1080 Ti and Octane
Forum rules
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Please add your OS and Hardware Configuration in your signature, it makes it easier for us to help you analyze problems. Example: Win 7 64 | Geforce GTX680 | i7 3770 | 16GB
Hi guys,
Just stuffed a 1080 Ti founder next to a Maxwell Titan Z and I am comparising them.
The first test was with the Octane Benchmark Trench 3.05.3.
I Have rendered the scene to 4000 maxsamples and the and the Ti just beats the Titan Z with 43 versus 44 seconds
Later I will post my findings for perfromance difference in my own projects as a workflow comparising,
because for me that is all that counts in practical way.
cheers,
Just stuffed a 1080 Ti founder next to a Maxwell Titan Z and I am comparising them.
The first test was with the Octane Benchmark Trench 3.05.3.
I Have rendered the scene to 4000 maxsamples and the and the Ti just beats the Titan Z with 43 versus 44 seconds

Later I will post my findings for perfromance difference in my own projects as a workflow comparising,
because for me that is all that counts in practical way.
cheers,
4090+3089ti & Quad 1080ti
ArchiCAD25, ofcourse Octane & OR-ArchiCAD plugin (love it)
http://www.tapperworks.com
http://www.facebook.com/pages/TAPPERWOR ... 9851341126
http://www.youtube.com/user/Tapperworks/videos
Hi Guys,rappet wrote:Hi guys,
Later I will post my findings for perfromance difference in my own projects as a workflow comparising,
because for me that is all that counts in practical way.
cheers,
I am not sure to be happy and or dissapointed


In three different project scenes (archviz exterior, object render (PMC) and interior panoramic render (DL and PMC) I have compared the Titan Z with the 1080Ti.
the results are -7%, -23%, -4%, -8% less performance for 108Ti compared to TitanZ..
I will check and do another test on the 23%, because that looks off.
to be continued
edited/added:
I did the test again with an Aquarium Object scecen.
Rendertimes Titan Z versus 1080Ti:
PMC 38/51 = -25%
DL 37/41 = -10%
PT 59/69 = -14%
Note about calculation:
the first example with 38/51 seconds ... is 0,745, then you might say difference is 25%
but if fact we also might say 51/38 = 1,34 and then you might say difference is 34%
Then the 1080Ti is 34% slower and the Titan Z 25% faster, right?
Is that tomatoes and tomatos?
It makes a big difference in saying 34% or 25%!
So in fact my first stated numbers ( -7%, -23%, -4%, -8%) are even bigger in that perspective.
cheers
4090+3089ti & Quad 1080ti
ArchiCAD25, ofcourse Octane & OR-ArchiCAD plugin (love it)
http://www.tapperworks.com
http://www.facebook.com/pages/TAPPERWOR ... 9851341126
http://www.youtube.com/user/Tapperworks/videos
Thanks for information Rappet ;)
I though it [1080Ti] should behave as TitanZ. Curious why there is difference on some scenes though, but I guess that was due to architectural differences.
If You happen to run OctanBench, could You please post those print-screen of that with final results + that extra table that pops in the end? (curious TitanZ vs 1080ti)
I though it [1080Ti] should behave as TitanZ. Curious why there is difference on some scenes though, but I guess that was due to architectural differences.
If You happen to run OctanBench, could You please post those print-screen of that with final results + that extra table that pops in the end? (curious TitanZ vs 1080ti)
The OctaneBench is 2.17 and does not support Pascal, right?glimpse wrote:Thanks for information Rappet![]()
I though it [1080Ti] should behave as TitanZ. Curious why there is difference on some scenes though, but I guess that was due to architectural differences.
If You happen to run OctanBench, could You please post those print-screen of that with final results + that extra table that pops in the end? (curious TitanZ vs 1080ti)
or do you mean another OctanBench...? I have read about an unofficial or 'cracked' OctaneBench somewhere around this forum.
4090+3089ti & Quad 1080ti
ArchiCAD25, ofcourse Octane & OR-ArchiCAD plugin (love it)
http://www.tapperworks.com
http://www.facebook.com/pages/TAPPERWOR ... 9851341126
http://www.youtube.com/user/Tapperworks/videos
Yeah, 2.17 does not support Pascal based cards, but You can run this one:rappet wrote: The OctaneBench is 2.17 and does not support Pascal, right?
or do you mean another OctanBench...? I have read about an unofficial or 'cracked' OctaneBench somewhere around this forum.
viewtopic.php?f=9&t=56108&p=289082&hilit=1080#p287699
would be very thankful if You could Run TitanZ; 1080TI & then both =)
strange thing, v3 behaves a bit different & thus when I'm running with pair of cards I always get lower than separately.
Hi glimpse,glimpse wrote:Yeah, 2.17 does not support Pascal based cards, but You can run this one:rappet wrote: The OctaneBench is 2.17 and does not support Pascal, right?
or do you mean another OctanBench...? I have read about an unofficial or 'cracked' OctaneBench somewhere around this forum.
viewtopic.php?f=9&t=56108&p=289082&hilit=1080#p287699
would be very thankful if You could Run TitanZ; 1080TI & then both =)
strange thing, v3 behaves a bit different & thus when I'm running with pair of cards I always get lower than separately.
TitanZ: 191.45 / 1080Ti: 165.45
1080 Ti ran at 1822 mhz, no overclock, 74 Celsius.
The precious TitanZ is still king with +15%

