6 ms·

> Probably many: the processor has to be made by millions (per model, not in total) to become reasonably priced, and it will still be slower and/or hotter than

by progman 10y ago

> Probably many: the processor has to be made by millions (per model, not in total) to become reasonably priced, and it will still be slower and/or hotter than an offering from Intel Why should R5 compete with the big players? It doesn't need to support the bloated OSes. It would be sufficient if Linux runs with acceptable performance. If Linux is able to run on a 12 MHz 68k [1], and SiFive already runs with 320 MHz [2] then I think we don't have to wait too much to get a suitable cpu, since it could also be a "weak" massive multicore, or we could stick multiple cpus together in a cluster. By the way, Fabrice Bellard has already published a Linux/R5 emulator [3]. As for me, I am really excited about the R5 movement! [1] http://www.bigmessowires.com/68-katy/ http://www.bigmessowires.com/68-katy/ [2] https://www.crowdsupply.com/sifive/hifive1 https://www.crowdsupply.com/sifive/hifive1 [3] http://bellard.org/riscvemu/ http://bellard.org/riscvemu/

zhemao 10y ago
The distribution of Linux running on a 68k microcontroller is very different from the distribution running on a laptop. Desktop operating systems have to support a much wider set of software, often with quite intensive costs. How well do you think a web browser would run on a 68K?
- nine_k 10y ago
  There's a lot of embedded hardware where running Linux makes sense but running a browser does not. This market may be even larger than the desktop market. OTOH this rules out laptops that OP wanted. OT3rdH ARM started with devices like these in mind, but went all the way to pretty powerful, desktop-comparable machines like current smartphones.
  - progman 10y ago
    Just imagine that the Raspberry Pi Foundation could be interested to switch from ARM to R5. They could actually design and produce their own R5.
    - nine_k 10y ago
      I can imagine that government agencies of various countries would be quite interested in a completely open, inspectable, patent-unencumbered, hardware design. Gives you a bit of certainty that your key infrastructure won't be remotely disrupted by a hidden silicon circuit.
      - nickik 10y ago
        India elected RISC-V as the new standard and they have multible projects around it. Open ISA and open hardware should be a major security concern for most countries.
    - zhemao 10y ago
      In fact, some people who were involved in the Raspberry Pi project are now working on exactly that, a RISC-V-based single-board computer. http://www.lowrisc.org/ http://www.lowrisc.org/
- progman 10y ago
  > How well do you think a web browser would run on a 68K? It depends. If you want something like Opera or Firefox it won't run at all. If you are satisfied with basic functionality then you can use browsers like w3m which work well on low-powered devices. I think some developers will find a good compromise between performance and functionality. And by the way, the very first R5 already runs with 320 MHz :-)
  - zhemao 10y ago
    I assume you are referring to SiFive's HiFive board. It's certainly a great accomplishment. I'm actually a Ph.D. student in the UC Berkeley lab that produced the RISC-V ISA and interned at SiFive over the summer. I've worked on the RocketChip SoC generator they used to produce the RISC-V processor on that board. So I definitely want them to succeed. But there is still a very long way to go. Clock speed is not the only determinant of processor performance. Even low end laptops like the ARM-based Chromebook have out-of-order processors. The Rocket CPU used in the HiFive board is an in-order machine. There is an out-of-order implementation called BOOM, but it is not quite as mature. Whereas Rocket has been taped out and validated many times in our Berkeley test chips and now in SiFive's commercial silicon, BOOM has not yet been taped out. And I'll be the first to admit that RocketChip is not nearly as reliable as you would like for a consumer product. SiFive's done a lot of work to improve the system, but there are probably still outstanding bugs in the memory system.
- chriswarbo 10y ago
  > How well do you think a web browser would run on a 68K? They run fine; I find Voyager 3 to be a bit bloated, but IBrowse runs quite well. As others have mentioned, text-based browsers like w3m will probably be even smoother (although I usually run w3m via emacs-w3m, which would probably be much slower than using a normal terminal interface).

by progman 10y ago

zhemao 10y ago

The distribution of Linux running on a 68k microcontroller is very different from the distribution running on a laptop. Desktop operating systems have to support a much wider set of software, often with quite intensive costs. How well do you think a web browser would run on a 68K?

nine_k 10y ago
There's a lot of embedded hardware where running Linux makes sense but running a browser does not. This market may be even larger than the desktop market. OTOH this rules out laptops that OP wanted. OT3rdH ARM started with devices like these in mind, but went all the way to pretty powerful, desktop-comparable machines like current smartphones.
- progman 10y ago
  Just imagine that the Raspberry Pi Foundation could be interested to switch from ARM to R5. They could actually design and produce their own R5.
  - nine_k 10y ago
    I can imagine that government agencies of various countries would be quite interested in a completely open, inspectable, patent-unencumbered, hardware design. Gives you a bit of certainty that your key infrastructure won't be remotely disrupted by a hidden silicon circuit.
    - nickik 10y ago
      India elected RISC-V as the new standard and they have multible projects around it. Open ISA and open hardware should be a major security concern for most countries.
  - zhemao 10y ago
    In fact, some people who were involved in the Raspberry Pi project are now working on exactly that, a RISC-V-based single-board computer. http://www.lowrisc.org/ http://www.lowrisc.org/
progman 10y ago
> How well do you think a web browser would run on a 68K? It depends. If you want something like Opera or Firefox it won't run at all. If you are satisfied with basic functionality then you can use browsers like w3m which work well on low-powered devices. I think some developers will find a good compromise between performance and functionality. And by the way, the very first R5 already runs with 320 MHz :-)
- zhemao 10y ago
  I assume you are referring to SiFive's HiFive board. It's certainly a great accomplishment. I'm actually a Ph.D. student in the UC Berkeley lab that produced the RISC-V ISA and interned at SiFive over the summer. I've worked on the RocketChip SoC generator they used to produce the RISC-V processor on that board. So I definitely want them to succeed. But there is still a very long way to go. Clock speed is not the only determinant of processor performance. Even low end laptops like the ARM-based Chromebook have out-of-order processors. The Rocket CPU used in the HiFive board is an in-order machine. There is an out-of-order implementation called BOOM, but it is not quite as mature. Whereas Rocket has been taped out and validated many times in our Berkeley test chips and now in SiFive's commercial silicon, BOOM has not yet been taped out. And I'll be the first to admit that RocketChip is not nearly as reliable as you would like for a consumer product. SiFive's done a lot of work to improve the system, but there are probably still outstanding bugs in the memory system.
chriswarbo 10y ago
> How well do you think a web browser would run on a 68K? They run fine; I find Voyager 3 to be a bit bloated, but IBrowse runs quite well. As others have mentioned, text-based browsers like w3m will probably be even smoother (although I usually run w3m via emacs-w3m, which would probably be much slower than using a normal terminal interface).