Unigine Heaven 4.0 Benchmark Scores

[manny167]

Here goes, my new oc'ed bechmark:

MSI GTX 970 @ 1500/1900 mhz and i5-2500k @ 4400 Mhz. A score of 1543 should move me up to 43rd postion (coincidence? 1543 -> 43 ). Sorry about the Intel HD Graphics showing up in there, Windows keeps reinstalling it after I remove it, I hope it's not a problem, if it is, let me know, I will redo the bench.

Nice! I also Overclocked my CPU to 4.4 and scored 1616, Will be Posting the SS tonight to update my score. GTX 970 is a beast for the price though

 

[manny167]

After an Hour of Oc'ing, Finally Broke the 1630 Mark, Moving up to the 30th Position

Core i5 4670K @ 4.4Ghz | GTX 970 1560/2018

 

[silentbogo]

Testing my new super-low budget rig. No OC ATM.

Xeon X5570 @ 3GHz | GTX 750Ti 1100/1350 (stock)

 

[brutlern]

After an Hour of Oc'ing, Finally Broke the 1630 Mark, Moving up to the 30th Position

Core i5 4670K @ 4.4Ghz | GTX 970 1560/2018
View attachment 61090

Very nice, you making me jelly. I wanna do 1600+ as well but for now, 1596. Stealing your 34th position (until leaderboard is updated of course).

i5-2500K @ 4.4 Ghz | MSI GTX 970 1515/1951

 

[manny167]

Very nice, you making me jelly. I wanna do 1600+ as well but for now, 1596. Stealing your 34th position (until leaderboard is updated of course).

i5-2500K @ 4.4 Ghz | MSI GTX 970 1515/1951

Thanks man, You can do it! just focus more on the Memory Clock . Oh, and run the bench without rivatuner running, that boosted my max fps quite nicely.

 

[brutlern]

Thanks man, You can do it! just focus more on the Memory Clock . Oh, and run the bench without rivatuner running, that boosted my max fps quite nicely.

I actually scored lower without RTSS, 1591 Ran it twice without RTSS, scored 1591 exactly each time. Than, I ran it with RTSS renabled, scored 1590. So I think I had something running in the background that I didnt have when I first scored 1596.

Edit: Yep, forgot to turn off the AV.

 

[manny167]

I actually scored lower without RTSS, 1591 Ran it twice without RTSS, scored 1591 exactly each time. Than, I ran it with RTSS renabled, scored 1590. So I think I had something running in the background that I didnt have when I first scored 1596.

Edit: Yep, forgot to turn off the AV.

I think this is the max score for my card, tried overclocking more but the display driver keeps crashing (unstable). Has your card reached its limit? If not, you still might break the 1600 barrier
Good luck!

 

[SkipP]

I have an FX-8150 and a GTX-970. I am curious to see how it performs.

 

[brutlern]

I think this is the max score for my card, tried overclocking more but the display driver keeps crashing (unstable). Has your card reached its limit? If not, you still might break the 1600 barrier
Good luck!

I haven't crashed yet, but running 3DMark Firestrike peaks the powerlimit to 109% (Heaven only goes up to 106%), if it reaches 110% it will start to throttle, so even if it doesn't crash, it won't be a valid overclock result (in my opinion it should not throttle to be called a "valid" result). So there's not a lot of OC headroom left for my card, at least not if I also consider 3DMark stability, but that's just a limit I set purely for myself, I like my overclocks to be 100% stable in everything, not just Heaven, but again that's just me. We will see if that last 1% can get me OVER 9000!!!.. i mean 1600

 

[manny167]

I have an FX-8150 and a GTX-970. I am curious to see how it performs.

Dude, set your settings same as the ones in the 1st Post (rules) and take a screenshot within Heaven (Also, in 1st Post)
your score is gonna be in the 1550-1600 range (with Overclock)
Good Luck!

 

[manny167]

I haven't crashed yet, but running 3DMark Firestrike peaks the powerlimit to 109% (Heaven only goes up to 106%), if it reaches 110% it will start to throttle, so even if it doesn't crash, it won't be a valid overclock result (in my opinion it should not throttle to be called a "valid" result). So there's not a lot of OC headroom left for my card, at least not if I also consider 3DMark stability, but that's just a limit I set purely for myself, I like my overclocks to be 100% stable in everything, not just Heaven, but again that's just me. We will see if that last 1% can get me OVER 9000!!!.. i mean 1600

Go hard or Go Home!

 

[Seth Collen]

Hit 3679 on my Tri-Fire R9 290x setup. It consists of an R9 295x2 and a R9 290x. Once my 1200 watt PSU gets here I'll start OC'ing but for now I'm running on a 1050 watt and it's barely enough power.

