This year's Hot Chips symposium on high performance chips will be held in California, August 19-21, 2018, and the conference's program lists sessions about NVIDIA's next generation mainstream GPU, and Intel's next generation "Cascade Lake" Xeon scalable processor.
At the first day of the conference, in August 20th, NVIDIA's Stuart Oberman (vice-president of GPU ASIC engineering) will talk about "NVIDIA's Next Generation Mainstream GPU". Obviously there will be an introduction to NVIDIA's next-generation computing solution under its gaming GeForce brand (GTX11). That date suggests Nvidia will tease or launch its next-generation gaming chip earlier in the summer.
The symposium's schedule also lists a session entitled "Next Generation Intel Xeon(R) Scalable processor: Cascade Lake". Intel's new server architecture is expected to support multi-core processors with Intel to offer up to octa-socket support with Cascade Lake to furhter increase the maximum physical core count of a single system. The number of the supported DDR4 memory channel is also expected to be high, as Intel wants to compete with AMD's EPYC server chips.
There are also some interesting subjects at the event, such as Intel's "High Performance Graphics solutions in thin and light mobile form factors," hinting at the Intel-AMD collaboration in Kaby Lake G, "ARM's First Generation ML Processor" and more.