Tag
ArchAgent v2 from Google DeepMind automates multi-level data prefetcher design using agentic algorithm discovery, outperforming hand-designed solutions in the Data Prefetching Championship.
The Assembly Hall of Shame is a GitHub project that ranks the slowest possible single x86 instructions by exploiting hardware quirks such as MMIO latency and microcode assists, achieving extreme latencies like a 62-second fxrstor64.
A performance evaluation of Go programs compiled with different amd64 microarchitecture levels (GOAMD64) using the Roaring Bitmap library, showing that enabling newer instruction sets like popcnt (v2) or AVX-512 can significantly improve performance.
This article uses high-resolution micrographs to deeply analyze the internal structure, transistor layout, memory interface, and major functional blocks of the Apple M1 chip, showcasing its revolutionary breakthrough in performance and efficiency.
Detailed reverse engineering of the 386 processor's prefetch queue circuitry, explaining the incrementer, alignment network, and dynamic logic used.