Du verwendest einen veralteten Browser. Es ist möglich, dass diese oder andere Websites nicht korrekt angezeigt werden. Du solltest ein Upgrade durchführen oder einen alternativen Browser verwenden.
vllt. hängts auch vom Treiber ab? mein nichtvalidierter UWP-Treiber hat ne gute Auslastung@DX12
130W+200W sind Beide ungefähr gleich gut ausgelastet mit meinem mickrigen i5-9400F
sprich, es wurde mit 200W nicht schlechter
Mal abwarten, ob AMD in nem späteren WHQL auch nachbessert.
btw.
Den ExplorerMode muss ich mal noch erkunden, ... habe dafür auch AFMF=on,
weil sonst 4k zu happig ist für mein 24/7@130W.
@GerryB why won't Steel Nomad run on my system? Timespy works fine, but when I run the new benchmarks, I just get an error message saying results missing.
For me it´s not possible to upload a older saved result. sometimes a run goes properly and sometimes not, who knows what they are doing
server to small?
today two Runs, one is ok and one has the non-web message in the red triangle
For me, the benchmark doesn't even run; it just skips to the result screen immediately and says "error result missing." I tried multiple drivers to no avail.
What's the ratio of the Steel Nomad score to TimeSpy? I can't tell if this score is good or not. Steel Nomad Light is much easier to run than TimeSpy, though it's not a good test for stability.
The highest I can do at 1175mv is 2770MHz, but that requires almost 400 watts to sustain; that is that is not possible on my cooler. What I find weird is that Steel Nomad, either light or regular, allows for like 30MHz above TimeSpy at each respective voltage level. I'm trying to find a stable setting for exactly 280 Watts; I had 1050mv at 2565MHz, which does 230 Watts in game load, and 1100mv at 2650MHz, which peaks at 300 Watts in game load. I'm trying 1075mv at 2600MHz, but before, I used to be able to do 2625MHz, maybe a specific driver I had back then allowed the card to run at unstable settings. I feel like I'm doing something wrong because only 2600 is stable now. In Steel Nomad, I can do 2635 MHz, which is how I noticed it being bad for stability tests.
BTW, it's summer in Texas, which is not exactly good for ambient temperatures for benchmarking on an air cooler. That's why I came here to see how to undervolt.
Wtf, now 1075mv isn't stable at 2600Mhz in my second run of TimeSpy. What's going on? I tried 1062mv at 2575Mhz since 1050/2565 was stable, but it crashed at the satellite. Am I stuck with just 2.5Ghz? Even your card can do 2560 at 1025mv.
1125mv is too power-hungry; I did 1100mv at 2650MHz, but I can't tell if it's fully stable.
Wtf, now 1075mv isn't stable at 2600Mhz in my second run of TimeSpy. What's going on? I tried 1062mv at 2575Mhz since 1050/2565 was stable, but it crashed at the satellite. Am I stuck with just 2.5Ghz? Even your card can do 2560 at 1025mv.
1125mv is too power-hungry; I did 1100mv at 2650MHz, but I can't tell if it's fully stable.
24/7 gaming is important
but
just for fun
He wants a good Number in the HoF. therefore a short run with a extra special setting, to see what the chip can do
in TEXAS with the high Temps not so easy
@GerryB Is there an easier way to find Fmax for a specific voltage instead of running TimeSpy, waiting for a crash, then doing it again and again until I can find the right clock speed?
Why do you test with TS? Do you really 24/7 play with 281W ?
this takes more Watt then Games, and in Games(x) may be, goes 10MHz more at my 6800, ... like TSE
and
is also very different to Nomad light or Port Royal
OK,
TSE takes more Watt then TS, but is easier with MHz/V stable.
(x) as Game may be, Hellblade II is not bad as Test (this game is available in gamepass)
it´s possible, that in real games not so much UV runs, ... first check a little bit more Volt in AdrenalinPaneel
and second less maxClock
I just finished Hellblade 2 last night at 4K High Settings TSR "Ulta Quality" (77%). I was getting like 35fps at the lowest, but usually around 50/60fps, which is my TV's max refresh rate; at 1440p UW, I was getting like 70. Using 1050mv at 2565MHz, which was stable throughout the throughout the whole playthrough using around 230–250 Watts.
GerryB schrieb:
Why do you test with TS? Do you really 24/7 play with 281W ?
I would prefer a configuration that required 260–270 Watts and gave me 2.6GHz, but I'll need to keep dreaming for that. TimeSpy is the most demanding test I know of, so usually if it passes, it's like an 80% probability of being stable. I used to use Superposition because it was way shorter than TimeSpy, but passing Superposition gives me like a 25% probability of the card being stable.
295 Watts :/ barely an increase over my 1050mv/2565Mhz setting.
What does it mean if the intro sequence skips midway through when the astronaut shoots the dragon thing?
may be, a fast Clock change, too fast with the min Volt or need more minClock, to be stable
I don´t know, because I bought the 3dMark full.
(better temps with only GT1+2)