cheers,
4090+3089ti & Quad 1080ti
ArchiCAD25, ofcourse Octane & OR-ArchiCAD plugin (love it)
http://www.tapperworks.com
http://www.facebook.com/pages/TAPPERWOR ... 9851341126
http://www.youtube.com/user/Tapperworks/videos
nice, thank You! ;) not sure if I posted this, but mentioned somewhere, that technically You should be able to get about 200 & in some cases even more if You would bump core clock +200 & memory +500. With curve adjusted or fan put in a fixed speed mode (70 to 100%), this card would still keep good temps. As far as I know there's not so much of possibilities to push TitanZ on air =)rappet wrote: Hi glimpse,
TitanZ: 191.45 / 1080Ti: 165.45
1080 Ti ran at 1822 mhz, no overclock, 74 Celsius.
The precious TitanZ is still king with +15% :lol:
cheers,
So yeah, TZ might still seem faster on paper, however in reality with some tweaks these are more or less the same in terms of performance ;)
Thanks again for this information, J!
Sinc.
tom
Hey T,
I did the overclocking +200 GPU Clock and +500 MEM Clock on the 1080 TI.
that gave 7% more performance: from default 165 to 177 on the Benchmark.
the fan was on auto, and the temp was max 75 celsius
And I also did the Titan Z and 1080Ti (default) together; result is 353
Compare adding numbers was 191+165= 356, so that is about the same.
The temp on the 1080TI now also was >80 celsius,
fanspeed of the 1080Ti was about 2000 rpm.
Gpu clock was 2012 mhz
Then I also did the Titan Z and 1080Ti overclocked together; result is 365
Compare adding numbers was 191+177= 368, so that is about the same.
The temp on the 1080TI now also was >82 celsius,
fanspeed of the 1080Ti was >2350 rpm.
Gpu clock was jumping between 1900-2000 mhz
the Titan Z is below the 1080Ti, so the heat spread form the Z surely influences behaviour of 1080Ti.
greetz,
Edited'/added 2:
But when doing an interior rendering (project) it is about same perfromance!!!
99 sec for 1080ti versus 95 sec for TZ
Edited/added:
I did overclocking on the 1080 Ti seperately on Mem and GPU and it looks they rare not related in performance increasing:
+200 GPU Clock: 175 (+7)
+500 MEM Clock: 172 (+5)
(together they ad up from default 165 to 177)
I did the overclocking +200 GPU Clock and +500 MEM Clock on the 1080 TI.
that gave 7% more performance: from default 165 to 177 on the Benchmark.
the fan was on auto, and the temp was max 75 celsius
And I also did the Titan Z and 1080Ti (default) together; result is 353
Compare adding numbers was 191+165= 356, so that is about the same.
The temp on the 1080TI now also was >80 celsius,
fanspeed of the 1080Ti was about 2000 rpm.
Gpu clock was 2012 mhz
Then I also did the Titan Z and 1080Ti overclocked together; result is 365
Compare adding numbers was 191+177= 368, so that is about the same.
The temp on the 1080TI now also was >82 celsius,
fanspeed of the 1080Ti was >2350 rpm.
Gpu clock was jumping between 1900-2000 mhz
the Titan Z is below the 1080Ti, so the heat spread form the Z surely influences behaviour of 1080Ti.
greetz,
Edited'/added 2:
mmm.... ? with this OC it does not come close to the 191 of the TitanZ in this Benchmarkglimpse wrote: So yeah, TZ might still seem faster on paper, however in reality with some tweaks these are more or less the same in terms of performance
But when doing an interior rendering (project) it is about same perfromance!!!
99 sec for 1080ti versus 95 sec for TZ
Edited/added:
I did overclocking on the 1080 Ti seperately on Mem and GPU and it looks they rare not related in performance increasing:
+200 GPU Clock: 175 (+7)
+500 MEM Clock: 172 (+5)
(together they ad up from default 165 to 177)
4090+3089ti & Quad 1080ti
ArchiCAD25, ofcourse Octane & OR-ArchiCAD plugin (love it)
http://www.tapperworks.com
http://www.facebook.com/pages/TAPPERWOR ... 9851341126
http://www.youtube.com/user/Tapperworks/videos