![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Post Your NEW 3DMark Sky Driver Benchmark Results
- Thread starter SixFootDuo
- Start date
Deacon_Jones1988
Limp Gawd
- Joined
- Dec 8, 2012
- Messages
- 159
My scores are: 24944
Gfx: 43898
Physics: 8889
Comb: 16315
Stay [H]
gamingarena23
Limp Gawd
- Joined
- Dec 7, 2006
- Messages
- 247
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
Wait... looking at Deacon_Jones1988's results... a 5 GHz octo-core AMD chip is only just-ahead of my 4.5 GHz quad-core i5?
Good thing this bench seems to be able to actually make use of 8 threads. Parallelism seems to be AMD's saving grace here.
Good thing this bench seems to be able to actually make use of 8 threads. Parallelism seems to be AMD's saving grace here.
DraginDime
[H]ard|Gawd
- Joined
- Jan 12, 2012
- Messages
- 1,464
Link to score http://www.3dmark.com/3dm/3377026
Was running my CPU at 4.6, will crank it back up to 4.8 when I get home today and see how it affects things.
I swear the first time I ran this benchmark I got a 28000-ish score, but I didn't save it, so maybe I'm just crazy?
Score 24110
GPU 36139
CPU 9697
Combined 19034
cageymaru
Fully [H]
- Joined
- Apr 10, 2003
- Messages
- 22,077
Seems like you need more cores. Time to upgrade!
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
I'm not sure how you've read the thread up to this point and reached to the conclusion that one needs additional cores to get a higher score in this benchmark
Only 645 points behind the octo-core I mentioned in that quote. OCing a bit further will mop up that difference. Sky Diver seems to like additional speed per-core just as much as it likes a pile of slower cores.
And for the purposes of this benchmark, HyperThreading appears to be just as effective as additional cores. Jumping up from a quad-core i5 to a quad-core i7 would immediately land me an extra 4000 points in the physics test.
Last edited:
jbltecnicspro
[H]F Junkie
- Joined
- Aug 18, 2006
- Messages
- 9,533
Link to score http://www.3dmark.com/3dm/3377026
Was running my CPU at 4.6, will crank it back up to 4.8 when I get home today and see how it affects things.
I swear the first time I ran this benchmark I got a 28000-ish score, but I didn't save it, so maybe I'm just crazy?
Score 24110
GPU 36139
CPU 9697
Combined 19034
Question (I don't have 3D Mark so forgive the ignorance), but how is there a Score and a Combined score? Shouldn't the combined score be the "Score?" How does the program differentiate?
jbltecnicspro
[H]F Junkie
- Joined
- Aug 18, 2006
- Messages
- 9,533
I'm not sure how you've read the thread up to this point and reached to the conclusion that one needs additional cores to get a higher score in this benchmark
Only 645 points behind the octo-core I mentioned in that quote. OCing a bit further will mop up that difference. Sky Diver seems to like additional speed per-core just as much as it likes a pile of slower cores.
And for the purposes of this benchmark, HyperThreading appears to be just as effective as additional cores. Jumping up from a quad-core i5 to a quad-core i7 would immediately land me an extra 4000 points in the physics test.
Call me stupid, but I read his post as an attempt at humor and sarcasm - not a suggestion to upgrade...
cageymaru
Fully [H]
- Joined
- Apr 10, 2003
- Messages
- 22,077
Call me stupid, but I read his post as an attempt at humor and sarcasm - not a suggestion to upgrade...
Yes, it was intended to be humorous and sarcastic.
Also it is bad form in threads where the OP asks to post your score on a benchmark, to trash talk another forum user's PC based on said benchmark results. It discourages other forum users from posting their scores for fearing of being goaded and made fun of. [H]ardocp prides itself on real world testing, and even though these threads are about a static benchmark, the information inside is still useful. But Unknown-One doesn't have the ability to recognize basic etiquette so I figured I'd sarcastically tell him to upgrade his PC.I know that he could overclock his CPU and get a better score. Nobody else in the thread made fun of another user's PC choice but him. I could see if someone had said, Hey Unknown-One! Why do you have the lowest CPU score so far?" Then his post may have been warranted. As is I just see him as a divisive person that hates everyone's choice that wasn't the same as his. He just can't wait to tell the world why you're stupid and he's brilliant.
