Brand: NVIDIA | Category: GPUs
SKU: 900-2G133-0000-100 | Part #: 900-2G133-0000-100 | MPN: 900-2G133-0000-100
Contact for Pricing — Request a Quote
Enterprise AI inference and real-time visualization workloads demand GPU architecture that balances memory capacity with sustained throughput—making the NVIDIA A10 24GB GDDR6 PCIe Gen4 Enterprise GPU an ideal foundation for data centers running concurrent inference tasks and high-resolution graphics pipelines. Built on the NVIDIA Ampere architecture, this GPU delivers 9,216 CUDA cores paired with 288 third-generation Tensor Cores and 72 second-generation RT Cores, engineered to accelerate mixed-precision inference, machine learning model serving, and professional visualization at enterprise scale.
The A10 offers substantial memory bandwidth—600 GB/s across a 384-bit memory interface—supporting batch inference and multi-stream workloads without memory contention. With 24 GB of GDDR6 memory, infrastructure teams can deploy larger models or process higher-throughput concurrent requests. The GPU's PCIe Gen4 x16 host interface ensures rapid data movement between compute nodes, while passive cooling (requiring an airflow-optimized server chassis) and a 150 W thermal design power footprint make it suitable for dense multi-GPU cluster deployments. Four DisplayPort 1.4 outputs enable 8K display output at up to 7680 x 4320 resolution, supporting remote visualization and digital content creation pipelines. IT procurement teams and AI infrastructure architects will recognize the dual-slot, full-height full-length form factor as standard for enterprise GPU acceleration.
For detailed technical datasheets, performance benchmarks, and deployment guidance, contact Omnixon Global to request a quotation.
| Brand | NVIDIA |
| Category | GPUs |
| SKU | 900-2G133-0000-100 |
| Part Number | 900-2G133-0000-100 |
| Condition | New |
| Manufacturer Part Number | 900-2G133-0000-100 |
| GPU Architecture | NVIDIA Ampere |
| CUDA Cores | 9,216 |
| Tensor Cores | 288 (3rd Generation) |
| RT Cores | 72 (2nd Generation) |
| GPU Memory | 24 GB GDDR6 |
| Memory Bandwidth | 600 GB/s |
| Memory Interface | 384-bit |
| Host Interface | PCIe Gen4 x16 |
| FP32 Performance | 31.2 TFLOPS |
| TF32 Tensor Core Performance | 62.5 TFLOPS (sparsity: 125 TFLOPS) |
| FP16 Tensor Core Performance | 125 TFLOPS (sparsity: 250 TFLOPS) |
| INT8 Tensor Core Performance | 250 TOPS (sparsity: 500 TOPS) |
| Thermal Design Power (TDP) | 150 W |
| System Interface Power | 250 W (total board power) |
| Form Factor | Dual-slot, full-height full-length (FHFL) |
| Cooling | Passive (requires airflow-optimized server chassis) |
| Display Outputs | 4x DisplayPort 1.4 |
| Max Display Resolution | 7680 x 4320 (8K) |
| vGPU Software Support | Yes (NVIDIA Virtual PC, Virtual Apps, RTX vWS, vCS) |
| ECC Memory | Yes |
| NVLink Support | Not supported |
Reference servers include Dell PowerEdge XE9680 / XE9712, HPE Cray XD670, Lenovo ThinkSystem SR685a / SR675 V3, Supermicro AS-A21GE / SYS-821GE, Gigabyte G593 / G894, ASUS ESC. Share your target platform in the RFQ and we will confirm chassis-to-GPU compatibility and recommended NIC pairing.
Highly model-dependent. L40S / RTX-class: typically 3-6 weeks. H100/H200/B200 in SXM form factor: 12-16 weeks for whole-platform allocations. We quote genuine-channel ETAs only — no grey-market promises.
Yes — Omnixon stocks the full NVIDIA networking lineup (Quantum-2 / Quantum-X InfiniBand, Spectrum-X Ethernet, ConnectX NICs, BlueField DPUs) so we can quote a complete training-cluster BOM, not just the GPUs.
Yes. We hold genuine-channels for NVIDIA AI Enterprise software subscriptions. Add it to your RFQ and we quote node-aligned licensing along with the hardware.