Support Matrix#
Hardware#
Unless specified otherwise, NVIDIA NIM for vision language models (VLMs) should, but are not guaranteed to, run on any NVIDIA GPU, provided the GPU has sufficient memory. They can also run on multiple homogeneous NVIDIA GPUs with sufficient aggregate memory and a CUDA compute capability of >= 7.0 (8.0 for bfloat16) unless otherwise specified. For more information, refer to Supported Models.
NVIDIA NIM for VLMs does not support NVIDIA Virtual GPU (vGPU) environments.
For information on the supported operating systems, drivers, and software, refer to the About Get Started page.
Supported Models#
Nemotron 3.5 Content Safety#
Latest supported release tag: 2.0.5-variant
The following section lists the supported configurations for
nvidia/nemotron-3.5-content-safety
(NGC catalog page).
Generic Configuration#
NIM for VLMs offers competitive performance through a custom vLLM backend. Any NVIDIA GPU with sufficient memory should be able to run this model, though this is not guaranteed.
The GPU Memory and Disk Space values are in GB.
GPU |
GPU Memory |
Precision |
# of GPUs |
Disk Space |
|---|---|---|---|---|
B300-SXM6-AC |
288 |
BF16 |
1,2,4,8 |
40 |
B200 |
192 |
BF16 |
1,2,4,8 |
40 |
GB300 |
2x288 |
BF16 |
1,2,4 |
40 |
GB200 |
2x192 |
BF16 |
1,2,4 |
40 |
H200 |
141 |
BF16 |
1,2,4,8 |
40 |
H100-80GB-HBM3 |
80 |
BF16 |
1,2,4,8 |
40 |
L40S |
48 |
BF16 |
1,2,4,8 |
40 |
RTX PRO 6000 Blackwell Server |
96 |
BF16 |
1,2,4,8 |
40 |
GB10 |
128 |
BF16 |
1 |
40 |