3/12/25
abstract background banner, ones and zeros --ar 16:9 --sref 455808632 --v 6.1
Note: The below represents my personal point of view, and is not an official opinion held by my employer, Tower Research, its affiliates, nor any of their private investments.
We have electronic markets for almost every commodity used in some kind of industrial process. We have markets for electricity; we have markets for silicon, aluminum and steel; we have markets for cobalt, lithium, copper, and nickel. In theory, computation is both abundant and commoditized, so why don’t people trade it on exchanges? The value proposition seems intuitive:
“Imagine if you could buy a spot instance on a machine anywhere in the world!”
“Imagine if your random SaaS company could hedge out its cloud contract by selling a future for its excess capacity!”
“Imagine if your semiconductor company could hedge out next year’s production!”
This is an idea I see a lot of, given what my employer does in (1) making markets in liquid assets and (2) building and using highly-performant compute systems. While the idea seems intuitive, we’re about 20 years into the hyperscaler cloud era, so it’s worth thinking through why we haven’t seen any true computation exchange emerge, whether we never will, or what needs to change if we do.