4 ms·

That is pretty awesome. So three 'open' RISC architectures (RISC-V, MIPS, SPARC) with varying amounts of ecosystem support. That the patent stuff seems to be d

That is pretty awesome. So three 'open' RISC architectures (RISC-V, MIPS, SPARC) with varying amounts of ecosystem support. That the patent stuff seems to be dealt with is key here, one of the challenges with working with MIPS and SPARC was always that their 'parent' organizations (SGI and Sun respectively) were essentially patent trolls when it came to third parties being successful with the architecture. Now that SGI and Sun are just footnotes in the computer industry, perhaps these architectures can flourish.

baybal2 8y ago
> challenges with working with MIPS and SPARC was always that their 'parent' organizations (SGI and Sun respectively) were essentially patent trolls when it came to third parties being successful with the architecture. And that was in part what has killed them. ARM only got so far thanks to their no questions asked attitude. I heard from people from Chinese fabless companies of ARM DHLing usb sticks to them in the early A8/A9 era, and flying support engineers to them the same day. Comparing MIPS to them, they had an obviously bad attitude. First, they barely sold complete solutions, nor worked on their integration (DMA controllers, memory controllers, bus interfaces and PHYs) Second, they were compete slowpokes. People are telling of them spending weeks to reply to very clear licensing applications, and them nor being able to say a word without going through their lawyers. You were essentially speaking not with sales, or engineers, but their legal department. Third, patently bad support. Only when they hit the terminal stage, they seemed to notice that they were doing something wrong
- ChuckMcM 8y ago
  > Only when they hit the terminal stage, they seemed to notice that they were doing something wrong Sadly this is the motivation for a lot of "correct" decisions. It is the first thing I look for when I read a startup pitch, a misunderstanding of where the value is and what the customer perceived risks are for adoption. Too often I've heard entrepreneurs or senior managers say, "But if we did that, this business model wouldn't work." And I patiently point out to them that if they can't get customer engagement with their current business model, it already doesn't work so they need to change something. Typically it is re-think the business model so they can focus on what is valuable to the customer and what isn't. ARM Holdings can be disrupted by a business model that can keep software coherence high while not requiring an architecture license to add special sauce. That puts more pressure on the foundation/management entity but it trades value (architecture extension) for adoption. The key is partnering with those who would extend the architecture to keep it aligned with key architectural invariants, and when possible allowing really useful extensions to move into the mainstream.
hawski 8y ago
And SuperH which patents have expired - http://j-core.org/ http://j-core.org/
snaky 8y ago
Maybe the future is not about fixed instruction sets at all [1]. Especially with the rise of high-level hardware synthesis methods, platforms and tools like Kami [2]. Of course, tooling and C compilers for the non-standard ISA are the main obstacles on this way. [1] "Synthesis of Processor Instruction Sets from High-Level ISA Specifications" IEEE Transactions on Computers 2014, 63(6), 1552-1566 [2] https://deepspec.org/entry/Project/Kami https://deepspec.org/entry/Project/Kami
- ChuckMcM 8y ago
  > Of course, tooling and C compilers for the non-standard ISA are the main obstacles on this way. Yup, see the Itanium architecture for how pernicious this can be to the survival of an architecture.
  - snaky 8y ago
    What exactly is the lesson of Itanium is debatable. > Q: I feel like deja vu - at Hot Chips, Intel introduced VLIW-concept Itanium that pushed complexity onto the compiler. I see traces of that here. What are you doing to avoid the Itanium traps? How will you avoid IP from Intel? > A: Itanium was in-order VLIW, hope people will build compiler to get perf. We came from opposite direction - we use dynamic scheduling. We are not VLIW, every node defines sub-graphs and dependent instructions. We designed the compiler first. We build hardware around the compiler, Intel approach the opposite. https://www.anandtech.com/show/13255/hot-chips-2018-tachyum-prodigy-cpu-live-blog https://www.anandtech.com/show/13255/hot-chips-2018-tachyum-... I don't think that's the way of the future anyway, but C is not the only programming language, fortunately. Generating the ISA, and then CPU from specification, the next logical step is to generate highly optimized lower-than-C-level code from high-level specifications too.
  - zozbot123 8y ago
    The Itanium folks actually put a lot of effort into tooling and compilers for the architecture, back in the early 2000s - even FLOSS tooling. The problem on that side was that a community never coalesced around Itanium, the way it did around e.g. amd64, or nowadays around riscv.
rurban 8y ago
POWER8 is also open now: https://openpowerfoundation.org/ https://openpowerfoundation.org/ By far the best of those. But IBM, and it's not POWER9.
- ksec 8y ago
  >and it's not POWER9 Surely POWER9 is open as well? Otherwise how did Raptor ship their POWER9 CPU?
  - bildung 8y ago
    Don't they just resell the CPUs? I think the mainboards are designed and produced by them, but not the CPUs.
  - rurban 8y ago
    Power9 only has open firmware and ISA, but not the verilog. They just resell it, to Raptor and Google-Rackspace mostly.
- floatboth 8y ago
  It's not clear what exactly is open about OpenPOWER. Firmware, sure. But is the ISA actually royalty-free?
ineedasername 8y ago
patent stuff seems to be deal with....SGI and Sun are just footnotes Eh... In Sun's case they were acquired by Oracle. I don't think that will have improved the situation. You'll see lots of bad anecdotes about Oracle here & there. You might wonder "huh, is all that really true?" Well, I've dealt with Oracle. I have yet to see something bad about them that didn't jibe with my own experience: I won't give info that might dox me, but there was a fairly visible lawsuit involved.

