WHATS UP WITH 3d11!!??

[dinos22]

have a good look at the results in 4way SLI and you will see why

 

 

Andre yang's run

http://3dmark.com/3dm11/652460

 

Vince's

http://3dmark.com/3dm11/1179592

 

it has all the frequencies there, you can see it's all CPU test

 

look at X also and you'll see the same thing except it has a smaller impact

Edited June 16, 2011 by dinos22

[K404]

Only reason anyone benches Heaven is because points are awarded for it. It's crap.

 

 

Uh-oh I actually kinda like Unigine, other than the over-used fly-past/ camera pans.

 

It's nice to have a couple of longer benches

[dinos22]

it's a bit too long though, would be good if it was a bit shorter and also if it actually had game tests and subtest results but i think massman was saying it wasnt possible. It's definitely not perfect but it's a start

[Massman]

Well, obviously we couldn't do too much with the Heaven BM. All we were allowed to do was make a front-end app that launches the original benchmark. No alterations (we could have cut scenes) to the BM whatsoever.

 

The panning is indeed a bit odd, but ... 3DM11 is kinda the same . Anyways, I found Unigine to be a lot more difficult to bench than 11, especially at the very edge of stability. With the 560 ti, at least, getting through with massive artifacting was doable in 11, but hells no in Unigine.

[BenchZowner]

Single GPU & dual-GPU ( not sure about the 2nd one ) is CPU-unaffected, and the fact that a multi-CPU config has a slight advantage in 4-way SLI/CF doesn't obstruct us from getting pts from the benchmark, especially since most of the points come from the single GPU race ( not many people can afford to have 4 GTX 580s anyway, let the WR points be for those who can ).

 

About Unigine and ATi.

I ran a few tests with my revived ( don't ask how, it just came back to life ) 5870 Super Overclock.

The 1021 score was obtained with the 980X @ 4G, and the card on air cooling at 965MHz GPU / 1300MHz MEM.

A few days ago, I gave it another try with the card on LN2 and a 5.2G Sandy Bridge ( same drivers, OS, etc )

at 1300MHz GPU & 1350MHz MEM it would come up with... X911, X997, X1011.

at 1350MHz GPU & 1300MHz MEM... X947, X902, X1010.

Why how, and wtf. I don't know.

All the other benchmarks were consistent and where they should be ( in respect to the CPU & GPU/MEM clocks ), so we can't blame the hardware.

[TiN]

Why we just cannot have both, Unigine is cool, and 3d11 is power-crazy-demanding on GPUs. Both are nice in own way. Just get it awarded same, top-guys anyway are very few who have best hardware in any benchmarks, no matter if it's superpi, 3d01 or 3d11.

 

11 now is just pulls out all juice what GPU can do, every single amp from it. 3d03 puts twice less loading on GPU VRMs.

Edited June 16, 2011 by TiN

[Christian Ney]

what I like with 3DM11 that is not possible with unigine is to run for exemple only one test(aka GT, aka pass for Unigine) to test our graphic card stability and clocks.

With unigine you have to run the whole bench

[dinos22]

Bill you are testing with an ATI card and also possibly a crippled card or one that's a bit "out". This benchmark is definitely not THAT inconsistent. It has a variation in score but not 10% or at least ive never seen it that crazy

[TaPaKaH]

personally, I have nothing against 3D11 as a benchmark, but I'm worried about its consequences in the team ranks.

Currently, there's a rather big disbalance between TPP global and hardware points and with the introduction of globals for 11 it's going to get worse, giving teams with marketing pimps / manufacturer employees onboard even more advantage.

[BenchZowner]

I just re-run a lot of game tests ( non-synthetic ) at 1000MHz-1350MHz like I did for the review back when the 5870 was the top dog.

Every single test matches the exact min-avg-max framerates from my initial tests.

I can match the score with any matching hardware combination on hwbot on every single benchmark ( including 3D Mark11 ).

[TiN]

Sam OCX for Pro OC leagues there is no difference, by idea?

[Smoke]

3DM2011 absolutely not interesting benchmark( No reason to get points for it...

[Christian Ney]

3DM2011 absolutely not interesting benchmark( No reason to get points for it...

 

I am sure if there is points you will find it interesting

[hokiealumnus]

Why we just cannot have both, Unigine is cool, and 3d11 is power-crazy-demanding on GPUs. Both are nice in own way. Just get it awarded same, top-guys anyway are very few who have best hardware in any benchmarks, no matter if it's superpi, 3d01 or 3d11.

 

11 now is just pulls out all juice what GPU can do, every single amp from it. 3d03 puts twice less loading on GPU VRMs.

 

What he said. I'm all for global/wr boints for 11, FWIW.

