Sleeping Dogs Bench Thread

Silent_Scone · 29 Apr 2014 at 15:20

LtMatt said:
Never mind him, what about me?

You've been beaten by a couple of air cooled reference 290's. That's just not cricket. How long do you give me?

Lol that run was stock

Edit: wait what score are you even talking about? Lol

Modified tess?

LtMatt · 29 Apr 2014 at 15:49

MjFrosty said:
Lol that run was stock

Edit: wait what score are you even talking about? Lol

Modified tess?

Dual gpu 1080P. Reckon i might have 5mhz more on the core in the tank if i need it. If it was colder the score would be higher i think.

No artifacts on these runs. Once i go past 1270 i do get a few though once temps hit the 80's. The problem i have is keeping temps in check. (shock horror

)

My PSU passed the Sleeping Dogs burn in test at least. 1270 Watts pulled at one point from my Superflower plat 1000W.

[email protected]
290 Crossfire at 1250/1625
14.4 WHQL
1440P

Uncle Petey · 29 Apr 2014 at 15:57

1270w? Christ... I'm afraid to put the watt meter on the trifire rig

Silent_Scone · 29 Apr 2014 at 16:01

Nice although in vain as I'll have better scaling now and can run 1300 core over two of mine lol. Or I could just disable world density as judging by some of the results some of you are : p

LtMatt · 29 Apr 2014 at 16:02

Uncle Petey said:
1270w? Christ... I'm afraid to put the watt meter on the trifire rig

That final scene made it reach that wattage. Crazy innit.

These reference cards can take it though. Its the only time i let them stretch their legs. Now they go back to -0.037v and -0.100mv @1050/1250 for 24/7.

LtMatt · 29 Apr 2014 at 16:05

MjFrosty said:
Nice although in vain as I'll have better scaling now and can run 1300 core over two of mine lol. Or I could just disable world density as judging by some of the results some of you are : p

Probably but still.

Kaap will get a performance boost if he runs two cards on 14.4 WHQL as well, so i expect he'll pull clear of me when he does. I'm bottlenecked at 1080P, need a 6 core then i could improve my score even more. Gpu usage did drop a lot during the 1080P run.

I didn't disable anything, so hope you're not saying i did. That's a legitimate run using the Extreme preset.

Silent_Scone · 29 Apr 2014 at 16:17

No, for the clocks used in comparison to your last run and drivers there doesn't look to be anything out of place lol. I was speaking about some of the other ones. should have kept the Lightning in, would have blazed Kaaps single card run

LtMatt · 29 Apr 2014 at 16:21

MjFrosty said:
No, for the clocks used in comparison to your last run and drivers there doesn't look to be anything out of place lol. I was speaking about some of the other ones. should have kept the Lightning in, would have blazed Kaaps single card run

Lighting would be great for this bench i think because of its cooler and the Samsung memory. How i wish i could send my cards off somewhere and have Samsung memory fitted. Bet id see a 5% or more performance boost.

Have you sent it back already?

Silent_Scone · 29 Apr 2014 at 16:22

No i'm half tempted to keep it and still get the block...part of me knows something the other part doesn't yet I think lol.

Kaapstad · 29 Apr 2014 at 19:47

Scoreboard updated.

LtMatt · 29 Apr 2014 at 19:48

Kaapstad said:
Scoreboard updated.

Kaap you should run dual card 1080P i got a bit of a perf boost from 14.4.

Silent_Scone · 29 Apr 2014 at 20:14

Quite a nice boost with these too

I'm having to run it three times and double check setting. might have to include a video for proof unless someone else runs this on the new NV magic drivers for a comparison lol

Kaap?

Uncle Petey · 29 Apr 2014 at 20:16

Bloomin eck! I was just having a dabble with benching this and looking at power consumption and here's what happens:

2600K @4.7 1.344v

7970's @1200/1600 1.244v = 655w peak
7970's @1260/1900 1.3v = 780w peak

:eek:

Either way, the game's playing funny buggers with me as it decides to CTD with no error message or nowt at the end of the first stage even at stock clocks... Ah, whatever.

Kaapstad · 29 Apr 2014 at 20:20

Uncle Petey said:
Bloomin eck! I was just having a dabble with benching this and looking at power consumption and here's what happens:

2600K @4.7 1.344v

7970's @1200/1600 1.244v = 655w peak
7970's @1260/1900 1.3v = 780w peak

Either way, the game's playing funny buggers with me as it decides to CTD with no error message or nowt at the end of the first stage even at stock clocks... Ah, whatever.

Have you got frame pacing on or off.

Uncle Petey · 29 Apr 2014 at 20:21

Off of course. It's the only way to bench

LtMatt · 29 Apr 2014 at 20:22

MjFrosty said:
Quite a nice boost with these too

I'm having to run it three times and double check setting. might have to include a video for proof unless someone else runs this on the new NV magic drivers for a comparison lol

Kaap?

Sleeping Dogs was one of the games that got a big perf boost on 337 so at 1080P id expect you to beat mine and Kaaps scores fairly easily. Might be closer at 1440P/1600P though. AMD cards are just too bottlenecked at 1080P but 337 improved this area for Nvidia.

Kaapstad · 29 Apr 2014 at 20:23

Uncle Petey said:
Off of course. It's the only way to bench

Sometimes it can be unstable when benching with it turned off.

Silent_Scone · 29 Apr 2014 at 20:23

I expected a boost, but

Can you re-run this on your Titans Kapitan for a comparison?

4960@ 4.5Ghz

Ref 780Ti @ 980/1750 (Stock bios)

Running a tight CL8-10-10-28 2133 for 24/7 at the mo so that might help in this bench, not tried it enough

Uncle Petey · 29 Apr 2014 at 20:29

Kaapstad said:
Sometimes it can be unstable when benching with it turned off.

Yeah, true. Maybe there's something going in 14.4, as having it switched off worked fine with the 13.11 b9.5's I used to post results in here before...

Just tried it with FP on and it still does it.

735w @1250/1800 1.288v btw.

Silent_Scone · 29 Apr 2014 at 20:36

Wow. Guessing there is a fair bottleneck in this game. Still stock GPUS

4960@ 4.5

Tri SLi 780Ti 980/1750