Brand: NVIDIA | Category: GPUs
SKU: 900-2G192-0000-000 | Part #: 900-2G192-0000-000 | MPN: 900-2G192-0000-000
Contact for Pricing — Request a Quote
The NVIDIA L4 is a purpose-built GPU accelerator designed for edge AI inference, video analytics, and real-time processing in space-constrained enterprise environments. With 24GB of GDDR6 memory and PCIe Gen4 connectivity, it delivers deterministic, low-latency inference performance across a broad range of deep learning frameworks and video codec transcoding workloads. The low-profile form factor enables deployment in compact edge servers, telecommunications infrastructure, and embedded systems where thermal and physical constraints are critical.
| Manufacturer | NVIDIA |
| Model | L4 |
| Part Number | 900-2G192-0000-000 |
| GPU Memory | 24 GB GDDR6 |
| Memory Bandwidth | 432 GB/s |
| Memory Interface | 192-bit |
| PCIe Interface | PCIe Gen4 x12 |
| Max Power Consumption | 70 W |
| Form Factor | Low-profile, Single-slot |
| CUDA Cores | 5120 |
| Tensor Float 32 (TF32) Performance | 120 TFLOPS |
| BFLOAT16 Performance | 240 TFLOPS |
| INT8 Performance | 240 TFLOPS |
| Max Thermal Design Power | 70 W |
| Supported Frameworks | TensorFlow, PyTorch, ONNX Runtime, TensorRT |
| Video Codec Support | H.264, H.265, VP9 (encode/decode) |
| Maximum Concurrent Video Streams | Up to 32 concurrent H.265 1080p streams |
| Launch Date | Q1 2023 |
Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.
| Brand | NVIDIA |
| Category | GPUs |
| SKU | 900-2G192-0000-000 |
| Part Number | 900-2G192-0000-000 |
| Condition | New |
| Model | L4 |
| GPU Memory | 24 GB GDDR6 |
| Memory Bandwidth | 432 GB/s |
| Memory Interface | 192-bit |
| PCIe Interface | PCIe Gen4 x12 |
| Max Power Consumption | 70 W |
| Form Factor | Low-profile, Single-slot |
| CUDA Cores | 5120 |
| Tensor Float 32 (TF32) Performance | 120 TFLOPS |
| BFLOAT16 Performance | 240 TFLOPS |
| INT8 Performance | 240 TFLOPS |
| Max Thermal Design Power | 70 W |
| Supported Frameworks | TensorFlow, PyTorch, ONNX Runtime, TensorRT |
| Video Codec Support | H.264, H.265, VP9 (encode/decode) |
| Maximum Concurrent Video Streams | Up to 32 concurrent H.265 1080p streams |
| Launch Date | Q1 2023 |