World’s first 12nm FFN GPU has just been announced by Jensen Huang at GTC17. The new Tesla has the second generation NVLink with a bandwidth of 300 GB/s. Tesla V100 utilizes 16 GB HBM2 operating at 900 GB/s.
The card is powered by new Volta GPU, which features 5120 CUDA cores and 21 billion transistors. This is the biggest GPU ever made with a die size of 815 mm2.
Volta GV100 features a new type of computing core called Tensor core. The purpose of this core is deep learning matrix arithmetics.
Jensen said that the cost of Tesla V100 development was 3 billion dollars.
 
Key features of Tesla V100:

• New Streaming Multiprocessor (SM) Architecture Optimized for Deep Learning
• Second-Generation NVLink™
• HBM2 Memory: Faster, Higher Efficiency
• Volta Multi-Process Service
• Enhanced Unified Memory and Address Translation Services
• Cooperative Groups and New Cooperative Launch APIs
• Maximum Performance and Maximum Efficiency Modes
• Volta Optimized Software

 

• New mixed-precision FP16/FP32 Tensor Cores purpose-built for deep learning matrix arithmetic;
• Enhanced L1 data cache for higher performance and lower latency;
• Streamlined instruction set for simpler decoding and reduced instruction latencies;
• Higher clocks and higher power efficiency.

 
 


It provides a 5x improvement over Pascal, the current-generation NVIDIA GPU architecture, in peak teraflops, and 15x over the Maxwell architecture, launched two years ago. This performance surpasses by 4x the improvements that Moore’s law would have predicted.
 
https://videocardz.com/69...0-with-5120-cuda-cores
 
look at the Die size 815mm
 
and Dang that isn't the full Die or cores
 
The chip itself is a behometh, featuring a brand new chip architecture that is just insane in terms of raw specifications. The NVIDIA Volta GV100 GPU is composed of six GPC (Graphics Processing Clusters). It has a total of 84 Volta streaming multiprocessor units, 42 TPCs (each including two SMs). The 84 SMs come with 64 CUDA cores per SM so we are looking at a total of 5376 CUDA cores on the complete die. All of the 5376 CUDA Cores can be used for FP32 and INT32 programming instructions while there are also a total of 2688 FP64 (Double Precision) cores. Aside from these, we are looking at 672 Tensor processors, 336 Texture Units.
 
no word on consumer GPU's
Expect late Oct at next GTC