"For example, we recently internally benchmarked an Inferentia instance (inf1.2xlarge) against a GPU instance with an almost identical spot price (g4dn.xlarge) and found that, when serving the same ResNet50 model on Cortex, the Inferentia instance offered a more than 4x speedup."
According to the Loadstar [1], this may be true for medical supplies, but demand for airfreight on a macro-level into Europe/US is dropping - a complete reversal from 3 weeks ago. The bullwhip effect [2] in action?
"The Loadstar warns air freight demand is expected to plummet as consumers in Europe and the US are forced to stop buying by shops closing. The automotive industry has already stopped requesting components as plants have closed; retail has cancelled significant volumes of orders - “Shopping is just not a priority,” said one forwarder. “The only real air freight demand now is medical equipment and some urgent products. “Just-in-time may be changed for ever. “There is also no cashflow.” He said air freight demand was now “the absolute reverse of two to three weeks ago; the dynamic has changed”. Another forwarder noted that while demand was, perhaps 15% of what it was three or four weeks ago, supply is at about 10%. “So there is still high demand in a relative sense.” [3]
"For example, we recently internally benchmarked an Inferentia instance (inf1.2xlarge) against a GPU instance with an almost identical spot price (g4dn.xlarge) and found that, when serving the same ResNet50 model on Cortex, the Inferentia instance offered a more than 4x speedup."
https://towardsdatascience.com/why-every-company-will-have-m...