
Final Round FiringSquad-Intel Edito... 

Round 3 Editors Challenge Sponsored... 

Top 10 Challenge Round Sponsored by... 

Editors Challenge Sponsored by Inte... 

FiringSquad Editors Challenge Round... 

indigo196 (249) 

DanT (357) 

OgreFade (150) 

jacobvandy (718) 

darrellwu (24) 

Itchyeyes (76) 

Droniac (108) 

CanadaDave (303) 

kevinSpiess (156) 

|

|
|
|
7 entry(ies) in this category
|
 ATI Radeon X800XL – Hanging Tough (18 comments ) by: DaugWok (85) | Posted in cluster Round 3 Editors Challenge Sponsored by Intel Posted 13 months ago ( edited 13 months ago ) in category DEFAULT
» MEDIA (8)
 The ATI Radeon X800XL

 Face front. Can you see the wrinkles?

 Look ma, no power connector!

 Figure 1 - S.T.A.L.K.E.R Settings

 Figure 2 - Dark Messiah Settings

 Figure 3 - Oblivion Settings

 Figure 4 - Supreme Commander Settings

 Figure 5 - Results Summary

| Introduction
Some gamers are facing a difficult dilemma. Their video cards are several generations behind the cutting edge, but they don’t want to pay for a top-of-the-line DirectX 10 replacement. Unfortunately, only the most expensive video cards offer DirectX 10 compatibility, meaning more affordable options are obsolete the moment they're purchased. If money were not a factor, the most powerful video card available would be the best choice for nearly everybody. Right now, that choice would be a card based on the nVidia GeForce 8800GTX chipset. Such products retail for approximately $500 to $700 US dollars, depending on features, and are undeniably impressive. For many people, however, cost is perhaps the most important consideration. Consequently, their best choice may be what already clings tenaciously to their PC’s innards. Although a large selection of antiquated graphics hardware is available, a particular specimen is the topic of this analysis.
Released in 2004, the ATI Radeon X800XL is a video solution which may be present in many systems. In some respects, the X800XL has aged gracefully:
• It has a respectable amount of memory by current standards.
• It was designed to accelerate game engines, such as those of Doom3 and Half-Life 2, which power many current titles.
• It is highly capable compared to many products released both prior to and concurrently with it.
Nonetheless, the X800XL is showing its age in many newer titles. In the rest of this article, the strengths, weaknesses, and overall usefulness of the card in current games will be assessed.
Features
Even though the X800XL is nearly three years old, its feature set is extensive. An overview is kindly provided by the manufacturer here:
» http://ati.amd.com/products/radeonx800/specs.html
The most conspicuously missing feature is support for Shader Model 3.0. Unlike its contemporary, the nVidia GeForce 6800, the X800XL’s most advanced shader revision support is 2.0b. Even though both products technically fall under the “DirectX 9” umbrella, Shader Model 2.0b corresponds to DirectX 9.0b while Shader Model 3.0 corresponds to DirectX 9.0c. Simply put, Shader Model 3.0 can provide improved performance and/or image quality for games that support it. While it is not a required feature in the overwhelming majority of games, its absence from the X800XL’s feature list is glaring. Games which support Shader Model 3.0 include Elder Scrolls IV: Oblivion, Far Cry, S.T.A.L.K.E.R., and F.E.A.R, to name a few.
Although the X800 chipset was developed for the PCI Express interface, AGP flavors exist thanks to Rialto’s PCIe-to-AGP bridge chip. Therefore, many X800XLs reside in systems which cannot accept newer, PCI Express-only video chipsets. This is a particularly strong incentive for many people to keep their existing video hardware, as upgrade costs could include a new motherboard, processor, and memory.
The X800XL card used for this analysis sports 256 megabytes of video memory. By current standards, this is a minimal but acceptable amount. There are 512 megabyte versions, however, which undoubtedly handle modern games more elegantly. The XL model has 16 rendering pipelines and 6 vertex shaders, the maximum for the X800 series. That means that aside from clock speed and memory total, the X800XL tested here is as good as an X800 card gets. Therefore, it should exemplify its line’s performance.
Finally, a significant feature of the X800XL is its lack of an external power connector. The overwhelming majority of modern gaming cards require power from both the expansion slot and a power supply lead. This is due to ever-climbing PC hardware power requirements, and often an expensive supply is the only solution. Fortunately, the X800XL lacks the high power requirements which would necessitate such a measure.
Overclocking
The stock speeds for the card's core and memory are 400 MHz and 500 MHz, respectively. The memory is Samsung brand and is rated at 2 nanoseconds, which is appropriate for the 500 MHz clock speed. Using ATITool’s “Find Max Core” and “Find Max Mem” functions in combination with in-game testing, the maximum stable overclocks found were 413 MHz for the core and 535 MHz for the memory. These settings will be used to determine if overclocking has a significant effect on the card’s ability to play modern games.
