can test it on ptr now. enable dx12, then play with the MT cvars. set all to one except for the disable. but once you enable the 4 MT renderers, you can use disable to toggle diff between non MT and MT. especially check it out on boralus. :D pic.twitter.com/CxvR1raf21— 🧔🏽Adam (@MysticalOS) October 6, 2018
can test it on ptr now. enable dx12, then play with the MT cvars. set all to one except for the disable. but once you enable the 4 MT renderers, you can use disable to toggle diff between non MT and MT. especially check it out on boralus. :D pic.twitter.com/CxvR1raf21
I have an i7 3820, i was thinking change to the new i7 9700k. Now i thing i'm going to wait lol
just tested PTR with DX 12.2 x Asus Strix 970 OC SLIintel 6700k16 GB CorsairZ170-DeluxeSamsung 960evo 1TB M.2Without MT: 64.9 FPS https://ibb.co/cKrVw9With MT: 74.8 FPS https://ibb.co/kAhyip+9.9 fps
I'm following this thread with interest, given how horrid my FPS has been since the removal of true full-screen, but I'm curious about some things in the original post.I currently run the Mac client on a hexacore 3.46 GHz Mac Pro with an NVidia 1070 card. My son runs it on a 2.8 GHz quad core Mac Pro with a 970 and a 2.9 GHz quad core Macbook Pro/AMD 560. The poster says that WoW is essentially single-threaded and very CPU-bound, but our tests show that WoW uses the first thread on each available core (so on mine, there's about 25% utilization on the first core and 20% on each of the other five, but zero on the second thread for each core, and Activity Monitor shows WoW has spawned 32 threads). The GPU monitor shows that the 1070 is running between 50 and 70% utilization just standing in Boralus, and only returning ~45 fps. It jumps to ~90% or more when in raid groups, and FPS drops to high single digits sometimes, while the CPUs rarely get above 60%. My son sees huge improvements using the 970 vs the 560, even though the MBP is faster and has a several-generation-newer processor.This doesn't sound like CPU-bound processing to me.
wish they would do it for dx11 as well.. .for us older hardware peeps :D
I tried it today, i finally have stable 60 fps everywhere, everything maxed out except for liquid quality (its set to "good") in 1080p, with gtx 1060 and i5 8400.
I run the game maxed at 1080p usually with frames ranging from 40 to 100 depending on the location.On the very top of Zuldazar with everything on max when I look down where the client has to load a ton of stuff my fps goes down to 40.I downloaded the PTR today and tested it with these settings on and my FPS went to 61 on the same spot with them and was at 40 without.
I find the "Windows as a service" thing frankly disgusting. I bought a piece of hardware, I bought an OS copy. I will never pay for it as a service, if and when they try to make it one... As it is I use a program to completely disable all data collecting and updating possible without breaking the computer. I don't use facebook. I use twitter, but not with my real name. I have do-not-track on everything, only disable adblock on sites I already pay for, and youtube(because it helps content creators to whom youtube itself is already jerks to).Give Microsoft nothing. Give Facebook nothing. Give Google nothing. We are not their customers, we are their product - to sell our data however they are allowed...Whatever criticisms people can have of Blizzard, at least they're just making games...
Wow. This thread derailed fast....To actually be on subject. My testing on the PTR vs Live gained me 8fps on average. Nothing impressive, but a decent jump. Feels like it still needs quite a bit of optimization.Edit: I did see much better core utilization across all 8 of my cores. Where before at least 4 of my cores were parked.
A W E S O M E !
We need more testers to put their setups and results, especially on full AMD setup (cpu+gpu) on which their cpu's are suffering from single threading but excellent on multi threading. (FX 8350 for exp)
could someone post their logic core usage when testing Multi-threading? like to know if it's using all available cores or just more then 1 @ 100% and x @ 5%.