Intel's Assembler Changes For JCC Erratum Are Not Hurting AMD
When writing about the Intel Jump Conditional Code (JCC) Erratum and how Intel is working to mitigate the performance hit of the CPU microcode update with patches to the GNU Assembler, there was some concern expressed by readers that it might hurt AMD performance. That does not appear to be the case.
On an AMD Ryzen 7 3700X box I installed Clear Linux and set it up in the same manner I used for this week's Skylake / Cascade Lake testing. In particular, comparing of Clear Linux builds 31470 and 31480 as that was the release where the patched version of the GNU Assembler was introduced and many bundles (packages) rebuilt as explained in the aforelinked article.
When running those tests on Clear Linux 31470/31480, there wasn't any real difference in performance. In just a couple of tests there were slight variations but nothing significant overall and also as pointed out in our earlier article the GNU Assembler behavior was odd in a few of the tests on Intel's own hardware.
So while some users were quick to speculate that it would hurt AMD or be some nefarious play by Intel, from this quick testing in repeating the same Clear Linux software setup as the JCC Erratum article, the updated assembler didn't introduce any real changes for the AMD Zen 2 system. Plus the assembler patches do add switches for controlling the behavior. Those wanting to look at the numbers in full anyhow can do so via this OpenBenchmarking.org result file.
On an AMD Ryzen 7 3700X box I installed Clear Linux and set it up in the same manner I used for this week's Skylake / Cascade Lake testing. In particular, comparing of Clear Linux builds 31470 and 31480 as that was the release where the patched version of the GNU Assembler was introduced and many bundles (packages) rebuilt as explained in the aforelinked article.
When running those tests on Clear Linux 31470/31480, there wasn't any real difference in performance. In just a couple of tests there were slight variations but nothing significant overall and also as pointed out in our earlier article the GNU Assembler behavior was odd in a few of the tests on Intel's own hardware.
So while some users were quick to speculate that it would hurt AMD or be some nefarious play by Intel, from this quick testing in repeating the same Clear Linux software setup as the JCC Erratum article, the updated assembler didn't introduce any real changes for the AMD Zen 2 system. Plus the assembler patches do add switches for controlling the behavior. Those wanting to look at the numbers in full anyhow can do so via this OpenBenchmarking.org result file.
28 Comments