4 ms·
Required or not, it is desirable.
by snvzz 5y ago
Required or not, it is desirable.
- krasin 5y agoMy guess would be that the first prototype that they made was too buggy to publish. The promise is that they will share the final design, and I am fine with that. (I am not affiliated with the project, but backed it)
- krasin 5y agoUpdate: marcan_42 apparently found it: https://efabless.com/projects/33 https://efabless.com/projects/33
- kop316 5y agoI don't think they are the same: Look at the picture in it versus in the link: https://groupgets-files.s3.amazonaws.com/Efabless/open-source-logo-on-caravel-min.jpg https://groupgets-files.s3.amazonaws.com/Efabless/open-sourc... They are different. Also, if you go to https://vlsicad.ucsd.edu/Publications/Conferences/383/c383.pdf https://vlsicad.ucsd.edu/Publications/Conferences/383/c383.p... Figure 6: You can find the image from the link (top left), and the on ehtat you linked (bottom left), so they are two different projects.
- kop316 5y agoAccording to this website: MPW-1 had a hold time violation in it's clock tree, so the RISC-V is unusable. https://www.zerotoasiccourse.com/post/mpw1-bringup/ https://www.zerotoasiccourse.com/post/mpw1-bringup/
- makomk 5y agoYikes. If I'm understanding that correctly every chip that came back from MPW-1 was more or less unusable - they were all required to gate access to I/Os behind the same management processor provided by the company organising the run, and because that management processor was broken they couldn't configure the I/Os to route through to their own designs.
- kop316 5y agoYeah that is my understanding too. You REALLY don't want a hold time violation, as you have to go in and edit the chip circuitry to fix that.
- commandpaul 5y agoLooks like this has been put together by Efabless themselves. I dont think the rtl would be too different from OpenFPGA(https://github.com/lnis-uofu/OpenFPGA https://github.com/lnis-uofu/OpenFPGA) married to the caravel f/w.