Note: Due to market demand, availability and pricing for memory and storage components may vary. Please contact our team to confirm current inventory and pricing
Nvidia 900-2H400-0010-000 Tesla P100 12GB HBM2 PCIe x16 3.0 GPU

* NOTE: Image may not exactly match the product
900-2H400-0010-000
2% Discount on Checkout
AED 12526.50 Incl. Vat
Get an extra 2% off on checkout

More Graphics Cards
From NVIDIA
Looking for a dependable GPU to boost your AI, deep learning or scientific computing setup? The Nvidia 900-2H400-0010-000 Tesla P100 12GB is a data center accelerator built on the Pascal architecture. Its 12GB of HBM2 memory and 3,584 CUDA cores let you process large datasets faster than a CPU-only server.
This card fits servers and workstations with a PCIe 3.0 x16 slot, making it a practical upgrade for businesses, labs and developers in the UAE. It is built for compute work rather than display output, so it suits training models, running simulations and accelerating analytics at a sensible cost of ownership.
Yes. With 3,584 CUDA cores, 12GB of HBM2 memory and FP16 support, it handles training and inference for many small to mid-sized AI models. It is a good fit for development, research and testing environments. Very large modern models may need more memory than 12GB provides.
No, it is a compute-only accelerator with no display connectors. You will need onboard graphics or a separate GPU to connect a monitor. The P100 then runs in the background, handling AI, rendering or scientific workloads.
It supports the NVIDIA CUDA platform and works with popular frameworks such as TensorFlow and PyTorch, plus many HPC applications. You need to install the correct NVIDIA data center driver for your operating system. Check framework version support before deployment, as newer releases may phase out older Pascal GPUs.