[elmor]

I'm all for globals for 11! Seems you'll have to reconsider it due to popular demand Pieter

[Vivi]

have a good look at the results in 4way SLI and you will see why

 

 

Andre yang's run

http://3dmark.com/3dm11/652460

 

Vince's

http://3dmark.com/3dm11/1179592

 

it has all the frequencies there, you can see it's all CPU test

 

look at X also and you'll see the same thing except it has a smaller impact

 

 

Yea but andre yang used old drivers. newer drivers give huge boost in 3D11 :>.

 

 

 

Lets look at the basics, DISREGARDING who will get points if its implemented.

 

It's a good bench for a GPU's to compete against eachother.

It equals out the scores between who has a good GPU, and who has an insane CPU. sure its kind of the same as heaven, but it will make the guys with insane cpu's less in the lead (so you cant just buy insane CPU and WIN hwbot). Which makes it more fair.

 

I really don't know why it's under discussion? FM benchmarks have always counted for points.. GLOBAL i mean.

 

And i dont think it's a SR2 only bench, you can compete well enough with a single CPU board, look at vantage, thats even MORE an SR2 bench and its for points.

Edited June 16, 2011 by Vivi

[K404]

Well, to be fair how many insane GPUs are in retail?

 

Best samples are, co-incidentally, in the hands of people close to the company

[Vivi]

i disagree, splave has bought all his own DCUii and lightnings. he has found like 2 1620+ cards :>

[K404]

One guy and two cards is.... one guy and 2 cards. There's CuIIs out there that clock like reference cards as well.....

 

I guess Smoke can be added to the list of guys using their own money to bin until they find something really worth competing with.

 

I probably should have started my comment with (The vast majority)

 

 

Anyway.... my point is.... your point is mute here vivi, sorry... whoever has the best hardware gets the rankings, be it CPU binning, GPU binning, or both. Having one 2D bench that only needs one supply of great hardware is neither here nor there when theres 6 that do.

Edited June 16, 2011 by K404

[dinos22]

Yea but andre yang used old drivers. newer drivers give huge boost in 3D11 :>..

 

new drivers are much better in single card and it does not affect CPU test mate, when you run 4way its a different story.

[k|ngp|n]

Dino with all respect dude...the sr2 does not have a big advantage in 11. Your comparison is really bad and is using different drivers. We BARELY beat the asus score by a hundred or two when running the same drivers 260 vs 270. Physics score has such little weight in 11, its nothing. and there are tweaks/methods to get physics score high on single cpu as well. I think some guys jsut arent running the bench right The scores goes up in quad as it does in single.

Its high mhz vs low mhz 2 cpus and it basically balances out. If you run it in quad you will see that buddy The difference is not nearly as much as vantage.

 

and besides all of that... so u vote no becuase u think the sr2 is faster? LOL! SO WHAT!!

Deal with mate! How about other manufactures make something that can compete with a faster "overclocking" motherboard :D

 

There is NO GOOD reason to not add 3d11 for world record points and global points in single/quad. It only give hwbot more to offer.

[TiN]

I guess Smoke can be added to the list of guys using their own money to bin until they find something really worth competing with.

Uh, with photos of trays of CPUs from Intel attached to scores, yea

[dinos22]

nah man im indifferent like i said, if it has points great, if it doesnt great i dont care

 

but there is a definite advantage having CPU tests and the fact andre ran an older driver highlights it even more as his GPU tests absolutely smash EVGA's score with newer drivers so what you've done is just made it up with CPU test scores pretty much.

[k|ngp|n]

nah man im indifferent like i said, if it has points great, if it doesnt great i dont care

 

but there is a definite advantage having CPU tests and the fact andre ran an older driver highlights it even more as his GPU tests absolutely smash EVGA's score with newer drivers so what you've done is just made it up with CPU test scores pretty much.

 

Your not getting it man

It's the same as it for vantage. If you run the SAME gpu clocks in quad on an sr2 that you do on a single gulftown that is clocked higher in mhz, the gpu score is ALOT higher. 11 is similar, but not as magnified. Thats all your seeing. Like I said run 11 in quad with a single cpu system and go form 260-270 drivers then you will see the difference in your comparison.

I'm not gonna argue that the physics test is higher on sr2, but the combined test is NOT its lower. The net effect of 1-2k higher physics score is barely anything in total score in 11 im tellin ya, its still very competitive between the two platforms. Its high mhz single cpu vs low mhz dual cpu. Much more balanced than vantage and look how close those scores are. Don't use my scores as compare:D Look at smoke/Andre

Edited June 17, 2011 by k|ngp|n

[dinos22]

fair enough mate

 

 

 

i loled here a little

Dont use my scores as comapare:D Look at smoke/Andre