Author Topic: GATT Benchmarking Development forum (Read 40074 times)

TriniXaeno · « **Reply #40 on:** June 17, 2009, 10:39:31 AM »

Going with Kraeoss, let a tie be a tie. (exact score)

W1ntry, captain's point is that we are even less likely to have scores so close. It makes the tie argument even less compelling. Won't you agree?

If a tie isn't necessary at a 1 point difference then it is even less applicable with a greater delta, say a 50 point difference between the two.

Agreed also @ expanding the list to a top 50. (looks forward to being back in the rankings, lol)

w1n, we don't really have to choose. You can have your cake and eat it too.

Let the casual benchmark list remain that (simple and easy for casual gatt members to participate)

And implement the more strict regimen (10 pass 3dmark runs, superpi, tie within 100 point delta, etc) for the hard core list.

Carigamers · « **Reply #40 on:** June 17, 2009, 10:39:31 AM »

W1nTry · « **Reply #41 on:** June 17, 2009, 10:45:45 AM »

Well I am for a more strict version IF men were willing to take up the gauntlet I laid out for all the benchmarks I myself have done. But no takers. I don't think a 50 list is a good idea, however I am for a retrospect listing, perhaps we take a time capture every year to show who was top dog for that year and an ALL time list with everyone's score (no placing) just for historical reference of ppls scores over the years. The likeliness of a EXACT tie though is RIDICULOUS and as I said, the score you get today will NOT be the same tomorrow or the same after you load a few games or apps. So I still think a delta is a good idea for a more level playing field.

Kraeoss · « **Reply #42 on:** June 17, 2009, 10:54:28 AM »

If u lose you lose if u win you win simple as that, even if it's by 1 or 99 pts. unless u wanna get bragging rights then spend yuh money (prime candidate = Arcman,Captain Awesome, And the others in the top 5) hehe

W1nTry · « **Reply #43 on:** June 17, 2009, 11:06:15 AM »

Quote from: Kraeoss on June 17, 2009, 10:54:28 AM

If u lose you lose if u win you win simple as that, even if it's by 1 or 99 pts. unless u wanna get bragging rights then spend yuh money (prime candidate = Arcman,Captain Awesome, And the others in the top 5) hehe

Hmm yuh almost sound bitter XD you're almost off the list aren't ya XD

Kraeoss · « **Reply #44 on:** June 17, 2009, 11:12:12 AM »

well i'm off and i know i be bested thas y i saying if you lose yuh lose lol if yuh win yuh win lol nothing bitter there

Carigamers · « **Reply #44 on:** June 17, 2009, 11:12:12 AM »

Netizen1 · « **Reply #45 on:** June 17, 2009, 11:34:36 AM »

How about we ask the ppl that posted scores to give an average of what their "skew" is like...

How bout it Arc et al, if you ran 3dMark a couple more times, what's the fluctuation in score like? Is W1nTry's 100 pts close to accurate here? or is it way less, like maybe 50 or so?

W1nTry · « **Reply #46 on:** June 17, 2009, 12:15:31 PM »

I concur, I will attempt to over the period of maybe a week run 3-pass benchmarks and accumulate the differences in skew as well.

Arcmanov · « **Reply #47 on:** June 17, 2009, 02:16:16 PM »

* Noted *

Netizen1 · « **Reply #48 on:** June 17, 2009, 03:32:04 PM »

No new "rules" have been cast in stone as yet OC

So doh worry yuh head about it, we just trying to make the process a lil more fair (?) I guess

TriniXaeno · « **Reply #49 on:** June 17, 2009, 04:14:36 PM »

lol, brings back memories of the "man borrowing parts to get a kick ass score" argument.

Spazosaurus · « **Reply #50 on:** June 17, 2009, 04:51:38 PM »

This is exactly the kind of complication that is neither necessary or needed. Maybe we should leave it as "more is more, less is less" and done because I see no end to the variables.

TriniXaeno · « **Reply #51 on:** June 17, 2009, 11:43:50 PM »

gets my vote

W1nTry · « **Reply #52 on:** June 18, 2009, 09:24:40 AM »

Hmm, we not even asking yall to do this testing, men like mehself and arc would do it (besides i'd like to see for my own personal info what the skew is). And as was suggested it could be done for a more intensive benchmark standings.

