Independent ALU chains¶

Generated workload catalogue · Execution contract

Purpose¶

Expose work that does not depend on the immediately preceding ALU result.

Setup¶

Property	Value
Workload ID	`independent`
Selection	`--workload independent`
Generator	`buildRV32IBenchmark() in RV32IBenchmarks.cpp`
End condition	EBREAK after the loop and result store

Eight accumulators x16..x23 start at 1..8. The input-data size does not set this fixed 64-operation body.

Use the shared address map and parameter bounds. N includes any training iterations; n means size.

Implementation¶

Issue 64 instructions in rounds across eight chains: ADDI +3, XORI 85, SLLI 1, SRLI 1, repeated twice per chain. The final reduction sums all eight accumulators.

for (unsigned round = 0; round < 2; ++round)
  for (auto& v : chains) v = uint32_t((v + 3) ^ 85) << 1 >> 1;

This is a behavioral summary of the generated assembly, not the host program being timed.

Expected result¶

Starting from chains={1,2,3,4,5,6,7,8}, apply the body N times and sum modulo 2^32. Shift-left truncates before logical shift-right. The generator computes this expected checksum independently of the instruction oracle.

The runner checks the generator's expected result in addition to retirement-by-retirement oracle validation. Output-array assertions, where described, check more than a checksum.

Measurement and running¶

Compare with the dependency-distance sweep. These are independent chains, not instruction-level parallel execution: all current cores remain single-issue.

build/rv32i_experiment --core advanced --workload independent \
  --iterations 2 --size 16 --seed 1  \
  --fidelity behavioral --period 4000 \
  --allocation-limit-mib 512 --max-cycles 4096

The explicit period is for a bounded functional check, not a hardware-speed result. For comparisons, use the serial campaign driver, identical case inputs, and a certified per-target clock.

Source and tests¶

src/analysis/RV32IBenchmarks.cpp: emitted instructions and independent expected results.
tests/cpp/RV32IBenchmarkTest.cpp: generator results, parameter bounds, and edge cases.
tests/checks/RV32IBenchmarkCLITest.py and tests/checks/BenchmarkCampaignCLITest.py: shared execution/campaign checks.