Although this seems to only affect the Chinese and Russian governments (as opposed to researchers), I'm generally against economic brinkmanship of this variety.

Although this seems to only affect the Chinese and Russian governments (as opposed to researchers), I'm generally against economic brinkmanship of this variety. It just festers resentment, and doesn't seem to achieve much: of what little I could find and scan in an hour, it appears that Nvidia GPUs strong-scale pretty well for neural networks ([1], [2]). Although this depends heavily on the NN architecture and training algorithm, Ampere architectures _seem_ to suffer from similar levels of atomic contention as Volta architectures [3]. That said, the throughput for half-precision floating point and integral data is much higher on Ampere [3]. Hopefully their final report includes NVLink bandwidth on Ampere. So what's stopping an adversary from just buying multiple V100s? Or potentially faster accelerators from countries outside the US [4]? If you're familiar with GPU strong-scaling, please chime in [5]! [1] - https://arxiv.org/abs/1807.09161 https://arxiv.org/abs/1807.09161 [2] - https://arxiv.org/abs/2111.04949 https://arxiv.org/abs/2111.04949 [3] - https://www.nvidia.com/en-us/on-demand/session/gtcspring21-s33322/ https://www.nvidia.com/en-us/on-demand/session/gtcspring21-s... [4] - https://arxiv.org/abs/1912.03413 https://arxiv.org/abs/1912.03413 [5] - https://meta.wikimedia.org/wiki/Cunningham%27s_Law https://meta.wikimedia.org/wiki/Cunningham%27s_Law