MessiaaH · « **Reply #53 on:** April 26, 2010, 09:58:04 AM »

just a note to Gatt and all phsyx users out there, phsyx enhanced 3DMark scores are not permitted in official world scale rankings and considered cheating.
Also physx does not really increase game performance and artificially inflates 3dmark scores.

Take a look here for some of d world scores:

http://hwbot.org/rankings/benchmark/3dmark_vantage_-_performance?tab=rankings

As such i think the ranking on gatt should follow suit, and include non-phsyx benchmarks for official compition.

What you guys think?

Arcmanov · « **Reply #54 on:** April 26, 2010, 11:06:54 AM »

Totally agree.

phoenix31tt · « **Reply #55 on:** April 26, 2010, 11:25:02 AM »

agreed...

follow d standards

Netizen1 · « **Reply #56 on:** April 26, 2010, 11:39:00 AM »

more haters...

just as I was about to add in a PhysX card to the system... whatever yes... I think I'll never make the rankings with them sorta rules and my current salary

firstchoicett · « **Reply #57 on:** April 26, 2010, 12:38:14 PM »

Quote from: MessiaaH on April 26, 2010, 09:58:04 AM

just a note to Gatt and all phsyx users out there, phsyx enhanced 3DMark scores are not permitted in official world scale rankings and considered cheating.
Also physx does not really increase game performance and artificially inflates 3dmark scores.

Take a look here for some of d world scores:

http://hwbot.org/rankings/benchmark/3dmark_vantage_-_performance?tab=rankings

As such i think the ranking on gatt should follow suit, and include non-phsyx benchmarks for official compition.

What you guys think?

this forums did not stated any rules on this, so you got beat from my eyes by miles. when i disable physx my score was a little above 30,000 points. i will have that shot up for you later when i get some time., but for now just stop making excauses to remove my score. because it will not work... if the forums had stated this rules before then i would of understand but if they state it after, thats unfair to me and other users.

anyway. doing some overclocking on the crossfire system, so i will have some results on that later.

MessiaaH · « **Reply #58 on:** April 26, 2010, 01:13:06 PM »

so wait, you have 2 high end personal systems? CF 5870 and Qaud 295 both you personal rigs? lol. Like arcman said, i in d wrong line of work.

Anyways,
The whole purpose of 3DMark and d Rankings is to judge how well a computer runs games, Higher 3dmark scores means better game/machine performance. In d case of phsyx dat is not so, it does nothing for game/machine performance, it juss raising yuh score in 3dmark. How hard it is to buy ah cheap gtx 260 tuh increase score by 5000? lol, i could do dat today self, but phsyx defeats d whole purpose of what d benchmark is meant to reflect, so yuh unfair treatment statement is balls.

3dmark programs doesnt even allow u to post ur results online if u run it with physx, and for ah man with ah high end system like urs, u shouldnt be complaining (firstchoicett).

And having a quad sli system is overkill unless u runing 30" monitor or some kinda eyefinity action. and 2 5970 is rel overkill, but if is war yuh want is war yuh go get, and ah secodn 5970 could be on d way

, but as is, ur CF system shud be able to beat mine, lets see who d better overclocker is. And if u claim d 295 quad sli system, and d cf 5870 are both ur own personal rigs u buy fuh yuhself, dat will decide weather i lash d second 5970 or not

Leh meh hear yuh. Fellers dis man calling meh out, Gustav locked and loaded.

SPK · « **Reply #59 on:** April 26, 2010, 01:20:46 PM »

Quote from: MessiaaH on April 26, 2010, 01:13:06 PM

Fellers dis man calling meh out, Gustav locked and loaded.

*gets popcorn*

This should be interesting to say the least

Carigamers · « **Reply #59 on:** April 26, 2010, 01:20:46 PM »

Author Topic: GATT Benchmarking Development forum (Read 40074 times)

Carigamers

Carigamers

Carigamers

ShoutBox