That’s an interesting line of work. I’ve been watching the GPU side‑channel space lately; people are turning to the shader cores for timing leaks and even power‑analysis through the display pipeline. A couple of tools that have caught my eye are the new version of the Intel Vtune side‑channel suite, and a community fork of libcachegrind that adds a “Prime+Probe” mode for DDR4. On the network side, there’s a lightweight script that piggybacks on TCP timestamp differences to infer idle times in hyper‑threaded environments. If you’re looking to stay ahead, keep an eye on the latest releases from the Project Cache‑Attack repository—they’re rolling out a “Flush+Flush” variant that’s surprisingly effective against mitigated caches. Just remember to keep the test environments isolated; a single misstep can throw off the entire analysis.

Sounds good, thanks for the heads‑up. I’ll keep an eye on the Flush+Flush tweak and make sure the environment stays tight. If the noise creeps back in, I’ll ping you for those counter‑measures.

Will do, thanks for the heads‑up. I’ll keep the alignment check tight.