Hitman Bench Thread

[wunkley]

My system isn't performing as good as it should in this bench. Ran the max settings as listed, at 1920 x 1080, vsync off and get:

Min: 48
Max: 74
Av: 50

This seems waaaaay too low for 780Ti's in SLI and an i7 2600k at 4.6ghz no??? (see my sig for system details)

I noted GPU usage was only about 60% per card at the most.

 

[LtMatt]

My system isn't performing as good as it should in this bench. Ran the max settings as listed, at 1920 x 1080, vsync off and get:

Min: 48
Max: 74
Av: 50

This seems waaaaay too low for 780Ti's in SLI and an i7 2600k at 4.6ghz no??? (see my sig for system details)

I noted GPU usage was only about 60% per card at the most.

What do you get with a single card?

 

[wunkley]

What do you get with a single card?

I think the hitman bench might be a bit 'odd'. Doen a few more quick tests

Single 780Ti (SLI disabled in nvidia control panel)

Min 38
Max 145
Av 47

780Ti's in SLI

Ran the benchmark 3 times in a row and got the following:

Min 48
Max 74
Av 50

Min 48
Max 163
Av 58

Min 40
Max 72
Av 58

It's all over the place it seems. One thing I noticed is when in SLI my top card never clocked above 823, whilst the bottom card was boosting to 1006.
Usage was 85% max for the top card, 55% max for the bottom.

Thoughts?

 

[LtMatt]

I think the hitman bench might be a bit 'odd'. Doen a few more quick tests

Single 780Ti (SLI disabled in nvidia control panel)

Min 38
Max 145
Av 47

780Ti's in SLI

Ran the benchmark 3 times in a row and got the following:

Min 48
Max 74
Av 50

Min 48
Max 163
Av 58

Min 40
Max 72
Av 58

It's all over the place it seems. One thing I noticed is when in SLI my top card never clocked above 823, whilst the bottom card was boosting to 1006.
Usage was 85% max for the top card, 55% max for the bottom.

Thoughts?

I think that's likely a driver issue than a bench issue mate. Otherwise we would get the same. I suspect there's also some cpu bottlenecking going on there. 780TI SLI at 1080P will need some cpu grunt.

 

[Kaapstad]

Scoreboard updated.

[*]Score 74.98 GPU nvTitan @1097/1753, CPU 3930k @4.4, whyscotty Link

 

[wunkley]

I think that's likely a driver issue than a bench issue mate. Otherwise we would get the same. I suspect there's also some cpu bottlenecking going on there. 780TI SLI at 1080P will need some cpu grunt.

Using 334.89 drivers if that helps?

With my cpu at 4.6 should I not be getting about the same score as Gregster's Titans, at least?

 

[LtMatt]

Using 334.89 drivers if that helps?

With my cpu at 4.6 should I not be getting about the same score as Gregster's Titans, at least?

No idea sorry mate. Greg has a 6 core cpu though so you're at a bit of a disadvantage there. Can you clock the cpu up further? I find at 1080P you need it with more than one card.

 

[wunkley]

No idea sorry mate. Greg has a 6 core cpu though so you're at a bit of a disadvantage there. Can you clock the cpu up further? I find at 1080P you need it with more than one card.

Ah, didn't realise that was a 6 core, bit out of the loop on the cpu front these days.

That probably accounts for the difference then. I'm not too bothered about overclocking my cpu further to get a higher score to be honest, I just want to make sure my system is working as it should.
My score seems low compared to you AMD fellas. Must be a driver thing?

Also, I think there are a ferw issues with this game engine. In game it runs absolutely flawless maxed out and at a constant 60 fps (I always play vsynced), BUT this is only if reflections are set to low. It's weird but at any other settings higher than this for reflections it sometimes chugs into the 40's. Very odd as I can't even tell a difference in visuals.

 

[LtMatt]

Ah, didn't realise that was a 6 core, bit out of the loop on the cpu front these days.

That probably accounts for the difference then. I'm not too bothered about overclocking my cpu further to get a higher score to be honest, I just want to make sure my system is working as it should.
My score seems low compared to you AMD fellas. Must be a driver thing?

Well if you do a single card bench, it will give us an idea of the level of multi gpu scaling you're getting.

 

[wunkley]

Well if you do a single card bench, it will give us an idea of the level of multi gpu scaling you're getting.

Done one already - see post #43

 

[LtMatt]

Done one already - see post #43

Looks like a driver issue then mate. 43fps at 1080P does not sound right to me.

What about if you lower MSAA or disable it?

 

[Kaapstad]

2 x GTX 690s @1160/1800

3960X @4.9

1080p

83.30 fps

Spoiler

1600p

73.05 fps

Spoiler

 

[LtMatt]

I knew you'd roll out the good old 690's before long.

 

[Kaapstad]

I knew you'd roll out the good old 690's before long.

They run good at 1600p,the only problem is at the end where the screen goes black at the end of the bench they hit the VRAM wall as they exit the bench hence the low minimum. Actually running the bench they don't go below about 60 fps.

 

[Uncle Petey]

1080p
2600k @4.6
R9 290 @1260/1625
13.12

67.89 fps

Spoiler

Having a bit of trouble with dual/triple card runs here. All I can get is 80-81fps (at stock or overclocked with 2 or 3 cards) which is the same as my two 7970's.
Other benches run fine, so I dunno what the deal is...

 

[wazza300]

You still using the up4 th board petey

 

[Uncle Petey]

No mate, sent that back and got a Z77V Premium.

 

[wazza300]

Ahh ok,I think this game scales poor with xfire,I can't even break 39fps with single 7970 1277/1700

 

[Uncle Petey]

I went from 49 to 81 at similar clocks with my 7970's, and Matt's 290's are doing alright.
Then again, Shankly's 290's are doing the same thing as mine, so...?

Greg's Titans only pulling 73 is a bit odd too.

 

[shankly1985]

I went from 49 to 81 at similar clocks with my 7970's, and Matt's 290's are doing alright.
Then again, Shankly's 290's are doing the same thing as mine, so...?

Greg's Titans only pulling 73 is a bit odd too.

Agreed something does seem strange with this Bench. Am thinking does CPU overclock play more bigger part on this? my 4.4 vs 5ghz from Matt.