Tensorflow

Qualcomm revealed the initial details of its cloud 7nm Cloud AI inference processor, saying it will deliver a 10x performance increase over competing devices.

Google launched its Coral dev board and USB Accelerator with embedded Edge TPUs, promising a large boost in machine learning inference performance for all IoT devices that integrate them.

AMD and Xilinx put Intel and Nvidia on notice with a new data center system that beat the world record for machine learning inference.

Habana Labs, a new AI chip startup, promises to deliver much higher inference performance than even a machine learning-optimized GPU.

Nvidia CEO Jensen Huang unveiled the new Tesla T4 GPU with the Turing architecture at GTC Japan.

Rockchip announces its new AI-focused RK3399Pro SoC, which an embedded AI performance of up to 2.4 trillion operations per second.

CEVA announced the new NeuPro AI processors for embedded devices ranging from wearables to connected cars, with a performance starting at 2 TOPS and reaching 12.5 TOPS.

Google announced TensorFlow Lite, a lighter-weight version of the TensorFlow software framework and a successor to TensorFlow Mobile that's more efficient on mobile and embedded devices.

Nvidia revealed the third generation of its TensorRT AI inferencing software, the DeepStream video analytics platform, and the ninth generation of Nvidia’s CUDA technology.