I’m happy to announce that starting today, I’m officially part of the NVIDIA DGX Spark team and can focus on making the DGX Spark experience better for everyone on a full-time basis. There will be a short transition period where I will not be as active as before, but I’ll be working more closely with other NVIDIA folks who will be here to support the community.
So, what does this mean for spark-vllm-docker and llama-benchy?
For the time being, both projects will continue as is. I will keep maintaining and enhancing them and will work more closely with the NVIDIA, vLLM and Flashinfer teams on bringing performance improvements and bleeding-edge functionality to the builds.
Joining NVIDIA will also unlock access to more resources, including hardware, so we will be able to test more hardware configurations and models going forward.
I will follow up with more updates as I settle into my new role and we work on the overall strategy.
I will also be switching to my work NVIDIA account, @eugr_nv, but I will keep monitoring this one as well.
@eugr and @eugr_nv You can always count on me to make your journey at NVIDIA very successful. Great news for the whole community. Couldn’t imagine a better person to be in the Spark team. Congrats and success.
Seeing this made my day. Your efforts have been genuinely instrumental for us, so glad to see that the NVIDIA team saw the value you provided and your clear passion for the field. Congratulations, we’re all lucky to have you doing this professionally now
Wow! I’m excited for you. I’m excited for the rest of us, too. We can rest assured that someone from nVidia cares as much as we do about getting that odd little box to do all that it is capable of.
Move both obvious and genius on nVidia management’s part. Corporate organizations are so seldom accused of executing either. This improves our perception of how the company regards DGX customers.
