Benchmarking a distribution (and some \-O3 results) | Why Ubuntu reverted move to -O3 compiler flag
-
This post did not contain any content.
-
System shared this topic on
-
This post did not contain any content.
The processors in these were 4-core AMD EPYC Rome CPU.
Surely they meant 48 core?
-
The processors in these were 4-core AMD EPYC Rome CPU.
Surely they meant 48 core?
I was wondering about that too. At first I assumed they were only allocated a few of the cores for their testing, but a typo seems more likely.
-
The processors in these were 4-core AMD EPYC Rome CPU.
Surely they meant 48 core?
There are 4 core EPYCs (
they probably make sense when PCIe lanes and memory channels are more important than raw CPU powernevermind, they actually have 16 PCIe lanes and 2 memory channels), and you don't need a very powerful machine to run benchmarks to find out how certain compiler flags affect performance. -
The processors in these were 4-core AMD EPYC Rome CPU.
Surely they meant 48 core?
There are 4 core romes, no idea why anyone would buy one (think they keep most of the ddr and pcie links) but still.
-
There are 4 core romes, no idea why anyone would buy one (think they keep most of the ddr and pcie links) but still.
That was my deleted comment, but then I realized the article specifies they are Rome CPUs, while these low cost ones are Raphael.
-
System shared this topic on