And that's just sad.
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
Ahh come on that was a bit harsh....I didn't interpret it that way at all..please edit your post
Last edited:
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
well because there different tests...its just giving you the score form each test...thats all...if memory serves me the combined test is the last one.....but what ever the case may be it just a score form that test......if you remember one of them is called combined and it works both the card and cpu quite hard at the same time....which 3dmark doesn't normally do
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
Uh, what?
I was pointing out the the fact that this bench doesn't seem to have its head screwed-on right when scoring performance gains by cores vs. performance gained by IPC. The scoring is waaaay out of whack.
That wasn't intended as a dig, the scoring simply doesn't make a whole lot of sense. We've got an instance where 4 cores are scoring the same as 8, which would suggest that the AMD chip has roughly 1/2 the IPC per-core (which isn't true).
And then you throw Hyperthreading into the mix and there are massive gains on two otherwise-identical chips (i5 vs. i7)... so your score bloats even though most games don't see much (if any) benefit from hyper-threading.
Last edited:
cageymaru
Fully [H]
- Joined
- Apr 10, 2003
- Messages
- 22,077
It's a culmination of his recent posting history. I think ideas should be free to flow on message boards without fear of others disparaging you. I'm just discussing how trollish his recent posts have become lately. He typically starts with something like this to get a reaction so he can start a troll fest. I love debate as much as the next person. I think it's great that people think differently about what's a good course of action than others. At the end of a debate it's fine to agree to not agree. It really is!
I'll just add him to ignore.
I'll leave my post up as it's exactly how I honestly feel. If anyone feels it's unwarranted then report it and Kyle will remove it. No harm intended and it's removal won't hurt my feelings in the slightest.
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
You act like I intend for people to blow up at nothing...
Seriously, the only thing here I'm pissy about is how little sense the CPU score makes in this benchmark. It seems to hand out higher scores simply because more cores (even if they're only logical cores) are present.
Go ahead, please explain how doing nothing but enabling Hyperthreading (which generally shows a 20% performance boost in the very-best-scaling applications) leads to a 50% higher CPU score.
Or miss the point, fly off the handle, and add someone to your ignore list for no good reason... that works tooI'll just add him to ignore.
Last edited:
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
If I had to guess:
GPU Score = Disregards CPU performance (performance data weighted towards test 1 and 2)
CPU Score = Disregards GPU performance (performance data weighted towards test 3)
Combined Score = Factors-in both CPU and GPU (performance data weighted towards test 4)
And then your final score is calculated from those three things... somehow.
jbltecnicspro
[H]F Junkie
- Joined
- Aug 18, 2006
- Messages
- 9,533
Makes sense, but there are four total categories:
Score
GPU Score
CPU Score
and Combined Score...
What's the difference between the first and last item then?
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
Last line in my post covered that. Still a guess, but this is what I'm figuring:
Combined Score = results from CPU + GPU test (Test #4).
Score = results from CPU + GPU test AND results from CPU-only test AND results from GPU-only tests (Test #1, 2, 3, and 4).
Basically, "combined" refers to the fact that Test #4 is a combination CPU + GPU test (where as the other tests tend to target either the CPU or the GPU).
Last edited:
jbltecnicspro
[H]F Junkie
- Joined
- Aug 18, 2006
- Messages
- 9,533
OH! Got you. I must have misunderstood the first time. Thanks for clarifying.
Deacon_Jones1988
Limp Gawd
- Joined
- Dec 8, 2012
- Messages
- 159
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
3DMark uses CPU-based physics. If it manages to max-out your CPU, your graphics card has to sit and wait for data
I've also got a 2500k, and I'm seeing exactly the same thing. Can only calculate physics fast enough to churn out ~80 FPS in test #4.
