3 ms·
It's extremely easy to achieve. It just costs money and takes effort.
by nickpsecurity 11y ago
It's extremely easy to achieve. It just costs money and takes effort.
- pjc50 11y ago"Going to the moon is easy to achieve, it just costs money and takes effort." I don't think it is easy to achieve verifiability to be above nitpicking of the kind we're seeing here. Assume we start with open RTL in some HDL. You then have to run it through layout and get someone with a Formality license to check it's equivalent (trusting that particular piece of software). You then take your GDSII to a fab. If you want to check that they've not inserted something, you have to resort to sampling production and sending it to the specialists with microscopes to check it - but then you have to trust their reports. And none of this is provable to doubting strangers on the internet. It's always possible for someone to assert that the whole group of people responsible is a clever fake by the CIA. You can't check it on your desk the same way you might be able to check software with reproducible builds (not that more than a few people do this anyway).
- nickpsecurity 11y ago"You can't check it on your desk the same way you might be able to check software with reproducible builds (not that more than a few people do this anyway)." Everything you just said applies to software as well because you have to trust what your machine shows you. And it might be subverted. "I don't think it is easy to achieve verifiability to be above nitpicking of the kind we're seeing here." The most basic verifiability for open HW is to use components that have documented, open HDL with rest of what was produced open for 3rd party verification. Gaisler already did it for a CPU and whole IP library. Rocket RISC-V team did their part. Magic-1 guy did it with wire-wrap on TTL chips. None of these people had $10+ million. Gaisler shows it was just a matter of getting money together to do the development cost-effectively, licensing to commercial users to sustain it, ensuring any I.P. could be re-synthesized for new use cases, and simply offering source for re-synthesis or just verification. Now, we can go much further than that and tools like Qflow OSS Synthesis are a step in that direction. However, most likely approach given EDA complexity is doing it with industry standard tools with third parties checking those results. Might even do it totally in high-end tools but in a way that end result can be verified to large degree with open tools. Many ways to do that but all need work. Good news is that there's no Snowden leaks, etc suggesting EDA tools are subverted. That helps. :) So, you're creating one strawman after another to justify either (a) not trying or (b) calling something open that's not. The minimal requirement has been done repeatedly in other projects by other people. It takes no extra money on top of a normal HW development. It can even be done both proprietary and open source as Gaisler illustrated for some of his stuff. The result merely has to be open and done in a way that 3rd parties can verify. Optionally, that they can integrate and put on fab of choosing if they desire. That's minimum for open-source hardware. A project does the minimum, I call it open-source hardware. If it is not doing it, I'm calling it something else. I'm calling a project that heavily depends on black boxes from DOD-connected and scheming companies as opposite of open in operation as is possible. If the board is open, I'll call it an open-board or open-PCB design. Novena seems to be an open-board system with OSS software. It's not open HW.