GPU Clocks:

GPU1 - 1018/1250
GPU2 - 1018/1250
GPU3 - 1050/1250

EDIT: Increased to 3679

 

[Henry Happeny]

 

[brendangreenninja]

do my scores look right????

 

[GreiverBlade]

do my scores look right????


View attachment 61183

yep for a dual 670 it looks totally alright. +/-200pts as the list show (SLI/CFX list under the single GPU)

 

[Zille]

 

[Jetster]

View attachment 61190

Read the first post in this thread

 

[MightyMission]

I must have be doing something wrong! so my 2500k was at 4.4 and the gtx 680 at 1348/3506 and this is the paltry score I got...

 

[Steevo]

Yayy performance.

 

[GreiverBlade]

this

Read the first post in this thread

for that

I must have be doing something wrong! so my 2500k was at 4.4 and the gtx 680 at 1348/3506 and this is the paltry score I got...

View attachment 61193

in case you didn't read the OP

***PRESS F12 for SCREENSHOT - Please attach a screen capture of your results for score verification.***
***Your submission will not be added if you fail to follow the rules stated below.***

Benchmark setup:

1.) 1920x1080, Fullscreen, 8x Anti-Aliasing
2.) Ultra Quality
3.) Extreme Tessellation
4.) No integrated GPU enabled, unless it's the only GPU in the build
5.) Tessellation correctly set up on AMD cards and not bypassed in CCC
6.) You must also provide correct GPU and CPU clocks
7.) Must Be a Screenshot from within Heaven to be valid (See bottom of post) Here's why.
8.) The only allowed "tweak" is overclocking. Absolutely no driver tweaks or operating system tweaks are permitted. The goal is to provide a chart that represents real world usage for people both in and outside of this community to use to compare their cards or for comparison when buying a new card.

It will ensure that we have consistent results.

 

[MightyMission]

this

for that



in case you didn't read the OP which is ... unforgivable...

I wasn't trying to get a valid! I have f12 screens from within the app once it's finished...whats the point of submitting a below par score?..help would've been more useful.

 

[GreiverBlade]

whats the point of submitting a below par score?..help would've been more useful.

you think so?

#CPUFrequencyGPUGPU ClocksScoreUser Name
Pentium G3258 4.3GHz GTX 680 1059/1552 950 offroadz
C2Q Q9550 3.4GHz GTX 680 1164/1552 910 Hugis







your result is within the 680 parameter ... CPU play nearly nothing in Heaven as you can see those 2 score are within a 40pts variance and the card are clocked just a bit differently, also Heaven report stock clock of the CPU no worries there

the only help: re run the bench and see the result variance (totally normal my 290 oscillate between 1390 and 1497 and clocked 1150/1500 all the time)

oh and i wrote:
"in case you didn't read the OP"

so you did read it no worries here too

 

[MightyMission]

I think I was agitated due to caning the cpu/ram and 680 to there limits with the cooling I have and still getting far lesser scores than a more or less equivalent setup posted here...and I still don't understand why my score was so low. the scores were heavily fluctuating between 800-950, I wanted 1000+

 

[GreiverBlade]

still getting far lesser scores than a more or less equivalent setup posted here

aherm ... the equivalent setup (aka the 2 only 680 of the list since the CPU play no role in that ) are between 910 and 950, you get 910, it's not a "far lesser" score but a "in the norm"

you want 1000+ get a 2nd hand cheap 770 (680 refresh/rebrand) the 680 is out of the race (for Heaven) but still a good card for nearly anything else, actually it's equivalent to a mid end

moreover your score line you with the 770/280X in the chart, which are the direct competitor of the 680 (well as i have said the 770 is just a "updated" 680 and the 280X is the counterpart of the 770) tho most of the 770/280X get above 1000

 

[the54thvoid]

Got my 2nd 780ti Classy from Ebay installed now (also ordered a 780 Hydro copper cooler for it).

Ran them on the stock BIOS at 1216Mhz Core and 7300Mhz effective memory. 1st card ran at 1.17v, 2nd ran at 1.16v. Really happy. Both stayed steady at 1216 through out the run. Not the highest score for dual gpu but I think the only dual gpu (not dual 'dual' cards, i.e. 2 x 7990) above me is at 1390Mhz.

Funny thing is when I removed the tubing to fit the new card (by way of a Koolance quick disconnect) the damn thing broke, pouring water all over my desk (not my pc though). So, quick disconnects? Pile of Shite. After a while in a loop they start to pick up micro particles and stick. When I get my new block I'll still use them but only on areas away from the mobo. Can't beat plain barbs/compressions for reliability.