Unigine Heaven 4 benchmark

[Guest2]

Nice, 441 extra points

...My slow old ass 4790k and DDR3 RAM is still quicker

 

[wookiee87]

Nice, 441 extra points

...My slow old ass 4790k and DDR3 RAM is still quicker

LOL ddr3 will always have an advantage, unless ddr4 can go above 5k stable everyday use then it'd be a different story, still happy as only 1 of 3 with AMD in the top 10, happy days

update set the max clocks i can get on my card and scored even higher , don't think i can go any higher with out risking damaging something (don't wanna push because I'm giving this card to a good friend) so below is my final 1080p run

3900x @4.4ghz all core 2080ti, @2115 core, @8125mhz ram

 

[Guest2]

LOL ddr3 will always have an advantage, unless ddr4 can go above 5k stable everyday use then it'd be a different story, still happy as only 1 of 3 with AMD in the top 10, happy days

update set the max clocks i can get on my card and scored even higher , don't think i can go any higher with out risking damaging something (don't wanna push because I'm giving this card to a good friend) so below is my final 1080p run

3900x @4.4ghz all core 2080ti, @2115 core, @8125mhz ram

Nice few extra points. I doubt your 2080Ti is staying at @2115 core if its on air.
What does it show stabalising at in MSI afterburner / other monitoring? If it was at 2115Mhz core all the time the score would be quite a bit higher. In the past I have sometimes noticed higher scores / fps in games with lower overclock. This way the card automatically boosts higher and stays at that core clock for longer with less heat. Higher clock boost initially to 2115 (or whatever) but drastically reduce down to 2000 / 1900s

 

[wookiee87]

Nice few extra points. I doubt your 2080Ti is staying at @2115 core if its on air.
What does it show stabalising at in MSI afterburner / other monitoring? If it was at 2115Mhz core all the time the score would be quite a bit higher. In the past I have sometimes noticed higher scores / fps in games with lower overclock. This way the card automatically boosts higher and stays at that core clock for longer with less heat. Higher clock boost initially to 2115 (or whatever) but drastically reduce down to 2000 / 1900s

everything's is under custom 360mm rads, i have 1 cooling down my gpu and the other is for cpu and vrm cooling , if my card goes above 47 degrees i see around 15mhz drop but normally can hold 2100-2115mhz, when my gpu on the very rare occasion tops 52 degrees i get drops down to 2020mhz, during stress tests the max i ever see is 49 and games is 45-47 degrees so i can hold those clocks quite easily

 

[Guest2]

everything's is under custom 360mm rads, i have 1 cooling down my gpu and the other is for cpu and vrm cooling , if my card goes above 47 degrees i see around 15mhz drop but normally can hold 2100-2115mhz, when my gpu on the very rare occasion tops 52 degrees i get drops down to 2020mhz, during stress tests the max i ever see is 49 and games is 45-47 degrees so i can hold those clocks quite easily

Are watercooled
Do GPUs reach a limit even when watercooling? 2040 stable I was hitting something like 86c on air. Could you not stick another 100 on the core and more volts and top it out at something like 60c? I dont start to see clock drops until about 60c

 

[wookiee87]

Are watercooled
Do GPUs reach a limit even when watercooling? 2040 stable I was hitting something like 86c on air. Could you not stick another 100 on the core and more volts and top it out at something like 60c? I dont start to see clock drops until about 60c

if i put my card back to stock it will boost up to 1890mhz and stay there, the temp of the card never breaks 42 degrees, to get the very best from these cards you need to watercool them, if i'd kept my card on air then it would have been a dog and got to 86 degrees and clocked around 1740mhz which is pants. The gains going to water are massive.

rule of thumb for Turing the cooler the better, up to 50 degrees will hold the max boost clock above and it will start to loose clock it to save the card from damage (aka too much heat)

 

[Kaapstad]

Scoreboard updated.

 

[Avathar77]

Edit(didnt do with 8xAA first time so edited with new results)
1800X @ 4GHz Palit 1080Ti Core 1987(boost) Mem 1376 MHz - default speeds, not overclocked

 

[Kaapstad]

Scoreboard updated.

 

[wookiee87]

4k run

3900x @4.45ghz, 2080 ti @2100mhz core 8150mhz ram

 

[Kaapstad]

Scoreboard updated.

 

[FireTwo]

hmmm score seems alittle low?!
expected closer to 5k any tips? guessing more gpu core or mem speed which does this benchmark pref?

https://ibb.co/R4jNKNt

 

[Boomstick777]

This an ok score for stock 2080 Ti and stock 3700X?

best free image hosting

 

[wookiee87]

This an ok score for stock 2080 Ti and stock 3700X?

best free image hosting

stock score isnt too bad at all, just be sure to set 100% gpu fan speed that way the card may hold it's max boost clock longer which will increase the score a bit, but to get the reail gains you need to push up the core and memory speeds, doing so will net a very healthy boost and play around with ram too, try and set the fastest sable timmings and again your score will go up.
cpu wise you could try setting a all core 4.3/4.4ghz, but keep any eye on temps, if they shoot up or the pc turns off then go back into the bios and set the cpu mulitplier back to auto, that means better cooling is required to again boost your overall score.

hmmm score seems alittle low?!
expected closer to 5k any tips? guessing more gpu core or mem speed which does this benchmark pref?
https://ibb.co/R4jNKNt

same as above really, gpu overclocking provides decent uplift, and if you have good fast ram then that'll help too, moniter your cpu temps too if they go over 70 degrees the tvb will drop back to a all core 4.7ghz which is fast but if you want to top the board keep the cpu as cool as possible and push a all core overclock 5.2/5.3ghz, if the cpu will allow of course, if the silicone is average then i'd say 5.0/5.1ghz is where it will crap out at but you may be lucky and have a golden chip

both tests need to be run in fullscreen mode just alter the resolution down to 1080p, 1440p or 4k

 

[Boomstick777]

stock score isnt too bad at all, just be sure to set 100% gpu fan speed that way the card may hold it's max boost clock longer which will increase the score a bit, but to get the reail gains you need to push up the core and memory speeds, doing so will net a very healthy boost and play around with ram too, try and set the fastest sable timmings and again your score will go up.
cpu wise you could try setting a all core 4.3/4.4ghz, but keep any eye on temps, if they shoot up or the pc turns off then go back into the bios and set the cpu mulitplier back to auto, that means better cooling is required to again boost your overall score.

Thanks dude, not looking to overclock just wanted to check it's all running where it should be at stock.

Installed this card today and nothing to compare too so that helps. Cheers.

 

[FireTwo]

ok 5k club! full screen mode must have helped,

https://ibb.co/18L1yyh

 

[Kaapstad]

ok 5k club! full screen mode must have helped,

https://ibb.co/18L1yyh

Have a go at 1440p if you really want to stretch your card.

1080p is just too CPU bottlenecked for a 2080 Ti

 

[justwondering]

Gigabyte RTX 2060 Super Windforce OC

1080p
Score 3060

i5 8600k @ 4.8Ghz
Drivers 451.48
GPU Clock 1590Mhz / boost 1800Mhz
Memory 1938Mhz

https://ibb.co/vXJNbmJ

 

[Kaapstad]

Scoreboard updated.

 

[C64]

rx 470 @1340/1750 that's about +20 frames on 2500k + rx 470
so the 5600xt is only +15 frames or so over a mildly overclocked rx470 ?