Blog: AI’s Memory Problem
Artificial intelligence, which allows computers to learn and perform specific tasks such as image recognition and natural language processing, is…
AI Inference Reaches 8.8 TOPS/W Due to PIM Techniques
Renesas increases accuracy, reduces power by adopting ternary memory structure.
NeuroBlade Preps Inference Chip
NeuroBlade emerged from stealth mode with plans to sample an AI inference accelerator by early 2020, with additional plans for…
HBM2e Top Contender for AI Applications
HBM2e and GDDR6 DRAMs are vying for AI applications. However, HBM2e has a competitive edge. But you be the judge
HoloLens 2, Neural Probes and More at Imec’s Summit
Speakers described work on everything from mixed-reality to deep learning, solid-state batteries and brain science at Imec's annual event
Imagination GPU Joins SiFive RISC-V Design Ecosystem
There are not a lot of other options for GPU IP that is not tied to a specific vendor
US, China Heading Towards Tech Cold War
The world might be divided between US-led and China-led spheres as the competition between the two countries carries over into…
Are GPU Limitations Slowing Innovation?
GPUs are widely used to accelerate AI computing, but are the limitations of GPU technology slowing down innovation in the…
AI Research Targets Nvidia V100
UT Austin says a chip it has developed can, in simulations, beat the V100 on training neural nets. The claim…
TinyML: Opportunities Bigger than the Challenges
200 engineers and researchers gathered in California to discuss forming a community to cultivate deep learning in ultra-low power systems,…