"Fluid Compute": Streaming requests to AWS lambdas for increased efficiency
theregister.comSo this tech has been out there for a while now and Vercel users are starting to see the benefits. If this IS working out well, it won't be long before non-Vercel solutions start popping up.
AWS have already calculated the pricing model of lambda to take advantage of this massive inefficiency of their lambda product, they charge multiple customers simulatnaeously to use for the same hardware for the compute that would otherwise be sitting idle 95% of the time due to Lambda's basic design.
If large numbers of customers start using this model to reduce AWS' ability to double-dip, it's going to severely affect the profitability of Lambda, I wonder how AWS will respond?
By increasing its price?
Yes, no doubt.
But a naive increase will punish the majority of customers who aren't yet using a similar efficiency mechanism.
My guess, if this approach becomes popular - we'll see a whole new layer of complexity added to Lambda pricing making it even harder to estimate lambda costs ahead of time.
True, and I can only imagine the types of tiered pricing that no one will understand.