Another round of speculative-execution vulnerabilities

That's exactly why there are multiple layers of cache. Each layer physically and architecturally closer to the core.

And also pre-fetching so that the slow memory access can happen while the CPU is doing other calculations.