kaivalnp commented on code in PR #15736:
URL: https://github.com/apache/lucene/pull/15736#discussion_r2835049070
##########
lucene/core/src/java25/org/apache/lucene/internal/vectorization/PanamaVectorUtilSupport.java:
##########
@@ -610,11 +610,10 @@ private static int int4DotProductSinglePackedBody(
prod8a.convertShape(ZERO_EXTEND_B2S, Int4Constants.SHORT_SPECIES,
0);
acc1 = acc1.add(prod16a);
}
- Vector<Integer> intAcc0 = acc0.convert(S2I, 0);
- Vector<Integer> intAcc1 = acc0.convert(S2I, 1);
- Vector<Integer> intAcc2 = acc1.convert(S2I, 0);
- Vector<Integer> intAcc3 = acc1.convert(S2I, 1);
- sum +=
intAcc0.add(intAcc1).add(intAcc2).add(intAcc3).reinterpretAsInts().reduceLanes(ADD);
+ ShortVector accShort = acc0.add(acc1);
+ Vector<Integer> intAcc0 = accShort.convert(ZERO_EXTEND_S2I, 0);
+ Vector<Integer> intAcc1 = accShort.convert(ZERO_EXTEND_S2I, 1);
+ sum += intAcc0.add(intAcc1).reinterpretAsInts().reduceLanes(ADD);
}
Review Comment:
@mccullocht In fact, I went one step ahead and removed all `convert`
operations, specially shape changing ones -- so we load and operate on the same
number of bytes (see [this
commit](https://github.com/apache/lucene/commit/9059dc570afb9440b0178faaf8baa90879decbf0)),
and I get much better performance on my machine (\~25% bump from this PR):
```
Benchmark (size)
Mode Cnt Score Error Units
VectorUtilBenchmark.binaryHalfByteDotProductSinglePackedScalar 1024
thrpt 15 2.450 ± 0.006 ops/us
VectorUtilBenchmark.binaryHalfByteDotProductSinglePackedVector 1024
thrpt 15 19.166 ± 0.028 ops/us
```
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
