![]() The photo editing benchmark measures performance with Futuremark's binaries using the ImageMagick library. Here we see the 3600/Stealth combo beat the H115i-equipped configuration, but the deltas are small enough to chalk up to expected run-to-run deviation. It also performs facial detection to model real-world usage. Our video conferencing suite measures performance in single- and multi-user applications that utilize the Windows Media Foundation for playback and encoding. In any case, the Ryis incredibly competitive given its price point, easily beating the -9400F. We see this trend repeat a few times throughout our test suite. The Ryactually loses a little performance when we kick on PBO with the Wraith Stealth cooler, but the result lands within the expected 3% variance of this test, implying there is no uplift from overclocking in some applications with the stock cooler. Other platform-level considerations affect this test as well, including the storage subsystem. The application start-up metric measures load time snappiness in word processors, GIMP, and Web browsers under warm- and cold-start conditions. ![]() The Core i5-9400F even trails the first-gen Ryzen 5 1600X, highlighting how much Intel's feature-trimming (in this case, Hyper-Threading) hamstrings its lineup against AMD's more well-rounded chips. The scripts are compiled using buildbot-benchmark.bash from the scriptlets repository, with a CPU frequency of 1.843 Ghz and 16G of RAM.The LLVM compiler benefits from extra threads, handing the 3600 an easy lead over the competing Intel chips, even after overclocking. ![]() Note: The tests here use the public version of the toolchain, if you use the special toolchain of thead, the performance is expected to be improved by more than 50%. Test results (1.85GHz) using the Thead toolchain, GCC version 10.4.0: 2K performance run parameters for coremark.Ĭompiler flags : -O2 -march=rv64gv0p7_zfh_xtheadc -O3 -funroll-all-loops -finline-limit=500 -fgcse-sm -fno-schedule-insns -msignedness-cmpiv -fno-code-hoisting -mno-thread-jumps1 -mno-iv-adjust-addr-cost -mno-expand-split-imm -DPERFORMANCE_RUN=1 -lrtĬoreMark 1.0 : 13219.644392 / GCC10.4.0 -O2 -march=rv64gv0p7_zfh_xtheadc -O3 -funroll-all-loops -finline-limit=500 -fgcse-sm -fno-schedule-insns -msignedness-cmpiv -fno-code-hoisting -mno-thread-jumps1 -mno-iv-adjust-addr-cost -mno-expand-split-imm -DPERFORMANCE_RUN=1 -lrt / Heap See README.md for run and reporting rules.ĬoreMark 1.0 : 9175.077154 / GCC13.1.0 -O2 -DPERFORMANCE_RUN=1 -lrt -funroll-all-loops -finline-limit=500 -fgcse-sm -fno-schedule-insns -msignedness-cmpiv -fno-code-hoisting -mno-thread-jumps1 -mno-iv-adjust-addr-cost -mno-expand-split-im / Heap Memory location : Please put data memory location hereĬorrect operation validated. Test results (1.85GHz) using the public toolchain, GCC version 13.1.0: 2K performance run parameters for coremark.Ĭompiler flags : -O2 -DPERFORMANCE_RUN=1 -lrt -funroll-all-loops -finline-limit=500 -fgcse-sm -fno-schedule-insns -msignedness-cmpiv -fno-code-hoisting -mno-thread-jumps1 -mno-iv-adjust-addr-cost -mno-expand-split-im funroll-all-loops -finline-limit=500 -fgcse-sm -fno-schedule-insns -msignedness-cmpiv -fno-code-hoisting -mno-thread-jumps1 -mno-iv-adjust-addr-cost -mno-expand-split-imm The official data from Flathead, C910 is around 5.6 out of 10.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |