Hello TG Forums...
Last years Forum Wars was fun but this year, it's gonna be a blast.
5 player teams, 5 teams max per forum....
There's gonna be some serious competition this year as I expect at least 20 or More forums to join in on the fun.... Each forum will have an A Team comprised of it's best 5 players, those 5 will play against the other forums best 5 for the title.
Teams 2-5 will play against each other in a winner takes all for the Tweaked Geek Team Title.
So thats 2 listings... The Top Dawgs and The Tweaked Geeks.... and you know there will be some very good systems on the Tweaked teams too.
Here's link to check out what's happening and what the rulz are....
Note.... there might be some minor rulz changes in the next month, but will be finalized by the end of march....
Well... I do all the admin work on Forum Wars.
this is my second Forum wars, I run quite a few Benching contests at our forums and at PC Pitstop forums...
Last Years FW2005 was a great success but the team size was a little too large to have competitive teams, this year will be different.
I am currently lining up Sponsers for the event and talking with about 30 forums who checked us out last year but didn't enter.
We currently have 6 forums entered and will only get larger as the contest approches. We hope you join us.
It will be fun...
First... let me say that I am Not trying to solicit members, this is YOUR Forum against the other Forums, I only provide a central location to place scores and for all the admins,mods and cops to meet for contest updates and stuff.
There will be 5 Teams Allowed with 5 players per Team (25 Total)
"A Team" will be the Top 5 Scoring Players.
"Teams 2-5" will be comprized of the rest of the Forums Players.
Each Forums Best 5 Players will go head to head against the other Forums Best
For the Forum Wars 2006 Top Dawg Title.
Teams 2-5 will Face off for the Tweaked Geek Crown.
Begining June 2nd.... All Players must submit a preliminary Score... I reccommend stock,
Just dont post your Best!!!
REASON... Every forum is guaranteed a score from every player entered... if a player doesnt participate after we start, their prelim score counts toward your total... So having reliable players is a must this time around.
All Participating Forums Must Provide a few Members to Police (verify and enter) the Posted Scores entered. These Forum Police (cops) Will have their Handle (Internet name) registered. They will also be global, meaning ... they can go to any forum to check any score... Cool (They do not have to play & Players can be cops also). Admins/Mods (Team Leaders) are Judges and have to make the local calls, we will have a private area called Forum Wars Admin set up soon where the cops and Admins/Mods (Team Leaders) can discuss things, tricks, how to's,Voting Polls, practice & the latest about whats going on. We will use an online database to enter the posted scores after they have been verified by a Cop or Team Leader. After that I will transfer them to my Excel spreadsheet that I make available at Our main site theraptorpit.com for download and Viewing.
"Ronnie Coleman" is our "Master of Arms" and in charge of the Forum Police Force.
He is very experienced, so if there's a problem... He's on it I assure you.
The TRP Staff is very helpful if anybody has any questions.
Once the Wars Starts... Players can Post in any participating forum.(taunting)
Once Preliminary testing begins June 2nd there will be NO UPGRADES ALLOWED
All Drivers are Allowed, Which one you use... is up to you.
Also, No Ram Drives
The Benchmarks we are going to use will be...
We may add One more before the Priliminary testing Starts...
All Participating Forums must be entered by the time we start Prelimanary testing
Prelimanary Testing Begins June 2nd 2006 and Ends June 23rd (3 weeks)
The War Starts July 1st 2006 and Ends August 11th.
All testing MUST be done during each testing phase... Meaning (No Old Tests)
If you overclock your video card it may have throttling technology in it, much like the Pentium 4 does, so by overclocking too far it may actually reduce performance.
Other differences may also be overall system configuration, usually (lack of) chipset drivers, and far less often, different video card drivers
Also you forgot to mention minimum texture quality, forcing of no FSAA in driver even if app enabled, minimum mipmap quality, etc in drivers..... this is how people get +15% higher on their 3DMark scores, then overclock another +25% or so (until you get massive artifacts but the test just passes) to get near +44% higher scores. (1.15 x 1.25 = 1.4375). - That is how 3DMark ORB was left in the state it is in now. No Artifact testing enforced during testing makes it useless for comparisons. Seriously....
As 3DMark has no Artifact testing how do you stop cheats ?
Or is overclocking the card, until it just passes the test (ie: Can barely make the test out as 97% artifacts) and just managers to generate a score aswell as hacking the registry settings fed to the driver permitted ?
Saying one can get ~infinite~ in 3DMark is useless, if the system can not play games at the same settings, or is not even Prime95 stable and and 3D Artifact free for 3 days each (well this is an 'extreme test' right ?)
Can I just run a 768 Gpixel/sec system from Evans & Sutherland for 48 hours for each of the tests ? (768 Gpixel/sec is about 50 times more powerful than the most powerful stuff listed in Futuremark ORB btw )
The whole thing sounds potentially like a giant waste of money via a poor excuse (output some numbers on your screen) to destroy some expensive hardware just to get a (fairly meaningless as all systems configured differently, some will artifact once or more every 500 frames, others not at all) number to output on your screen.
My CPU score in '3DMark 06' (as the 2nd test benefits from 4 cores, the 1st CPU test doesn't though) and possibly result in 'Crystalmark version 9' should be quite high, but Futuremark is the poorest method to compare two systems for the above, and so many other, reasons.