baybal2 8y ago

> challenges with working with MIPS and SPARC was always that their 'parent' organizations (SGI and Sun respectively) were essentially patent trolls when it came to third parties being successful with the architecture. And that was in part what has killed them. ARM only got so far thanks to their no questions asked attitude. I heard from people from Chinese fabless companies of ARM DHLing usb sticks to them in the early A8/A9 era, and flying support engineers to them the same day. Comparing MIPS to them, they had an obviously bad attitude. First, they barely sold complete solutions, nor worked on their integration (DMA controllers, memory controllers, bus interfaces and PHYs) Second, they were compete slowpokes. People are telling of them spending weeks to reply to very clear licensing applications, and them nor being able to say a word without going through their lawyers. You were essentially speaking not with sales, or engineers, but their legal department. Third, patently bad support. Only when they hit the terminal stage, they seemed to notice that they were doing something wrong

ChuckMcM 8y ago
> Only when they hit the terminal stage, they seemed to notice that they were doing something wrong Sadly this is the motivation for a lot of "correct" decisions. It is the first thing I look for when I read a startup pitch, a misunderstanding of where the value is and what the customer perceived risks are for adoption. Too often I've heard entrepreneurs or senior managers say, "But if we did that, this business model wouldn't work." And I patiently point out to them that if they can't get customer engagement with their current business model, it already doesn't work so they need to change something. Typically it is re-think the business model so they can focus on what is valuable to the customer and what isn't. ARM Holdings can be disrupted by a business model that can keep software coherence high while not requiring an architecture license to add special sauce. That puts more pressure on the foundation/management entity but it trades value (architecture extension) for adoption. The key is partnering with those who would extend the architecture to keep it aligned with key architectural invariants, and when possible allowing really useful extensions to move into the mainstream.

hawski 8y ago

And SuperH which patents have expired - http://j-core.org/ http://j-core.org/

snaky 8y ago

Maybe the future is not about fixed instruction sets at all [1]. Especially with the rise of high-level hardware synthesis methods, platforms and tools like Kami [2]. Of course, tooling and C compilers for the non-standard ISA are the main obstacles on this way. [1] "Synthesis of Processor Instruction Sets from High-Level ISA Specifications" IEEE Transactions on Computers 2014, 63(6), 1552-1566 [2] https://deepspec.org/entry/Project/Kami https://deepspec.org/entry/Project/Kami

ChuckMcM 8y ago
> Of course, tooling and C compilers for the non-standard ISA are the main obstacles on this way. Yup, see the Itanium architecture for how pernicious this can be to the survival of an architecture.
- snaky 8y ago
  What exactly is the lesson of Itanium is debatable. > Q: I feel like deja vu - at Hot Chips, Intel introduced VLIW-concept Itanium that pushed complexity onto the compiler. I see traces of that here. What are you doing to avoid the Itanium traps? How will you avoid IP from Intel? > A: Itanium was in-order VLIW, hope people will build compiler to get perf. We came from opposite direction - we use dynamic scheduling. We are not VLIW, every node defines sub-graphs and dependent instructions. We designed the compiler first. We build hardware around the compiler, Intel approach the opposite. https://www.anandtech.com/show/13255/hot-chips-2018-tachyum-prodigy-cpu-live-blog https://www.anandtech.com/show/13255/hot-chips-2018-tachyum-... I don't think that's the way of the future anyway, but C is not the only programming language, fortunately. Generating the ISA, and then CPU from specification, the next logical step is to generate highly optimized lower-than-C-level code from high-level specifications too.
- zozbot123 8y ago
  The Itanium folks actually put a lot of effort into tooling and compilers for the architecture, back in the early 2000s - even FLOSS tooling. The problem on that side was that a community never coalesced around Itanium, the way it did around e.g. amd64, or nowadays around riscv.

rurban 8y ago

POWER8 is also open now: https://openpowerfoundation.org/ https://openpowerfoundation.org/ By far the best of those. But IBM, and it's not POWER9.

ksec 8y ago
>and it's not POWER9 Surely POWER9 is open as well? Otherwise how did Raptor ship their POWER9 CPU?
- bildung 8y ago
  Don't they just resell the CPUs? I think the mainboards are designed and produced by them, but not the CPUs.
- rurban 8y ago
  Power9 only has open firmware and ISA, but not the verilog. They just resell it, to Raptor and Google-Rackspace mostly.
floatboth 8y ago
It's not clear what exactly is open about OpenPOWER. Firmware, sure. But is the ISA actually royalty-free?

ineedasername 8y ago

patent stuff seems to be deal with....SGI and Sun are just footnotes Eh... In Sun's case they were acquired by Oracle. I don't think that will have improved the situation. You'll see lots of bad anecdotes about Oracle here & there. You might wonder "huh, is all that really true?" Well, I've dealt with Oracle. I have yet to see something bad about them that didn't jibe with my own experience: I won't give info that might dox me, but there was a fairly visible lawsuit involved.