"Google's AI chips outperform Nvidia and Qualcomm in efficiency tests"

TL;DR Summary
Qualcomm's Cloud AI 100 chip beat Nvidia's H100 chip in two out of three measures of power efficiency in AI chip tests conducted by MLCommons. Qualcomm's chip was more power-efficient in classifying images and object detection, while Nvidia's chip was more efficient in natural language processing. Neuchips, a Taiwanese startup, took the top spot in power efficiency for natural language processing. The market for data center inference chips is expected to grow as businesses put AI technologies into their products, and power efficiency is a major cost consideration.
- Qualcomm, Nvidia spar for top spot in AI chip efficiency tests Reuters
- Google reveals its newest A.I. supercomputer, says it beats Nvidia CNBC
- Google takes AIM at Nvidia, claims AI chips are faster, greener CNBC Television
- Alphabet Says Its AI Chips Can Beat Nvidia's. It's Not Just Taking on Microsoft. Barron's
- Nvidia slips as Google touts prowess of its own custom chips (NVDA) Seeking Alpha
Reading Insights
Total Reads
0
Unique Readers
12
Time Saved
1 min
vs 2 min read
Condensed
77%
387 → 88 words
Want the full story? Read the original article
Read on Reuters