Nvidia recently earned $14 billion in profit in one quarter thanks to AI chips. It is now planning to design new chips annually instead of every two years. The company has traditionally introduced a new architecture about once every two years, but Nvidia’s management says they will now accelerate the release of new chips, including new CPUs, GPUs, networking NICs, and switches. Nvidia’s new generations of AI GPUs are designed to be fully backward-compatible, allowing customers to easily transition from one generation to the next in their existing data centers. The company anticipates strong demand for its AI GPUs, with some customers having purchased over 100,000 Nvidia H100 GPUs. With ever increasing AI demand, those once sizable customer quantities may soon look small in comparison.
The ainewsarticles.com article you just read is a brief synopsis; the original article can be found here: Read the Full Article…
