I'm back... you smoked me on sssandra. Great score!!!
Hehehe, Cool, You should see my memory scores. (Over 10 GB/sec, and in SANDRA at that
, while using Registered [ECC] Memory which adds a transparent CAS tick [from being Reg, and it is key'd that way, so no choice on the matter
]).
I am running a dual proc, dual-core machine though, and
once tweaked out I can get over 180,000 in CrystalMark... just how much over I am not comfortable saying 'quite' yet though.
My submitted result to CrystalMark was
only just over 150,000, as I wanted some 'headroom' to spare on the score, ie: by not playing my full hand before the game had even started 8) , (eg: I am currently sitting in the #2 place, for CrystalMark for (AU) Australia 8O , so as not to attract much attention to myself 8) . However, I've got heaps of headroom :twisted: and could 'without doubt smash' 1st place in (AU) Australia for CrystalMark without really trying at all). I only need +6.5% more on score to take 1st place, and believe me when I say I can even smash past 180,000 without breaking a sweat. (As I am sure you're aware).
I deliberately loaded a few apps to reduce my score, and made sure not to take 1st place when submitting the results. Base clocks, no tweaks, on my most unclean OS partition, etc
If my config is against the rules... well, I might just run all the tests this weekend tweaked out a little, CrystalMark and SiSoft SANDRA look good, I'd need 2 x 7900 GTX (XFX #1 model) to compete in the video subsystem tests though (eg: All Futremark 3DMark tests).... Bear in mind I have no problems overclocking until the test is 85% artifacts and using every driver hack I know for texture and mipmap performance (that is legal under the rules) to just generate a few high numbers if that is what everyone else is doing.
I'd prefer each machine require (multiple instances of) Prime95 for 24+ hours, and pass that without any failures, aswell as no [zero] artifacts during all the 3D tests at all (which is very hard to test). As in 'stable overclocks', vs 'if I log off I can kiss my registry good-bye overclocks'. 8)
[eg: 1+ artifact in 216,000 frames = failure, Futuremark don't enforce this, or other 'tweaks' thus the ORB DataBase is useless for 'general comparisons' these days. MadOnion was their glory days IMHO, but even then ORB 'was not for n00bs'. eg: Look at all the 'Why is my 3DMark Score so low ?' posts on every forum on the net. --- Well, that's the answer
--- poor bastards.
]
I am down with it, even if I can't compete, but my rig 'breaks the rules' (Being a 4-way system, over 2 x CPU sockets and all) I guess I'll have to sit on the bench and watch the game.
I'll still post scores for those interested, and to help boost interest as a whole.
However, if my rig breaks the rules it'll save me money on video cards that are hard enough to get in Australia anyway. (Only one in every 14 people who can afford, and would pay for, one GeForce 7900 GTX card in Australia will actually get one because of stock levels here currently, that won't improve by July either
, I even offered AU$1200 (+33% 'tip) just for one such card, calling the extra a 'guaranteeing fee' [instead of a bribe
], just to get put higher (as in: move to first place) in the order queue, and I still don't have one).
Still, the CrystalMark scores possible on a Radeon X800 XL are quite good, but they'd be slightly better with a GeForce 7900 GTX. As for 3D Mark, I'd need far more than a X800 XL just to 'compete' in that area, but it does the job well enough... for now anyway).
:arrow: Another 'world wide web forum war' I'd like to see is a system builders war, where there are over 100 tests, some [highly] desirable, some mandatory, that need to be passed... all using the same configuration. With zero attempts (ie: disqualification) to 'get it right' if failing mandatory tests inbetween the desirable tests. - Now that would truely rock. Esp if done around March/April in 2007.
8) Looking at that video from last year (Forum Wars 2005):
It really should've been in 640 x 480 using Windows Media Video V9 codecs, instead of 320 x 240 using Windows Media Video V8
. The 1-pass constant bit rate audio, although using a V9 codec, should've been 96 Kbps (average) variable bit rate, and 2-passes. However I am happy to download 10 GB of data come time for a new video and 'assist' to create a higher quality one if need be. (My rig is good for video editing, as you've no doubt noticed). 8)
http://www.microsoft.com/windows/windowsmedia/forpros/encoder/versions.aspx
http://www.microsoft.com/windows/windowsmedia/forpros/encoder/faq.aspx
http://www.microsoft.com/windows/windowsmedia/forpros/encoder/default.mspx
I've studied AMVs for quite awhile
, understand 3:2 & 5:4 deblocking (aswell as how it would look in NTSC, PAL, aswell as TFT/CRT, HD formats, interlacing, [Inverse] Telecine, multi-pass variable bit rate with good minimum bit rate encoding, etc, so happy to offer my services & hardware as we have 'similar enough' agendas.
That and been a user of Jasc (well Corel now) Paint Shop Pro since it was made for Windows 3.x. Happy to work with 8192 x 8192 (64 MegaPixel+) multi-layer images.
If I can't 'paricipate' in the benchmarking, I'd still like to participate in ways that I can and test my skills, aswell as work with an ace team, on a more global scale. (Video + Image processing, Scene, Footage, Soundtrack, etc planning, and the like).
Heavily suspect MadModMike might even like to compete against you guys, for whatever reasons. :twisted:
I can't speak for him but you may want to PM him and check with him personally that's all. :idea: It may actually be good for morale around here with people pimping rigs and all. (He might be working on a new rig and have it ready for testing, if he's interested, but have missed this thread).