shopify analytics ecommerce

NVIDIA Tesla P100 - Front

More Views

NVIDIA Tesla P100 Passive GPU

NVIDIA Tesla P100 Passive - 12 or 16GB HBM2 PCIe GPU Module - 9.3 TFLOP SP/4.7TFLOP DP

Price as configured: £0.00 Excl. Tax


Price as configured (excl. VAT): £0.00


* Required Fields

Price as configured (excl. VAT): £0.00



Most modern datacentres rely heavily of on interconnected computing nodes to meet the demands of High-Performance Computing (HPC) workloads. NVIDIA Tesla P100 is the most advanced GPU ever built. This Professional Graphical Processing Unit is design for some of the world’s critical challenges that demand tremendous amount computing power to be implemented. Some of these challenges include Artificial Intelligence / Deep Learning for self-driving cars, future climate prediction machines, machines for developing new drugs cancer treatment.

The NVIDIA Tesla P100 adopts the new NVIDIA Pascal GPU architecture that is designed to offer more performance than 100 computing nodes based on Intel Xeon Processors. The high-performance GPU ensure high throughput while saving a lot of money. There are now over 400 HPC applications that can take advantage of GPU acceleration such as deep learning frameworks, and many organizations are now developing their application to widen support. Click here to see a full list of compatible applications.

NVIDIA Tesla P100 Features and Benefits

The Tesla P100 is well designed for both hardware and software compatibility delivering high feature innovations at every level. Therefore, it combines various modern high-end technologies to deliver a dramatic jump in performance and come up with world fastest computing node.

NVIDIA Tesla P100 offers the following benefits:

- Adopt the Pascal Architecture to achieve exponential performance:
With the NVIDIA Pascal architecture design, Tesla P100 can deliver absolute high performance for hyper scale performance and HPC. It features over 21 TeraFLOPS with FP16 performance and is optimized to drive existing and new possible deep learning applications. The Pascal architecture is also optimised to deliver more than 5 and 10 TeraFLOPS with double and single precision performance for HPC.

- It includes Page Migration Engine that allows Simple Programming:
NVIDIA Tesla includes a parallel computing platform and programming models called CUDA that allow simple programming. This helps to ensure that developers focus on tuning the computing performance and managing data movement rather that concentrating in programming.

NVIDIA TESLA P100 for Mixed-Workload HPC:
This type of NVIDIA TESLA P100 is designed to meet the demands of mixed-workload HPC data especially in datacentres, which help to realize a dramatic improvement in throughput at low cost. For instance, a single GPU-accelerate node can be powered by four Tesla P100s interconnected using the PCLe to achieve the performance of 32 commodity CPU nodes. This makes it easy to complete all jobs with fewer powerful nodes and save 70% in the cost of the datacentres.

NVIDIA® CUDA® Cores: 3584
Peak Double Precision Floating Point Performance: 4,7 Tflops (GPU Boost Clocks)
Peak Single Precision Floating Point Performance: 9,3 Tflops (GPU Boost Clocks)
Peak Half Precision Floating Point Performance: 18,7 Tflops (GPU Boost Clocks)
Capacity: 12 or 16 GB CoWoS HBM2 with ECC
Memory Interface: 4096-bit
Memory Bandwidth: 12 GB at 549 GB/s & 16 GB at 732 GB/s 
Bus Type: PCI-Express 3.0 x16
Thermal Solution: Passive
250 W
1 × 8-pin CPU power connectors

Write Your Own Review

You're reviewing: NVIDIA Tesla P100 Passive GPU

How do you rate this product? *

  1 star 2 stars 3 stars 4 stars 5 stars
Build Quality
Turnaround Time
Likely To Recommend
Overall Experience

NVIDIA Tesla P100 Passive GPU