https://www.phoronix.com/scan.php?page=article&item=windows10-october-linux I just happen to come across this review and found it interesting. No mention about meltdown and spectre if both or all OS had microcode loaded that I could find.
All tests are performed in up to date installations. That means meltdown/spectre patches are enabled on all OSes.
While interesting, those comparisons are ultimately useless unless you're looking for the best performing Linux distro. Even then, it's possible that the benchmark results won't be indicative of real world performance.
That's a pleasant opinion. Seems like a very reasonable statement. Thanks for taking the time to share. I guess Michael can just quit now.
I only posted the link cause I thought some here might be interested, I only run windows so not really big thing, but I rarely come across BM like this.
The usefulness of direct comparisons between operating systems is dependent on the desired functionality of the operating systems involved. For most of us, the comparisons made in the linked article are useless. If you rely on software that performs the same operations as the benchmarks used, it's possible that the comparisons are valid. However, the benchmark results may not always be indicative of which distro actually performs the operations faster due to the code used. Personally, I prefer Ubuntu over Windows. Unfortunately, Ubuntu lacks functionality that I require. If I performed the operations used in those benchmarks on a regular basis, I'd be running Clear Linux instead of Windows. So, while the article was an interesting read, for me (and most others here) the benchmark results and thus, the comparison, are useless. Doesn't change the fact that it was a well written article, presents the data in a very clear manor and provides data that may be useless to those that actually need such data. It's just useless to gamers and people like me that don't use their system for such tasks.
@sykozis The Phoronix OS benches have a long history. They are re-done every couple months. People have several uses for them. One of them includes tracking how the distros change over time in terms of perf. They happen to be a source of useful information by the distro developers, just like the kernel developers look at the kernel benches to see if there's unintended perf regressions. Another use is simply informing the public whether or not Linux runs games better or not. When a Random J. Nobody claims that, generally, Windows games run faster than Linux games, you don't have to take their word for it. You can just look at these benches. When another Random Nobody says Linux is faster at server related workloads (databases, HTTP, etc.), you also don't need to take their word for it, you need some proof. These benches are useful, just like any other bench out there. When you say that a 1080 Ti is faster than a Vega 64 in the vast majority of games, I don't want to just take your word for it. I need some proof. I'm gonna look at benches.
