The NVIDIA® GeForce RTX™ 5090 is the most powerful GeForce GPU ever made, bringing game-changing capabilities to gamers and creators. Tackle the most advanced models and most challenging creative workloads with unprecedented AI horsepower. Game with full ray tracing and the lowest latency. The GeForce RTX 5090 is powered by the NVIDIA® Blackwell architecture and equipped with 32GB of super-fast GDDR7 memory, so you can do it all.
Features
NVIDIA Blackwell Architecture
32GB GDDR7 Memory
3x Display Port 1x HDMI
21,760 CUDA® Cores
600 W Thermal Design Power
Warranty 3 years manufacturer included
Model: TUF-RTX5090-32G-GAMING
Video Memory: 32GB GDDR7
CUDA Core: 21760
Memory Speed: 28 Gbps
Memory Interface: 512-bit
Resolution: Resolution 7680 x 4320
✔ FP64: 34 TFLOPS
✔ FP64 Tensor Core: 67 TFLOPS
✔ FP32: 67 TFLOPS
✔ TF32 Tensor Core²: 989 TFLOPS
✔ Architecture: Blackwell
✔ BFLOAT16 Tensor Core²: 1,979 TFLOPS
✔ FP16 Tensor Core²: 1,979 TFLOPS
✔ FP8 Tensor Core²: 3,958 TFLOPS
✔ INT8 Tensor Core²: 3,958 TFLOPS
✔ GPU Memory: 141GB
✔ GPU Memory Bandwidth: 4.8TB/s
✔ Decoders: 7 NVDEC, 7 JPEG
✔ Confidential Computing: Supported
✔ Max Thermal Design Power (TDP): Up to 600W (configurable)
✔ Multi-Instance GPUs: Up to 7 MIGs @16.5GB each
✔ Form Factor: PCIe
✔ Interconnect: 2- or 4-way NVIDIA NVLink bridge: 900GB/s, PCIe Gen5: 128GB/s
✔ Server Options: NVIDIA MGX™ H200 NVL partner and NVIDIA-Certified Systems with up to 8 GPUs
✔ NVIDIA AI Enterprise: Add-on
✔ Warranty: 3 years manufacturer parts or replace
Ships in 2-3 weeks from payment. All sales final. No returns or cancellations. For bulk inquiries, consult a live chat agent or call our toll-free number.
✔ Product SKU: ARS-111GL-NHR
✔ Motherboard: Super G1SMH-G
✔ Processor: NVIDIA 72-core Grace CPU on GH200 Grace Hopper™ Superchip
✔ Memory:
✔ Chipset: System on Chip
✔ I/O Ports:
✔ BIOS Type: AMI 64MB SPI Flash EEPROM
✔ Power Management:
✔ Power Supply: 2x 2000W Redundant Titanium Level (96%)
✔ Available: Used Condition
The Stock is 3Months Used, and available in USA for Immediate Pickup. All sales final. No returns or cancellations. For bulk inquiries, consult a live chat agent or call our toll-free number.
The H100 NVL has a full 6144-bit memory interface (1024-bit for each HBM3 stack) and memory speed up to 5.1 Gbps. This means that the maximum throughput is 7.8GB/s, more than twice as much as the H100 SXM. Large Language Models require large buffers and higher bandwidth will certainly have an impact as well.
NVIDIA H100 NVL for Large Language Model Deployment is ideal for deploying massive LLMs like ChatGPT at scale. The new H100 NVL with 96GB of memory with Transformer Engine acceleration delivers up to 12x faster inference performance at GPT-3 compared to the prior generation A100 at data center scale.
Unprecedented performance, scalability, and security for every data center. Designed for deep learning and special workloads.
Part Number: 900-21010-0300-030
The SXM4 (NVLINK native soldered onto carrier boards) version of the cards are available upon request only, and are attached permanently to their respective motherboards via a complete system only, with longer lead times.
The NVIDIA RTX™ 5000 Ada Generation GPU, powered by the NVIDIA Ada Lovelace architecture, unlocks breakthroughs in generative AI and delivers the performance required to meet the challenges of today’s professional workflows. With 100 third-generation RT Cores, 400 fourth-generation Tensor Cores, 12,800 CUDA® cores, and 32GB of graphics memory, the RTX 5000 excels in rendering, AI, graphics, and compute workload performance.
Supported Technologies
AMD Infinity Guard , AMD Infinity Architecture
Workload Affinity
Analytics , App dev|test , Content mgmt , HPC , Media streaming , Networking|NFV , Security , VDI , VM Density , Web Serving , CDC
