Questions
Read the code, predict what happens, then see why. Hand-written questions on C++, Rust, and the systems underneath them.
- Single-core bandwidth for independent random DRAM loadsHard
- Branchy vs. branchless code on unpredictable dataMedium
- A vectorized scale loop whose buffers differ by 4 KB plus 64 bytesHard
- Store-to-load forwarding for two store/load overlap patternsHard
- Reading back an address whose store hasn't drainedMedium
- Exiting a spin-wait loop on a flag, with and without `pause`Hard
- L3 eviction of a line in a strictly inclusive hierarchyHard
- One running sum vs. four partial sumsHard
- Two instructions that both write R1Hard
- Message-passing litmus test on ARMv8 with a DMB between the storesHard
- BTB behavior for monomorphic vs. megamorphic virtual callsHard
- Two counters eight bytes apart, two coresMedium
- Typical access latencies from L1 through DRAMEasy
- A buffer first-touched on one socket, consumed on the otherMedium
- Striding down a column of a 1024x1024 matrixHard
- Doubling a VIPT L1D from 32 KB to 64 KB at fixed associativityMedium
- Auto-vectorizing `a[i] = a[i - 1] * 2 + b[i]` at -O3Medium
- Driver read of a DMA-written buffer on a non-coherent platformMedium
- Sorted vs. unsorted iteration through a branchy loopMedium
- Return prediction after a longjmp out of a deep call chainMedium
1–20 of 47
Page 1 of 3