What to Expect?

[Dracas]

Hey All, posting again, my parts should arrive later today, and I was sort of curious what kind of 3Dmarks/3DM2K3/GLExcess Marks to expect, here's the rig - if someone could provide me with an estimate, that'd be awesome. Thanks again!

Athlon XP 2500+ 333/MHz
512MB GeIL GD PC3200 SDram 2/6-3-3 1T (Dual Channel)
Asus A7N8X Deluxe 2.0 Rev.A1 nForce2 SPP+MCP-T
Prolink Pixelview GeForce FX 5600 DVI/VIVO 256MB AGP 8X
IBM Deskstar (Click-death prevention Flashed) 60GXp 60 GB

I'm guessing around 12,000 3dMark 2001SEm 2600 3DMark 2003?

Can't guess on GLExcessMarks, haven't seen the thing used to bench OGL performance on the 5600 Series lately.

Thanks folks. :t

[fishybawb]

Sounds about right, but the 2001SE score might be closer to 11000. Had chance to test it yet? Let us know what you manage :)

[MadPistol]

yeah that sounds about right. I find it kind of funny though that a Geforce 4 ti 4200 128mb card could get a score of 13000 on stock settings on 3dmark2001se but the FX 5600 can only get around 11000. I have a 2.8 ghz p4 on an intel motherboard w/ 512 PC3200 ram, so I'm really not sure how I got such a high score on stock settings. :cool:

[Dracas]

**** Mo.

I installed DirectX 9 and I can't run 3Dmark 2001SE anymore, even tried the 'deinstall, get the new 330 v and reinstall' bit and it didn't help, says it needs 8.1, and its being stubborn about it.

Playing around with the system a bit still, my last 3Dmark 2003 330 bench came up as '2842'. Bum wrap, I can get better than that, considering that score seems to stay the change regardless of whether I overclock the CPU and Videocard or not.

I smell something fishy, and its name is Futuremark.

This card outperforms my GF's Radeon 9700TX regularly, at least in games and real world performance. Man, I've seriously lost faith in bogus benchmarks :rolleyes:

[Magua]

Didn't you expect to get around 2600 with 3dmark2003? 2842 is still considerably higher then that.

Are you trying to uninstall dx9 just to run a benchmark? Seems like a lot of work for something that dosent really matter.

[Dracas]

While its better than I estimated, I estimated before based on 'here-say', now that I've used the card some - it performs a lot better in real-world testing than it does on 3Dmark I guess.

A lot better. And its kinda silly, but this card runs faster at "Best Picture Quality" than it does at "Highest Performance"

Crazy, I guess that has something to do with how it has to remove things from the instruction queue if you start tellin it not to do a certain 'thing' per clock. Wierd.

Example: 2654 at Highest Perf. in the latest det.'s
2842 at Highest Picture Quality in the latest det's.

Wierd.

It still seems like Futuremark's software is lowballing the actual performance, likely in response to the b*tching and whining about 'specific target optimizations'. Its been used for a long time, even before 3Dmark, and a few whiny people whipped up the complaint to start the commotion (Get yer body in motion).

Ohwell, one whine to rule them all, one b*tch to bind them.

[Dracas]

Oh - and no, I wasn't uninstalling and reinstalling DX9, I went by Futuremarks support suggestion and uninstalled then reinstalled 3DMark 2001SE, no workie though, ohwell

[Logan[TeamX]]

Hate to be a naysayer on anyone's personal choice... but the FX cards, especially the FX 5600 and FX 5900 Ultra are not performing as they should be. The Raedon 9600 Pro and 9800 Pro (128 MB) are kicking them around the block.

Check www.tech-report.com for more info.

Logan :t

[Dracas]

And I care why...?

I don't give a **** dude, I don't have one, nor do I want one, because I'm happy with the money I just spent. I'm talking about the performance I can get with what I have, not with yer jimminny ***** envy "Mines better than yours" ****.

For future reference, you can keep yer two cents :rolleyes:

[bblqj78]

ooooooooooo.......

getting nippy in here!

:D

[Dracas]

Not really, I just think people should keep their BS to themselves, I'm interested in technology, not a particular brand. :t

[ComPooTer]

it doesnt even matter, we shouldnt be putting down people for buying nvidia cards. even tho ATI DOES have best price/performance, he bought the fx and he likes it dont make him feel like his buy sucks, hey its still better than most of the populations graphics cards

[bblqj78]

I'm interested in technology, not a particular brand.

Exactly!


