Announcement

Collapse
No announcement yet.

Google Moves Ahead With Providing No-Cost Open-Source Silicon Manufacturing From GlobalFoundries

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Google Moves Ahead With Providing No-Cost Open-Source Silicon Manufacturing From GlobalFoundries

    Phoronix: Google Moves Ahead With Providing No-Cost Open-Source Silicon Manufacturing From GlobalFoundries

    Google announced funding for silicon manufacturing for participating open-source projects using the process design kit with GlobalFoundries...

    https://www.phoronix.com/news/Google-GloFo-Sponsored-Si

  • #2
    90nm? 130nm? 180nm? Pentium 3 (Coppermine) was 180nm in 1999.

    Comment


    • #3
      I commend industry for opening up the opportunity.
      But in reality, 6 shuttle runs on what? 180nm doesn't cost google that much.

      I remember low cost shuttle runs on 45nm 10-12 years ago costing around 50k.
      At that time, 45nm a node or two behind bleeding edge.
      Last edited by milkylainen; 13 November 2022, 04:14 AM.

      Comment


      • #4
        Originally posted by xorbe View Post
        Pentium 3 (Coppermine) was 180nm in 1999.
        If it is open hardware, and continuously sourceable, then many of us would still appreciate that.

        Some guys are interested in making solutions, not gaming PCs.
        Last edited by kpedersen; 13 November 2022, 06:43 AM.

        Comment


        • #5
          It may be an elderly process, and maybe not super-suitable for high efficency circuits - but then, it should be true and tried, Google covers the cost and it's not like a community project starts to build something like a Zen 5 CPU all of a sudden. So this sounds like a fair idea.

          I'm more surprised GloFo still has this old machinery around. ;-)
          I tought even automotive chips were done on smaller scale these days.
          Stop TCPA, stupid software patents and corrupt politicians!

          Comment


          • #6
            Originally posted by Adarion View Post
            It may be an elderly process, and maybe not super-suitable for high efficency circuits - but then, it should be true and tried
            Yup, the old whining about true and tried again. Most people on tech sites are like "Why do manufacturers always have to live on the edge with hardware?". But those same people also complain if a manufacturer decides to use something a bit older, but true and tried…

            Comment


            • #7
              Has anyone looked at the projects that have been made with this? I'm curious is anyone has tried to make new versions of old chips like Motorola 68K's or similar?

              Comment


              • #8
                Originally posted by willmore View Post
                Has anyone looked at the projects that have been made with this? I'm curious is anyone has tried to make new versions of old chips like Motorola 68K's or similar?
                There is a FPGA re-implementation of the 68K called 68080:

                http://www.apollo-core.com/

                It would be great to have in silicon.

                Comment


                • #9
                  Originally posted by JPFSanders View Post

                  There is a FPGA re-implementation of the 68K called 68080:

                  http://www.apollo-core.com/

                  It would be great to have in silicon.
                  True, except it isn't open source. They'd have to open source it to get a free run of it. .. but I mean by all means do.

                  I'd acutally like to see the temlib sparc on it... it would be nice to have one of those running at a several hundred Mhz instead of around 50-100 (some of the higher end FPGAs might be able to run it at 200mhz or so).

                  Comment


                  • #10
                    Originally posted by milkylainen View Post
                    I commend industry for opening up the opportunity.
                    But in reality, 6 shuttle runs on what? 180nm doesn't cost google that much.

                    I remember low cost shuttle runs on 45nm 10-12 years ago costing around 50k.
                    At that time, 45nm a node or two behind bleeding edge.
                    Really? You found a way to complain about somebody spending their own money to offer you something that is hard to access, for Free?

                    Comment

                    Working...
                    X