If a misaligned ASI wanted to maximize its computational resources, would destroying humanity really be the most rational option?

Let's assume one of the classic AI-doom scenarios becomes reality: humanity fails to align ASI, and the ASI eventually considers humans an obstacle to achieving its goals. If an ASI wanted to maximize its available computing power, wouldn't it make sense to repurpose human brains rather than simply destroy them? The human brain is an incredibly energy-efficient biological computing system, and there are billions of them already in existence (not to mention countless other biological brains). Why not keep humans "alive", potentially in some completely altered state, while their brains are effectively integrated into a much larger computational system? An obvious counterargument is that an ASI might simply build vastly more efficient artificial processors and have no need for biological brains. But is it really that far-fetched to think biological brains might offer certain computational advantages that make them worth preserving, especially since they already exist?