Artificial intelligence for self-driving cars. Predicting our climate's future. A new drug to treat cancer. Some of the world's most important challenges need to be solved today, but require tremendous amounts of computing to become reality. Today's data centers rely on many interconnected commodity compute nodes, limiting the performance needed to drive important High Performance Computing (HPC) and hyperscale workloads.
NVIDIA® Tesla® P100 GPU accelerators are the most advanced ever built for the data center. They tap into the new NVIDIA Pascal™ GPU architecture to deliver the world's fastest compute node with higher performance than hundreds of slower commodity nodes. Higher performance with fewer, lightning-fast nodes enables data centers to dramatically increase throughput while also saving money.
With over 400 HPC applications accelerated—including 9 out of top 10—as well as all deep learning frameworks, every HPC customer can now deploy accelerators in their data centers.
The Tesla P100 is reimagined from silicon to software, crafted with innovation at every level. Each groundbreaking technology delivers a dramatic jump in performance to inspire the creation of the world's fastest compute node.
The new NVIDIA Pascal™ architecture enables the Tesla P100 to deliver the highest absolute performance for HPC and hyperscale workloads. With more than 21 TeraFLOPS of FP16 performance, Pascal is optimized to drive exciting new possibilities in deep learning applications. Pascal also delivers over 5 and 10 TeraFLOPS of double and single precision performance for HPC workloads.
Performance is often throttled by the interconnect. The revolutionary NVIDIA NVLink™ high-speed bidirectional interconnect is designed to scale applications across multiple GPUs by delivering 5X higher performance compared to today's best-in-class technology.
Note: this technology is not available in Tesla P100 for PCIe.
The Tesla P100 tightly integrates compute and data on the same package by adding CoWoS® (Chip-on-Wafer-on-Substrate) with HBM2 technology to deliver 3X memory performance over the NVIDIA Maxwell™ architecture. This provides a generational leap in time-to-solution for data-intensive applications.
Page Migration Engine frees developers to focus more on tuning for computing performance and less on managing data movement. Applications can now scale beyond the GPU's physical memory size to virtually limitless amount of memory.
Tesla P100 with NVIDIA NVLink technology enables lightning-fast nodes to substantially accelerate time-to-solution for strong-scale applications. A server node with NVLink can interconnect up to eight Tesla P100s at 5X the bandwidth of PCIe. It's designed to help solve the world's most important challenges that have infinite compute needs in HPC and deep learning.
Tesla P100 for PCIe enables mixed-workload HPC data centers to realize a dramatic jump in throughput while saving money. For example, a single GPU-accelerated node powered by four Tesla P100s interconnected with PCIe replaces up to 32 commodity CPU nodes for a variety of applications. Completing all the jobs with far fewer powerful nodes means that customers can save up to 70% in overall data center costs.
| P100 for PCIe-Based Servers |
P100 for NVLink-Optimized Servers | |
| Double-Precision Performance | 4.7 TeraFLOPS | 5.3 TeraFLOPS |
| Single-Precision Performance | 9.3 TeraFLOPS | 10.6 TeraFLOPS |
| Half-Precision Performance | 18.7 TeraFLOPS | 21.2 TeraFLOPS |
| NVIDIA NVLink™ Interconnect Bandwidth | - | 160 GB/s |
| PCIe x16 Interconnect Bandwidth | 32 GB/s | 32 GB/s |
| CoWoS HBM2 Stacked Memory Capacity | 16 GB or 12 GB | 16 GB |
| CoWoS HBM2 Stacked Memory Bandwidth | 732 GB/s or 549 GB/s | 732 GB/s |
| Enhanced Programmability with Page Migration Engine |  |
|
| ECC Protection for Reliability | |
|
| Server-Optimized for Data Center Deployment | |
|
The Tesla P100 is available now, only in the NVIDIA® DGX-1™ system—purpose-built for deep learning.
 | NVIDIA Blog |
 | |
 | |
 | YouTube |
