ATI X800XT PE vs. eVGA 6800 Ultra: The Battle at the Top

**MIG-31** · Feb 10, 2005, 07:35 PM

3DXtreme is extremely pleased to present our latest review. Thanks to our friends at both eVGA and ATI we are finally able to take a look two of the fastest stock AGP cards available on the market. We are not the biggest hardware review website on the internet so for us to be able to bring this type of review is only due to the generosity of the kind folks at each of the respective companies. Availability has been an issue with this latest generation of high end Video Cards, producing pieces that are able to run at blistering speeds is no doubt very difficult so understandably samples to a smaller site like us would is tough.
_______________
Review: 3dxtreme

redsolar · Feb 10, 2005, 09:25 PM

Not a bad read, but man I hate the way some people do graphs. Visually it looks like x800XT or 6800U beat the crap out of each other in each test. Poor graph scale cutoff, very poor.

Take BF graph on this page. While the difference is a measly (and statistically insignificant) 1.2 FPS, the graph makes you believe that an x800XT is waaaay ahead. Yet Doom 3 graph on the same page has exactly the same representation of a 6800U beating x800XT almost twice (99 fps vs 52).

Sandok · Feb 10, 2005, 09:29 PM

ya but that's why you should always look at the scale! Nice review!

BWX · Feb 11, 2005, 12:13 AM

I agree with red, it is ridiculous- if you are going to do graphs like that, then leave the graphs out. Graphs are supposed to be accurate and comparable graphic representations, not wildly exaggerated or on completely different scales..

I'm not picking on that website or author, but all sites that make graphs like that.

They should make it more consistent at the very least. At least 50 FPS in each direction, (50 from the top, and 50 from the bottom score) Or come up with some kind of algorithm for some kind of continuity. It's not really misleading, just annoying- and it defeats the purpose of the graphs in the first place..

The Battlefield 1942 graph goes from 97.5 to 99.5? Why not go from 25 to 125? The difference is 1 FPS but the graph shows it being 3 times as fast..
Pet peeve I guess.

krips · Feb 11, 2005, 02:19 AM

Graphs should start at 0 IMO.

BWX · Feb 11, 2005, 06:22 PM

Quote:

Originally Posted by krips

Graphs should start at 0 IMO.

yeah that would be a good place to start...

Feb 10, 2005, 07:35 PM	#1
MIG-31 HardwareHeaven News Mod Join Date: Dec 2002 Location: Nottingham,UK Posts: 56,860 Rep Power: 471 System Specs	ATI X800XT PE vs. eVGA 6800 Ultra: The Battle at the Top 3DXtreme is extremely pleased to present our latest review. Thanks to our friends at both eVGA and ATI we are finally able to take a look two of the fastest stock AGP cards available on the market. We are not the biggest hardware review website on the internet so for us to be able to bring this type of review is only due to the generosity of the kind folks at each of the respective companies. Availability has been an issue with this latest generation of high end Video Cards, producing pieces that are able to run at blistering speeds is no doubt very difficult so understandably samples to a smaller site like us would is tough. _______________ Review: 3dxtreme __________________ Please Post Your News Submissions Here

Feb 10, 2005, 09:25 PM	#2
redsolar Off duty Join Date: Jul 2003 Location: Las Vegas, USA Posts: 1,242 Rep Power: 0 System Specs	Not a bad read, but man I hate the way some people do graphs. Visually it looks like x800XT or 6800U beat the crap out of each other in each test. Poor graph scale cutoff, very poor. Take BF graph on this page. While the difference is a measly (and statistically insignificant) 1.2 FPS, the graph makes you believe that an x800XT is waaaay ahead. Yet Doom 3 graph on the same page has exactly the same representation of a 6800U beating x800XT almost twice (99 fps vs 52). __________________ Real_Men_Do_Not_Use_Space_Bar Main rig: Intel Core i7 930 @ 4.0 Ghz \| Corsair H50 Push/Pull \| 12GB OCZ DDR3 \| ASUS Radeon HD 5850 \| 3x 1TB Samsung F3 RAID5 on Dell PERC 5i SAS + 1TB Hitachi \| MSI X58 Pro-e \| 850W Antec Truepower Quattro \| RocketFish by Lian Li Full Tower Case \| Gateway XHD3000 30" LCD \| Dell 2408WFP 24" LCD \| Dell 2407WFP 24" LCD \| Logitech Z-5500 Speakers \| Windows 7 Ultimate x64 Laptop: Samsung x360-34P \| Core 2 Duo SU9400 1.4 GHz \| 4GB DDR3 \| 128GB Samsung SSD \| Intel 4500M w/ HDMI \| WiFi \| Bluetooth \| 13.3" WLED 1280x800 \| Windows 7 Professional x64

Feb 11, 2005, 12:13 AM	#4
BWX unplugged Join Date: Nov 2002 Location: USNY Posts: 19,669 Rep Power: 110 System Specs	I agree with red, it is ridiculous- if you are going to do graphs like that, then leave the graphs out. Graphs are supposed to be accurate and comparable graphic representations, not wildly exaggerated or on completely different scales.. I'm not picking on that website or author, but all sites that make graphs like that. They should make it more consistent at the very least. At least 50 FPS in each direction, (50 from the top, and 50 from the bottom score) Or come up with some kind of algorithm for some kind of continuity. It's not really misleading, just annoying- and it defeats the purpose of the graphs in the first place.. The Battlefield 1942 graph goes from 97.5 to 99.5? Why not go from 25 to 125? The difference is 1 FPS but the graph shows it being 3 times as fast.. Pet peeve I guess. __________________

Feb 11, 2005, 02:19 AM	#5
krips UD Cruncher Join Date: Sep 2003 Location: Alberta, Canada. Posts: 144 Rep Power: 0	Graphs should start at 0 IMO. __________________ [color=navy]3200+ Winchester \| XP-90 \| LANParty UT nF4 Ultra-D \| Enermax 535W \| BBA X800XL \| 1GB OCZ PC4000 \| 2x80GB RAID1 \| Audigy 2 ZS \| Logitech Z5300E[/color]

Feb 10, 2005, 09:29 PM	#3
Sandok DriverHeaven Extreme Member Join Date: Jul 2004 Posts: 9,472 Rep Power: 0	ya but that's why you should always look at the scale! Nice review!