The largest ASIC chip ever built, an AI capable machine. This is a very specialized hardware and a marvellous of engineering.

By the end of 2Q 2020 there will be already some use cases tested, hopefully.

The WSE has 18 GB of on-chip memory, all accessible within a single clock cycle, and provides 9 PB/s memory bandwidth. This is 3000x more capacity and 10,000x greater bandwidth than the leading competitor. More cores with more local memory enable fast, flexible computation, at lower latency and with less energy.

Specifications:

Supports TensorFlow and PyTorch frameworks
Sparse Linear Algebra Compute (SLAC) Cores: 400,000
On-chip Memory (SRAM): 18 GB SRAM
Memory Bandwidth: 9.6 PB/sec
Interconnect Bandwidth: 100 Pb/sec
System I/O: 1.2 Tb/s
Dimensions: 15 rack units (15U)
Colling System: Water-cooled with redudancy
Power consumption: up to 20Kw
Price: Can't disclosure :( sorry