-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
**The New Official THIEF-Incorporating Bench mark Thread**
- Thread starter humbug
- Start date
Soldato
- Joined
- 14 Feb 2011
- Posts
- 7,490
Nvidia GTX Titan @ 1346/1925, 4770K @ 5.0Ghz
E: 1080P
Before i do anything with this;
Min FPS
Kaap on a 4.8Ghz Haswell only got 61.3 FPS.
Greg on a 5Ghz Ivy got 60.5
On a 5Ghz Haswell you got 65..6
Your way ahead of both Greg and Kaap on Similar CPU cores, and;
Avr FPS;
Way ahead of Greg on a similar GPU.
At this point it doesn't look right to me.
Soldato
- Joined
- 14 Feb 2011
- Posts
- 7,490
Gregs on Sandy bridge E, not Ivy and at lower clocks on both Core & Memory with his Titan.
Last edited:
Soldato
- Joined
- 28 May 2007
- Posts
- 10,147
It's really does not look in line tbh. Only 40mhz more than Greg on the core should not relate to 13-14% more performance. It's obvious there's no Cpu bottleneck in both systems. There's not enough Nvidia results to really compare though as Greg could be scoring on the low side. In the last thread there was some weird results appearing as well.
Last edited:
It doesn't explain how he's 8% faster than Kaap on the cores.
And 13% Avr than Greg on the GPU and 18% faster on the Max FPS. those are huge differences.
Kaap has Ivy, Greg has SB, both have considerably lower single core performance as haswells, especially considering mine was using x50 cache/uncore.
Besides, min fps is mostly luck anyway, it's usually the first frame of the bench that gives you the min numbers.
And thief loves single thread performance at low resolutions. My GPU was clocked 5% higher than Gregs (AB lies a bit, it's actually ~1360MHz), I had shader cache off in the NV control panel because while it can speed up loading times (note it's not an image quality option, just for loading times) it sometimes lowers performance by a couple of fps, and my CPU is significantly faster as far as single thread goes. And my memory is also clocked higher.
5% higher GPU clock
~2.5% higher GPU memory clock
~10-20% higher single thread performance (depends on the application I don't know the exact difference in UE3)
SC Off
And a custom bios that removes a couple of power saving features from use.
bleh.
Besides, min fps is mostly luck anyway, it's usually the first frame of the bench that gives you the min numbers.
And thief loves single thread performance at low resolutions. My GPU was clocked 5% higher than Gregs (AB lies a bit, it's actually ~1360MHz), I had shader cache off in the NV control panel because while it can speed up loading times (note it's not an image quality option, just for loading times) it sometimes lowers performance by a couple of fps, and my CPU is significantly faster as far as single thread goes. And my memory is also clocked higher.
5% higher GPU clock
~2.5% higher GPU memory clock
~10-20% higher single thread performance (depends on the application I don't know the exact difference in UE3)
SC Off
And a custom bios that removes a couple of power saving features from use.
bleh.
Kaaps is on IVY-e and Greg is on SB-e
Soldato
- Joined
- 28 May 2007
- Posts
- 10,147
Using a single card would it really matter what they are using at those clocks. Surely any one of those cpu's are more than enough for it not to matter.
Caporegime
Something seems amiss there. Perhaps scotty can run with one of his titans to compare. If i trust anyone, its that man from Wigan. 
Soldato
- Joined
- 14 Feb 2011
- Posts
- 7,490
His scores look genuine to me judging by the equipment he's using and clocks on both Cpu & Gpu. I don't see anything fishy going on.
Caporegime
Would love to oblige - alas my PSU on that rig is DEAD
Oh dear.
Time for a superflower then.
Soldato
- Joined
- 28 May 2007
- Posts
- 10,147
To be honest the last thief bench thread had some really weird results mainly on the NV side.
I found these results looked off for obvious reason's.
1. MjFrosty-780ti@1315/1965, [email protected]
2. khemist-780Ti@1410/1905, [email protected]
3. gregster-Titan@1267/3767, [email protected]
5. whyscotty-Titan@1241/1802, [email protected]
Note how the 6 core cpu's seem to do better even if they are slower core per core.
I found these results looked off for obvious reason's.
1. MjFrosty-780ti@1315/1965, [email protected]
2. khemist-780Ti@1410/1905, [email protected]
3. gregster-Titan@1267/3767, [email protected]
5. whyscotty-Titan@1241/1802, [email protected]
Note how the 6 core cpu's seem to do better even if they are slower core per core.
Last edited:
Soldato
- Joined
- 14 Feb 2011
- Posts
- 7,490
Something seems amiss there. Perhaps scotty can run with one of his titans to compare. If i trust anyone, its that man from Wigan.
To compare he would need to be able to run both his 4770k at 5GHz and his Titan at the same clocks.
Caporegime
I'm sure even similar clocks would be fine. Typically a slight gpu clock speed difference does not equal a massive fps difference, especially on single cards.
This game seems to favour 6 cores, but will need a bigger sample size i guess.
Soldato
- Joined
- 14 Feb 2011
- Posts
- 7,490
Other factors you also have to consider are system tuning... Ie ram speeds and timings/drivers used etc. Who knows what he's using. Taking everything into account his score will be correct i'd say.