Lines Matching full:arithmetic
598 # latency is lower, especially on Intel P4. While arithmetic601 # so that "arithmetic throughput" remains virtually the same. And
