Today, NVIDIA has announced that with NVFP4 support, the DGX Spark delivers up to a 2.5x boost in the Qwen 235B model (two DGX Sparks paired). Boost from what, running on a CPU? Early adopters have been patiently waiting for proper software support. It makes it even worse to read this marketing s…

Yup – hadn’t set the TIKTOKEN source. Duh. Anyways, last update on this with GPT-OSS-120B @eugr vllm | model | test | t/s | ttfr (ms) | est_ppt (ms) | e2e_ttft (ms) | |:--------------------|----------------:|----------------:|-----------------:|…

I’m unsure if you’re strictly talking about the recent announcement/live stream, but Nvidia hosts the model nvidia/Qwen3-14B-FP4 which is NVFP4. It’s much faster than vanilla Qwen3-14B.

Is it faster than Qwen3-14B-AWQ though? Of course 4-bit quant will be faster than full fp16 model.

Yeah, nice prefill speeds, but generation speeds are bad even in vLLM, should be in mid-high 50’s on low context…

Did you run their Docker container or built it from the source?

Their container. I didn’t feel like putting myself through whatever torture of ENVS were required to get it built 🤣

right.- that’s the misleading part I think. Being 2.5x faster than a full weight model with 4 bit quant is a ridiculously low bar. Particularly with sparks limitations. The real comparisons are other quants and inference providers — and it loses there.

I tried the NVFP4 quant of Flux2 dev in ComfyUI and it was twice as slow as the fp8 version, which was disappointing. If anyone is a ComfyUI guru maybe you could critique my launch parameters: python3 main.py –listen 0.0.0.0 –disable-mmap –use-sage-attention –supports-fp8-compute –gpu-on…

And here are the results of the German Jury. ### nvcr.io/nvidia/tensorrt-llm/release:1.2.0rc7 ### latest llama-benchy --base-url http://localhost:8355/v1 --model nvidia/Llama-3.1-8B-Instruct-FP4 --depth 0 4096 8192 16384 32768 --latency-mode generation | model | …

NVIDIA folks -- where is this promised nvfp4 speedup?

Accelerated Computing DGX Spark / GB10 User Forum DGX Spark / GB10

ferran.duarri March 26, 2026, 2:59pm 27

I accept an nvidia dgx spark as a gift/developer machine to try tune it and make custom software for it, my actual project is not meant for nvidia dgx spark (which already has an integrated memory bus) >> https://forums.developer.nvidia.com/t/nvidia-greenboost-kernel-modules-opensourced/

Topic		Replies	Views
We unlocked NVFP4 on the DGX Spark: 20% faster than AWQ! DGX Spark / GB10	145	8341	March 28, 2026
New bleeding-edge vLLM Docker Image: avarok/vllm-nvfp4-gb10-sm120 DGX Spark / GB10 Projects	35	3144	December 31, 2025
Help: Running NVFP4 model on 2x DGX Spark with vLLM + Ray (multi-node) DGX Spark / GB10 mistral-large	18	2548	December 25, 2025
NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 DGX Spark / GB10 nemotron	89	9678	March 31, 2026
DGX Spark performance DGX Spark / GB10	50	5526	February 27, 2026
PSA: State of FP4/NVFP4 Support for DGX Spark in VLLM DGX Spark / GB10	234	12478	May 15, 2026
Two-Spark cluster with vLLM using tensor-parallel-size 2 causes one node to drop while the other's GPU goes 100% forever DGX Spark / GB10	36	1608	February 13, 2026
DGX Spark + Qwen3-Next-80B: Proven Performance, But Missing Clear Path to NIM, TensorRT-LLM & Web UIs DGX Spark / GB10 cuda , nim , llama	16	4519	March 6, 2026
Benchmark Report: Qwen3.6-35B-A3B-NVFP4 on NVIDIA DGX Spark, Jetson Thor, Blackwell 6000 Pro DGX Spark / GB10 Projects	10	1492	June 2, 2026
Question on Inference Performance Results of Qwen3 235B A22B on 2× DGX Spark DGX Spark / GB10 cuda	5	760	December 19, 2025

NVIDIA folks -- where is this promised nvfp4 speedup?

Related topics