Many this really helps. You have no idea how many places I went trying to find these answers and all I found was people asking the same thing. I guess if your system is powerful enough that you see the frames shoot up high enough you know for sure whats what but if you are already maxing around your frame cap or in my case less. Then how could you possibly know. And the is no indicator or explanation of what the sounds are, in fact the AMD stuff doesn't even mention the sounds just that their is the F11 hotkey. So you are giving some great help here!
My monitor is 1440p 60hz native with a FreeSync range of 48 to 78. But in BF One for instance I run between 55 and 62 and that's about the best my card can do.
I have an XFX Black RX 580 that is a factory capable of 1425, I have mine at 1460 with notable improvement. I can run the card to 1475 stable but at that speed it is slower than 1460. My memory wont do as well as yours unfortunately 2020 is as high as I can go without instability. Was hoping for better as I even have Samsung ram. However it doesn't really seem to be memory restricted anyway. So unless I could push the GPU core a lot higher, I don't think it would really matter much. Again thanks for all the help and info, and your hardware info share too!
I have been having a little bit of trouble with knowing if my VRAM is stable just because it runs through benchmarks (Heaven, Valley, FurMark, etc.) just fine maxed out at 2250mhz, but then it will randomly crash my PC after I have been playing a game for an hour or two. I dropped it to 2200 and haven’t had it happen again so far. I was still seeing performance gain at 2250 vs 2200 in the benchmarks, never got any artifacting, it just randomly crashes at that speed when I play games. My PC will just randomly reboot itself… I kind of wish there was an easier way to tell if your VRAM was stable. When I OC core it is pretty obvious when you start getting artifacts in FurMark, they show up fast and because it is such a regular pattern they are easy to spot.
Sent from Mail for Windows 10
Mine only crashed when it was unstable too. Never saw any glitches or anything like that. I'd say 2200 is pretty decent. I'd be very happy with that. I will add that I could run through the benchmarks to without issue, but certain games would crash. Some are just way more temperamental than others. I like the heaven at this point to test stability as it runs fast enough now to know if it's a stability issue and not that I am taxing the card too much. I like the superposition benchmark to test the best overclock as you can really evaluate the differences down to an individual frame change because it is so taxing.
For checking core stability I really prefer MSI Kombustor under Combined FurMark & Tessmark load. I have to manually set my fans to 100% at all times and take the side panel off my case to prevent thermal throttling while I do it, but running the 1080p benchmark with the built in artifact scanner running is very good at picking up on any artifacts. I could get my core to 1435mhz before I noticed artifacts on my third run-through using Heaven, but when I ran it through kombustor it picked up artifacts inside of 20 seconds. I had to back it down another 10mhz to get it totally stable. Each run of kombustor only takes one minute, and once I found my max stable I only ran it 2-3 times to check. Works like a charm. It does get toasty though, my card maxed out at about 82c on the third run through w/ side panel off and fans going full blast. I do like that it only takes me 2-3 minutes to check for total stability, and that it has a built in artifact detector that picks up on artifacts that are to small or go by to fast for the eye to catch. It is also good for seeing improvements memory clock changes, as every time I brought up the memory by 25mhz it would give me a drastic change in FPS and benchmark score. Here are my scores for the 1080p benchmark on the Combined setting.
Score Core VRAM
8133 - 1425 - 2175
8192 - 1425 - 2200
8284 - 1425 - 2225
8339 - 1425 - 2250
I had used that utility a couple years ago. Neat what it does. I don't usually worry too much about it to get too technical. My workday is like that every day. When I get home I don't mind a dip the toe in the water approach to things.
Looks like the memory changes do affect that score quite a bit. Just don't know how that translates in a games fps. I will definitely have to check that out again sometime. Thanks for the share!