Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Nvidia DGX Spark: great hardware, early days for the ecosystem

simonwillison.net

Post reply on HN
Print view View original

1–10 of 116 posts

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#4

Post by p_l »

ur-whale wrote: post #3

As is usual for NVidia: great hardware, an effing nightmare figuring out how to setup the pile of crap they call software.

And yet CUDA has looked way better than ATi/AMD offerings in the same area despite ATi/AMD technically being first to deliver GPGPU (major difference is that CUDA arrived year later but supported everything from G80 up, and nicely evolved, while AMD managed to have multiple platforms with patchy support and total rewrites in between)

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#5

Post by kanwisher »

ur-whale wrote: post #3

As is usual for NVidia: great hardware, an effing nightmare figuring out how to setup the pile of crap they call software.

If you think their software is bad try using any other vendor , makes nvidia looks amazing. Apple is only one close

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#6

Post by simonw »

It's notable how much easier it is to get things working now that the embargo has lifted and other projects have shared their integrations.

I'm running VLLM on it now and it was as simple as:

  docker run --gpus all -it --rm \
    --ipc=host --ulimit memlock=-1 \
    --ulimit stack=67108864 \
    nvcr.io/nvidia/vllm:25.09-py3
(That recipe from https://catalog.ngc.nvidia.com/orgs/nvidia/containers/vllm?v... )

And then in the Docker container:

  vllm serve &
  vllm chat
The default model it loads is Qwen/Qwen3-0.6B, which is tiny and fast to load.

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#7

Post by jasonjmcghee »

ur-whale wrote: post #3

As is usual for NVidia: great hardware, an effing nightmare figuring out how to setup the pile of crap they call software.

Except the performance people are seeing is way below expectations. It seems to be slower than an M4. Which kind of defeats the purpose. It was advertised as 1 Petaflop on your desk.

But maybe this will change? Software issues somehow?

It also runs CUDA, which is useful

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#8

Post by enoch2090 »

kanwisher wrote: post #5
ur-whale wrote: post #3

As is usual for NVidia: great hardware, an effing nightmare figuring out how to setup the pile of crap they call software.

If you think their software is bad try using any other vendor , makes nvidia looks amazing. Apple is only one close

Although a bit off the GPU topic, I think Apple's Rosetta is the smoothest binary transition I've ever used.

Re: Nvidia DGX Spark: great hardware, early days for the ecosystem

#10

Post by airstrike »

jasonjmcghee wrote: post #7
ur-whale wrote: post #3

As is usual for NVidia: great hardware, an effing nightmare figuring out how to setup the pile of crap they call software.

Except the performance people are seeing is way below expectations. It seems to be slower than an M4. Which kind of defeats the purpose. It was advertised as 1 Petaflop on your desk. But maybe this will change? Software issues somehow? It also runs CUDA, which is useful

it fits bigger models and you can stack them.

plus apparently some of the early benchmarks were made with ollama and should be disregarded

Post reply on HN