for dav1d there is https://github.com/memorysafety/rav1d although it reuses the dav1d assembly and performance is typically slower by a single-digit percentage.
Vera Rubin will have Groq chips focused on fast inference so it points toward a trend. Also, with energy needs so high, why not reach for every feasible optimization?
> In 1997, ASML began studying a shift to using extreme ultraviolet and in 1999 joined a consortium, including Intel and two other U.S. chipmakers, in order to exploit fundamental research conducted by the US Department of Energy. Because the Cooperative Research and Development Agreement (CRADA) it operates under is funded by the US government, licensing must be approved by Congress.
As far as I understand, people using this site to contact their elected officials were instrumental in making lawmakers back down from ChatControl v2.0. Hoping the same will be true this time around.
My understanding is that they fell behind on offering the latest gen Nvidia hardware (Blackwell/Blackwell Ultra) due to their focus on internally developed ASICs (Trainium/Inferentia gen 2).
If you're interested in low-level, resource-constrained and privileged execution environments, consider joining our team of compiler, OS, and Rust enthusiasts.
At Immunant, we spend much time staring at weird build errors, ISA or ABI documentation, or digging through obscure build systems or C code written 20 years ago. In any case, we strive to make the systems we're working on more secure, more efficient, and have more Rust in them (whenever we get a chance).
We also maintain and enhance the fully open source c2rust translation framework.
If you have recent, professional experience with C/C++ and Rust and know about operating systems, hypervisor internals, or compilers, we'd love to hear from you. Candidates with only superficial knowledge in all of these areas are not encouraged to apply. Please note that we can only consider candidates resident in the US.
We offer a remote-friendly, collaborative work environment with a high degree of autonomy and competitive benefits.
I didn’t see any mention of EU’s last attempt to build a sovereign cloud: Gaia-X. I think that didn’t really go anywhere due to disagreements over technical direction. Much like Quaero.
Don’t get me wrong, I hope it goes better this time and the author acknowledges the many pitfalls of such a complex undertaking.
well put! making sure that the new, memory safe code interoperates with the old code and is equally well supported by tooling takes an incredible amount of work.