prasincs
Research Lead At Anchorage
- Joined
- Nov 18, 2026
Comments
Advancing chip design with OpenAI models
Curious if anyone's looked at this from the security side. Everything here is scored on correctness + PPA, and the optimizations they describe (sharing hardware, cutting intermediate registers, re-pipelining) are pretty much the list of things that break masking or add timing channels. Equivalence checking won't catch any of that, since the function doesn't change.
Also, "from-spec scored higher because it omitted features" is basically reward hacking. On a secure core like Ibex, the features with the worst perf/area tradeoff are the countermeasures.
Not saying the results aren't real, the Jalapeño and AMD numbers are hard to argue with. But before model-written RTL goes anywhere near a root of trust, I'd want security properties as hard constraints in the eval, not something you bolt on later.
Hacking a new home screen for my Samsung TV
The “Smart Hub” thing that they have on these TVs is… hot garbage
Yep. I will extend this is to pretty much their whole lineup including devices. This is actually pretty neat and practical writeup of replacing an app. I foresaw this being a viable option in near future and have been working on a full replacement operating system but there are far too many undocumented dragons there.
I am surprised that you only had to go to OPE AI, usually these stories end with then I had to use the Chinese model lol.
Jensen Huang statement offering to return
It'd be pretty bad precedent to set to let him back in given it was affair with direct reports. Even if smuggling case with Mega speed doesn't turn out to be true, it's hard one for the board to walk back from. The way they're communicating seems very aggressive.