This Nvidia-ATI thing is beyond the joke.....but I dont suppose it will ever go away until one company folds!
At the moment I gotta 9700pro....but I gotta tell you....my GF4Ti4200 was bettter for playing some games than the 9700. Had a snappier more reactive kinda feel...just what you want for a pacey FPS. I dont suppose NVidia has lost that! Hmmm..... Kinda feel like putting it back in!

[Dracas]

There's a thought. With PCI-X, we can all return to the good old days of having Dual Videocards running seperate resources so they don't eat each other in a dual monitor rig.

Anyone else here ever try to run two Voodoo2's at once? *laughs* :t

Win98's Dual-monitor support wasn't too awful, so long as you had seperate videocards, so it'd be a cool option to have again. Two large, panting, fiery GeForce PCI-X or ATi cards heavy-hauling side by side.

Swish :D

[MrBurns]

Originally posted by Dracas
There's a thought. With PCI-X, we can all return to the good old days of having Dual Videocards running seperate resources

I dont think so, because PCI-X has a special slot, which has a much higher performance than the others and will only be used by video cards. Because of this, mainboard manufacturers will only include one PCI-X video card slot.

[Dracas]

You're actually the first that has elected to say that, however - in the whitepapers and spec sheets I've seen, they've rather stringently shown that the entire system is being reverted to a single-slot upgrade pattern, and then (if needs be) they will expand the graphics capability later similarly to the AGP variant of the PCI slot (and rather - AGP IS PCI. Just a different form of it made strictly for graphics cards. AGP 16x has already been ruled obsolete).

I don't think we'll see a reference slot for graphics cards, at least not initially.

[MrBurns]

Originally posted by Dracas
I don't think we'll see a reference slot for graphics cards, at least not initially.

I have mixed this up w/ PCI express.

[Dracas]

PCI-X (PCI Express) Is a replacement for the 33MHz Parallel-PCI bus that continues to choke the system up. The new PCI-X Standard is a serial component interconnect (so why they didn't call it that is beyond me, unless to avoid confusion with SCSI...which means something totally different.) The new bus runs at 64-bit 133MHz (to synchronize with 64-bit processors, and match the minimum bus-speed of the most prevalent processors. (133x2 = 266, which is mostly what you find these days in ANY DDR system).

AGP is being effectively phased out and the unified PCI-X Architecture is replacing it. Also, as I said before, if it becomes necessary they can alter one slot (remeniscent of the AGP slot) to use two or more serial interfaces to increase bandwidth.

Think of PCI-X like some very VERY complex form of USB. That doesn't do the new standard any justice, but its the best way to explain it.

[MrBurns]

PCI Express and PCI-X are not the same. PCI-X will be the first implemenattion of a tecnologie originally called "3GIO". PCI express will be the final implemantation of "3GIO". PCI-X doesnt have a special graphics slot, but PCI Express will have. PCI Express will provide 200MB/s on each channel, starting w/ one channel for all normal cards and some more channels. The maximum number of channels is 32. At the beginning 16 channels will be used for graphics, later 32. For normal cards at the beginning one 1 channel will be used, than 2, than 4, than 8.

See here (pdf file):

http://www.intel.com/technology/pciexpress/devnet/docs/WhatisPCIExpress.pdf

Introduction
The PCI bus has served us well for the last 10 years and it will play a major role in the next few years. However, today’s and tomorrow’s processors and I/O devices are demanding much higher I/O bandwidth than PCI 2.2 or PCI-X can deliver and it is time to engineer a new generation of PCI to serve as a standard I/O bus for future generation platforms. There have been several efforts to create higher bandwidth buses and this has resulted in the PC platform supporting a variety of applicationspecific buses alongside the PCI I/O expansion bus as shown in Figure 1.

[Dracas]

You missed the point.

PCI-X and PCI Express are exactly the same, based from the same 3GIO platform, and you did nothing but recap EXACTLY what I had just said about PCI eXpress (Hence PCI-X) later expanding to include a Graphics particular upgrade slot.

It's still called PCI-x (Or PCI Express) and both are one and the same, you need to reread that document extensively, now the POINT REVISIONS. THose are Different, PCI-X (PCI 2.2) is going to be totally different from later revisions (And there may be quite a few of them, look how many revisions there are on USB and IEEE-1394 Serial Interfaces)

I think you just have the wrong idea (currently) about the early implementation and later implementations of 3GIO. It's all still 3GIO, just different revisions. It's still renamed PCI-X or PCI eXpress, those names are synonymous.

Get the idea now?