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
Here is mine currently haha.
On my laptop (A10-5757m 2.5ghz base, 3.5ghz turbo) Fastest Amd mobile cpu out currently until the new FX steam roller chips hit market.
are you able to play games on it? wonder if bf4 would run on it lol
would be grate if mantel helped right?
dr.stevil
[H]F Junkie
- Joined
- Sep 26, 2008
- Messages
- 9,265
Everything at stock speeds. Not too shabby.
=Gunslinger=
Gawd
- Joined
- Jan 25, 2009
- Messages
- 533
Tessellation ON in CCC
Tessellation OFF in CCC
Tessellation OFF in CCC
Bluesun311
2[H]4U
- Joined
- Sep 21, 2013
- Messages
- 2,523
Wow that new AMD driver makes this shit fly.
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
double post
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
Tessellation ON in CCC
Tessellation OFF in CCC
Over 20K physics score...impressive...except when buying the cpu $1,049.99 ouch...i'll say one thing its an upgrade from what i have...soon as i can buy one or comparable xeon at $200 it be time to rebuild again
I just don't see the point in rebuilding unless i can get that kind of upgrade...truthfully it is exactly what im holding out for....thats a hell of a score
cageymaru
Fully [H]
- Joined
- Apr 10, 2003
- Messages
- 22,077
Over 20K physics score...impressive...except when buying the cpu $1,049.99 ouch...i'll say one thing its an upgrade from what i have...soon as i can buy one or comparable xeon at $200 it be time to rebuild again
Yes, see that's the type of upgrade that the PC enthusiast community has been waiting for. A few more frames, a few less watts; it just doesn't excite us. Jumping up an order of magnitude? Hell yea!
dr.stevil
[H]F Junkie
- Joined
- Sep 26, 2008
- Messages
- 9,265
Ok, so I updated to the latest AMD drivers and got worse performance, but that was also running it remotely. Just ran it in person and scores were MUCH better
Last edited:
Last edited:
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
just thought it was kinda interesting...the video card running in low power mode during the physics test......and it never even reaches 100% on the cpu......think it spiked to about 94% at most......funny how it pegs some of the cpus to 100% thru the hole test....
And the card running in low power mode...wasn't expecting that...course i guess it wouldn't matter since its really just a cpu test..
[/URL][/IMG]
And the card running in low power mode...wasn't expecting that...course i guess it wouldn't matter since its really just a cpu test..
DraginDime
[H]ard|Gawd
- Joined
- Jan 12, 2012
- Messages
- 1,464
This is really just a wonky benchmark for high end rigs. Scores seem to be all over the place. I guess it's because it's supposed to be for gaming laptops?
primetime
Supreme [H]ardness
- Joined
- Aug 17, 2005
- Messages
- 7,337
I don't personally agree with that but at the same time I think its been a very long time if ever 3dmark has ever graded the cpu power and gpu power it seems equally...Some would argue hyper threading shouldn't be relevant in coming up with a physics score...everyone is entitled to there opinion...I have turned it off to play certain games and just found it to get less performance
It seems the people getting very high scores have very powerful cpus and gpus where as some of the older 3dmark test really only tested the gpus.....is pretty much my take on it
Unknown-One
[H]F Junkie
- Joined
- Mar 5, 2005
- Messages
- 8,909
Well, most of us are running quad cores (or quad cores with hyperthreading), where as you're running a hexa-core with hyperthreading.
It's possible their physics engine isn't set up to make use of more than 12 threads, so you're right on the cusp of diminishing returns. Just a guess.
Yeah, I've re-run Sky Diver with my GPU forced to run at full boost clock all the way through, and it doesn't make any difference to the score. This bench is just really, really CPU limited.
I'm fine with Hyperthreading being included, as long as the performance gains map linearly to the point-gains.
Problem is, 3DMark seems to be just giving out free points because it sees more cores, not because it's actually running any faster because of them.
Last edited: