various other companies have offered free access to tools that normally cost USD $80m to license. this on the basis that the Indian local market is so... INSANELY LARGE that they recognise that they will make literally billions of dollars over the next few years. so it's getting very interesting.

Shoothe 9y ago
Wow, that's really interesting. Is there a web page where one could watch the progress or read the news about this?

monocasa 9y ago

Does anyone here know why OpenCores atrophied so much between now and then? It had so much going for it.

mlinksva 9y ago
I don't really know, but maybe for one thing 2001 was still when x86 was destroying its RISC competition and getting much faster every year, and maybe before it was obvious ARM would be huge; there wasn't appetite for alternatives, let alone room for bottom-up adoption of an open architecture? OpenCores seems to still exist as does a relatively recent (2015) splinter LibreCores/Free and Open Source Silicon Foundation (FOSSi). Wikipedia says OpenCores was always a commercial organization; maybe it was not entirely clear how other commercial entities could relate? OpenCore's OpenRISC is LGPL/GPL which could have further put off commercial entities (whether it should have or not)? There's an associated annual conference https://orconf.org/ https://orconf.org/ which seems to be about any open chips topic now, unsurprisingly including several RISC-V talks last year.
Analemma_ 9y ago
I don’t think it was ever viable to begin with. Remember that this was when “Linux” was what “Bitcoin” is now: a word you stuck in your company name to boost your share price based on nothing but bubble hype. That looks to be what happened here. Also, it takes stupefying amounts of capital to be in the CPU business, especially when starting from scratch, and if that looked possible at one point, it certainly didn’t when the dotcom bubble ended.
- BenoitP 9y ago
  I think the timing is pretty perfect for a RISC-V open eco-system to flourish in a Cambrian manner. With silicon scaling laws getting walls, Intel's ability to have an efficient general purpose CPU for every workload is fading. The only way to provide better performance is to provide specialization, now. Processing will specialize in the following: * databases: SQL projections in special hardware data slicing CPU instructions. Hardware-accelerated hashing, sorting, filtering. * dynamic languages: Special code caching for JITted code. Memory models that won't have to be emulated. Memory barriers and copy on write to help the GCs do the work. * AI: We're already seeing this with Google's TPUs: Low precision tensor operations. We're going to see more specializations inside this segment too. Tensorflow today has CPU, GPU and TPU backends. In a stable high volume configuration such as vision for an autonomous car, why not provide a fourth backend for a Tensorflow pipeline: VHDL/Scala? * IoT/Sensors: Extremely low power modes, on-chip LoRa. * Network: this is already happening with microsoft using FPGAs to help do the routing. Once settled, some use cases will have their silicon edition. ---- This will lead to a thining of the margins of Intel, NVidia and ARM, to the fragmentation of the monopolies, and less money will go to the lawyers.

olfactory 9y ago

Ironically now that it is known that the major vulnerabilities found have to do with processor features that have probably been the most protected IP in the entire design. I'd argue that for most applications fab isn't really all that relevant as a secret sauce. For any consumer device the benefit of an open core and a custom SoC that includes the CPU and other key parts will likely be acceptable even with 5 year old fab technology.