NVIDIA A10
Universal mainstream data center GPU based on Ampere — 24 GB GDDR6, 150W, single-slot. Accelerates graphics, AI inference, virtual workstations, and video in the broadest range of enterprise servers.
Showing all 15 results
-
-
NVIDIA A100 80GB
Ampere Tensor Core GPU with 80 GB HBM2e and 2 TB/s memory bandwidth. The proven workhorse for AI training, HPC, and data analytics with Multi-Instance GPU (MIG) for secure partitioning.
-
NVIDIA A2
Low-profile, low-power edge inference GPU with 16 GB GDDR6 ECC in a 40–60W envelope. Brings NVIDIA AI to space- and power-constrained edge and enterprise servers.
-
NVIDIA AI Enterprise
The end-to-end software platform for production AI. Includes NIM microservices, NeMo, RAPIDS, Triton, TensorRT, and full enterprise support across cloud, data center, and edge.
-
NVIDIA B200
Blackwell architecture flagship — 192GB HBM3e, 8 TB/s bandwidth, up to 18 PFLOPS FP4 sparse per GPU. The next generation of AI training and inference.
-
NVIDIA H100 NVL
Dual-GPU PCIe accelerator with 94 GB HBM3 per card and an NVLink bridge delivering 600 GB/s GPU-to-GPU bandwidth. Purpose-built to supercharge LLM inference in mainstream PCIe servers.
-
NVIDIA H100 SXM
The GPU that launched the AI era — 80GB HBM3, 3.35 TB/s bandwidth, 3,958 TFLOPS FP8, NVLink 900 GB/s in an SXM form factor for HGX and DGX systems.
-
NVIDIA H200 NVL
The first PCIe GPU with HBM3e — 141 GB of memory at 4.8 TB/s. Hopper architecture with NVLink bridge, purpose-built to deploy large language models and generative AI in mainstream enterprise servers.
-
NVIDIA H200 SXM
Hopper architecture with HBM3e — 141GB memory, 4.8 TB/s bandwidth, up to 2x faster LLM inference vs H100. The memory-optimized AI GPU.
-
NVIDIA L4
Ada Lovelace inference accelerator with 24 GB memory in a compact 72W low-profile, single-slot form factor. Delivers universal acceleration for AI inference, video, and graphics at the edge and in the cloud.
-
NVIDIA L40S
The universal Ada Lovelace GPU for the modern data center — 48 GB GDDR6 ECC, 18,176 CUDA cores, and 350W TDP delivering breakthrough performance for LLM inference, training, graphics, and video.
-
NVIDIA RTX PRO 4500 Blackwell Server Edition
Single-slot Blackwell data center GPU with 32 GB GDDR7 ECC in a 165W envelope. Engineered for high-density edge, telco, and VDI deployments where power and density matter most.
-
NVIDIA RTX PRO 6000 Blackwell Server Edition
The universal data center GPU powered by Blackwell — 24,064 CUDA cores, 96 GB GDDR7 ECC, and 5th-gen Tensor Cores for AI, graphics, digital twins, and VDI all on a single server-grade card.
-
NVIDIA Rubin CPX
A new class of NVIDIA GPU purpose-built for massive-context inference. Rubin CPX accelerates million-token reasoning, video understanding, and long-context agents with disaggregated inference design.
-
NVIDIA Vera Rubin NVL72
Rack-scale AI factory built on the Rubin platform: 72 Rubin GPUs, 36 Vera CPUs, NVLink 6 fabric, and up to 3.6 EFLOPS of NVFP4 inference. Engineered for agentic AI and trillion-parameter reasoning models.