Yes, even if I had 2 x GeForce 7900 GTX with 2 GHz x 256 bit GDDR3 Video RAM, or Quad-SLI (what a ****ing waste of money), or something from http://www.es.com I'd still have that opinion.
If you used more varied benchmaks, such as (image below), then maybe it would be interesting:
As 3DMark lacks interaction the cache hit rate for the CPU and GPU (memory switching predictions) is far higher than in games, even though the test uses heavier shading, textures, etc than games. The test is also not that CPU intensive and only indicates Direct 3D (totally lacks OpenGL) video subsystem performance.
However, If you can address my concerns I'll assemble an (Australian based) crew.
But this is ment to be fun and we dont encourage cheated and , if caught it would be disasterous for an entered team...
There have been a few changes to accommedate some of the testing problems some people are having... New ATI & Nvidia cards are having touble finding Certified drivers that actually work, so The driver selection is now unlimited (Null cheat is not allowed and will be delt with if someone is caught using it, so beware if your using one of the drivers that uses it, it will be examined closely)
Also... the Top 5 Players will be using all the tests
The B Teams will not be using 3d06 because it doesn't appear to work in some of the older systems.... The Scoring Site is open.
http://www.theraptorpit.com/html/fw_2006.html As you can see... we have a few sponsors and winners from the Top Dawgs Class as well the Tweaked Geeks Class will recieve some stuff.
Several Class winners and random drawings will take place as well.
If your forum would like to enter, you need atleast 5 Players, 1 cop and a Mod Or Admin to be the leader...
Just say so here... 10 forums are entered so far...
including Samino And Frog (3dmark world record holders)
Sorry to bother you TabrisDarkPeace (ed: You must have 4096 MB of Registered PC3200 ECC DDR-SDRAM). But It's crazy to hava 4 gig Plus too huh?
Crazy to have 1 TB in a Raid array too Huh?
Crazy to have a fast VGA card to play anything better than Video Poker too Huh?
Wow, that totally answered all of my concerns about 3DMark not testing for Artifacts. Thanks for not being a total jackass and explaining the benefits of this test in a professional manner. So you're saying people are not penalised if the whole test is an artifact fest ?, Sounds 100% professional to me... Just to get +50% higher scores in 3D Mark. You've also addressed my registry hack feeding the drivers 'ultra high performace, but one texture = one shade of colour' texture and mipmap quality settings like a true professional.
PC100 (800 MB/sec) can't do 10240 MB/sec in SANDRA using 2 x NUMA nodes can it ? - No it can't.
8) The Radeon X800 XL is ample for most games btw.
6.4 Gpixel/sec GPU + 32 GB/sec VRAM.
:? Amazing how many people think they need 10,000 3DMarks just to play Doom 3, Quake 4, Fear or Battlefield 2. Guess some people do a good job at driving the industry and encouraging ppl spend more than they feel comfortable spending. I wondered where those 'ideas' came from, but it appears we now know.
When you can top these scores, or learn what SANDRA is, let me know on this thread:
Tip: Oh, and if your low res monitor won't expand it in full screen, disable AutoResize in brower, or use FireFox to left click and get 1:1 zoom / fit to window toggle. After expanding the above image so you can clearly see the text in 1:1 sizing that is. If that is to hard, then just right click and to a Save As to get 1:1 viewing
PS: This board doesn't take SDRAM (PC100), and no machine which can generate the above screenshot (taken from the time Gomer challenged me, but he's cool, with the thread in the background just to taunt him).
(Trying to break the ice with this image I did in PSP 9 - 3 days ago, the original is far higher resolution in .pspimage format though 8) )
I hope you find part funny at least.... ..... and then maybe we can get down to the question of how are they testing for Artifacts ?, as I saw no such testing software mentioned in the testing methodology.
You've only mentioned some rather 'lame' ways of 'cheating'. Does overclocking until it is barely stable (eg: It can run 3DMark say 1 in 25 times from end to end without crashing) count as cheating ?, Because if it doesn't .... and as I've got until July or so to save for parts I may be interested. (Bearing in mind Australia like like 6 weeks behind USA stock on parts availability, so how does the 'test' correct for that ?).
Are Alienware, Dell, etc and pre-fab 'Quad-SLI' systems excluded from the testing ?, or in in their own league ? (eg: The Very Bad Investments League).
I like the Vista graphic. As far as the competition goes think of it this way. When you drag race cars you do everything withen the rules to go as fast as possible. These cars arent daily drivers and wouldnt make it on a cross country drive but they can run a 1/4 mile fast as all get out. The competition isnt about who has the most stable daily use computer, its about who can tweak there rig to the edge and score the best. Many of the things you are concerned about are very tough to discern and thus enforce so the fewer rules to enforce the easier it is to police.
Well Said Tgic...
I truly ment no offence, just a little kick in the ribs about cutting loose and letting it fly. there are rulz to follow. 2 proc setups are legal but dual 2 proc set ups are not. no restrictions on video cards. No cascades or LN.
Alienware is for wussies with too much money... and for grins I'll take your SSsandra challenge. Your team could use You and your rig... we got some healthy rivalry's going on, wish you guy's could join the fun..
PC100 (800 MB/sec) can't do 10240 MB/sec in SANDRA using 2 x NUMA nodes can it ? - No it can't.
..... and then maybe we can get down to the question of how are they testing for Artifacts ?, as I saw no such testing software mentioned in the testing methodology.
