Nvidia will now design new AI chips annually instead of biennially, with upcoming architecture 'Rubin' expected in 2025, accelerating all chip types to match this cadence for various applications.
New generations of AI GPUs are backward-compatible, easing transitions for customers in data centers, with overwhelming demand due to cost efficiency and profit opportunities.
Nvidia CEO highlighted incredible demand for AI GPUs, expecting supply shortages during transitions to new chip generations, emphasizing savings and revenue generation for customers.
The company plans to release a 'mountain of chips' including CPUs, GPUs, networking NICs, and switches as part of the accelerated design strategy.
[
add
]
[
|
|
...
]