hardware-effects
hardware-effects copied to clipboard
Kobzol
Demonstration of various hardware effects.
Hey, I've run the bandwidth-saturation on ryzen 9 5950x, debian bullseye and got some concerning results: ``` ./bandwidth-saturation 0 32 -> 224mS ./bandwidth-saturation 1 32 -> 1979mS ``` Those benchmarks...
Respected Admin, Kindly add a citation for this repository to make it citable in any publication. Regards, Deepak
I am curious what is the 'correct' perf counter for 4K aliasing. You have mentioned `ld_blocks.store_forward`, but I was wondering about the other counter `ld_blocks_partial.address_alias` as well. Here is the...
This isn't really an issue, but more of a request. I think it would be really nice to have a program (much like cache-hierarchy-bandwidth) which determine the cache line size....
remove the cache effect (I try several times, not as obvious as directly write to memory). So that the misaligned access effect can be seen. (Correct me if I am...
Hi again, here's my plot for the cache-memory-bound-test:  Do you have any explanation for the measurements after increment 32? The processor is a Xeon(R) Gold 6130...
I'm toying around with the cache-conflicts code, and I'm using a higher count than you did in your benchmark. However, the repetition count than makes the program run for absolutely...
- [x] non-temporal stores - [x] multiple threads saturating the memory bus - [x] hardware prefetching with indexed accesses - [x] floating point handling (denormals etc.) - [x] 4k aliasing...
This is in reference to #4. It makes direct use of perf events in Linux for measuring TLB misses, but also times each run. Overall, the results I got seem...
when run it on ``` perf stat -e resource_stalls.sb store-buffer-capacity ``` the resource_stalls.sb result varied a lot(each time, got quite different result). It seems can not use this way to...
