Multi core performance matters?

[Griefs-emerald-dream (Griefs)]

Good to have you back, Kagthul

[Kagthul-bleeding-hollow (Kagthul)]

just here for Classic.

I quit posting all that time ago because there had apparently been a bug that was letting people with lapsed accounts continue to post, that they ‘fixed’ in May or whenever it was i quit posting, since i didn’t have an active account (and hadn’t for some time).

I re-activated today for Classic.

Good to be back, i guess. Still wish there was an ignore feature.

[Griefs-emerald-dream (Griefs)]

Still wish there was an ignore feature.

Having the script on only my home desktop keeps me from checking these specific forums very often. While I feel bad for the people who come here and get crap advice from the resident fools, it’s just not worth it to me to step in as often as needed to provide a rational position.

The WPvP and Classic General forums are much better places, anyway.

[Aribeth-arygos (Aribeth)]

This is literally like the 30 billionth time you’ve posted something that quite literally proves the exact opposite of what you’re claiming it does.

You and Griefs REALLY need to have your eyes checked.

1 - This is new data so who know what you’re thinking of.
2 - 4 cores vs 8 core is a 10% performance improvement. How is that ‘opposite’ of more cores = better performance?

Also… “using 16 cores” does NOT mean it performs better with more cores…

Going from 4 core to 16 cores having an 18% performance. 18% isn’t an improvement over 0? You’re math impaired.

A faster clocked Intel chip will BURY a Ryzen chip in sustainable minimums (can literally be the difference between maintaining a solid 60+ in a busy raid setting and bottoming out in the 50s).

Great more BS and no proof.

The CPU CANNOT ISSUE THE DRAW CALLS FAST ENOUGH , so the GPU starves. THATS A CPU BOTTLENECK. Like, the very definition of one.

Lowering the graphics to 1 would still have poor frame rates. Guess what? Lowering graphics to 1 is smooth as butter, proving your dribble useless (ie the GPU is the bottleneck). A CPU can feed draw calls and not have the GPU render it (ie GPU bottleneck). Man you’re clueless.

I mean… seriously, loved watching you claim in another thread that a Ryzen 3000 series chip running 600+ mhz slower than an Intel chip, given IPC parity (as IPC between Zen 2/Ryzen 3 and 8000/9000 series chips is within the margin of error) will somehow outperform an Intel chip, despite a 600mhz deficit.

Maybe because I actually watched the benchmarks while you didn’t… Shock! People who read actually know more than you.

Did you just use mhz as a comparison? ROFL ok man, maybe next you’ll tell us how the 5000mhz Bulldozer will be better than the ‘4000mhz modern chips’.

I mean… seriously. What is wrong with you? Is it like some form of information dyslexia? You read the information, but it turns backwards in your head?

Says the guy who thinks 18% improvement is ‘no improvement’. You need to figure out what improvement means.

Whatever you do, seek help. Serious, professional help. You need it.

You should take some reading comprehension classes first. Start by learning how to read graphs. Maybe they’ll have a 2 for 1 sale with you and Griefs, we all know you need it.

The WPvP and Classic General forums are much better places, anyway.

Then stay there and stop offering people bad advice with 0 proof attached.

[Kodiack-barthilas (Kodiack)]

The multi-core performance enhancements in the 8.x patches have been fairly substantial. I benchmarked some quick and dirty results on a 4.5 GHz 9900K with a 2080 Ti and render scale set to 0.25 to ensure that CPU was the bottleneck.

[Album] imgur.com

I probably should have thrown a 2-core/4-threads result in there as well. I’ll do that if I rerun it sometime. Regardless, WoW scales very comfortably up to four cores, and then sees marginal improvements beyond that. There are no doubt situations where scaling will be somewhat different, possibly pushing even higher minimums with additional cores.