Performance
The goal of this article is simply to determine if the X800XL can produce playable frame rates in current PC games. Instead of recorded time demos, common areas of each game were played to generate performance data.
The Test System
• Athlon64 3000+ CPU running at approximately 2.5 GHz
• ATI Radeon X800XL 256MB PCI Express Video Card
• 2 Gigabytes of DDR400 System Memory
• Windows XP Professional Operating System
Fraps 2.8.2 provided the listed minimum, maximum, and average frame rate values. Additionally, ATI’s Catalyst 7.3 drivers were installed, which are the most recent as of this writing.
S.T.A.L.K.E.R
Starting from the initial trading bunker, this test included a sprint through the opening camp, down the road to the north, and concluded with a firefight with approximately ten bandits at their warehouse to the northeast. Bad dudes were killed gruesomely, rest assured.
Settings
Screen Resolution - 1280x960
No full-scene anti-aliasing or anisotropic filtering
Maximum detail settings, static lighting
Complete Settings - Fig. 1
Results
Stock (400 MHz core/500 MHz memory)
» Minimum: 22 - Maximum: 153 – Average: 52.427
Overclocked (413 MHz core/535 MHz memory)
» Minimum: 27 - Maximum: 146 - Average: 59.075
The bane of the X800XL in S.T.A.L.K.E.R is dynamic lighting. When it is enabled, frame rates drop to the single digits, and the game is categorically unplayable. As mentioned in the previous article “Stalking through the paces”, this is likely due to the X800XL’s lack of Shader Model 3.0 support. Therefore, the results above were obtained using static lighting. The game performed very well, averaging nearly 60 frames per second at the listed settings. Considering the resolution and high video settings, the X800XL seems to be a viable accelerator for this game.
Dark Messiah of Might and Magic
In this test, a handful of enemies were engaged in melee combat in a Stonehelm courtyard. Water, fire, blood, and swift kicks to the face were all present.
Settings
Screen Resolution - 1280x960
No full-scene anti-aliasing or anisotropic filtering
Game-recommended detail settings
Complete Settings - Fig. 2
Results
Stock (400 MHz core/500 MHz memory)
» Minimum: 23 - Maximum: 56 - Average: 37.916
Overclocked (413 MHz core/535 MHz memory)
» Minimum: 25 - Maximum: 63 - Average: 39.792
Once again, the X800XL was able to deliver satisfactory frame rates at visually appealing graphics settings. Because this game averaged about 20 frames per second slower than S.T.A.L.K.E.R, there may be less room to increase graphics options. Unlike in S.T.A.L.K.E.R, however, dynamic lighting was enabled. Anyway, the X800XL pulls it off a second time. Huzzah.
Elder Scrolls IV: Oblivion
This game may be a year old, but it’s still one of the most intensive around. For testing, an outdoor area surrounding an oblivion gate to the east of Leyawin was used. Shadowmere aided in a battle against two Storm Atronachs and a Daedroth. It was a scuffle for the ages, to be sure.
Settings
Screen Resolution - 1024x768
No full-scene anti-aliasing or anisotropic filtering
Custom detail settings
Complete Settings - Fig. 3
Results
Stock (400 MHz core/500 MHz memory)
» Minimum: 17 - Maximum: 32 - Average: 23.695
Overclocked (413 MHz core/535 MHz memory)
» Minimum: 16 - Maximum: 33 - Average: 23.829
Compared to the previous two games, Oblivion gave the X800XL a battering. Even at the reduced resolution of 1024x768 there was noticeable slowdown, and overclocking did nothing to improve the situation. Oblivion is not as fast-paced as S.T.A.L.K.E.R or Dark Messiah, however, so at no point did it feel unplayable. Nonetheless, persons unable to cope with low resolutions, a lack of High Dynamic Range lighting (Shader Model 3.0 required), or occasionally choppy frame rates will need to consider a faster, more capable card than the X800XL.
Supreme Commander
No dual display this time. For information on how that feature influences performance, refer to the previous article “The War Widens: Supreme Commander Dual Display". Skirmish mode was used to play against three enemy AI teams on the 10km map “Four-Leaf Clover". Some light tanks and fervors, accompanied by the commander, attempted to destroy an enemy land factory guarded by erupters and seekers. Sparks did fly, and robots did die.
Settings
Screen Resolution - 1024x768
No full-scene anti-aliasing or anisotropic filtering
High detail settings
Complete Settings - Fig. 4
Results
Stock (400 MHz core/500 MHz memory)
» Minimum: 12 - Maximum: 29 - Average: 19.937
Overclocked (413 MHz core/535 MHz memory)
» Minimum: 11 - Maximum: 29 - Average: 20.317
Supreme Commander’s results are similar to Oblivion’s. Even overclocked, the X800XL failed to deliver liquid-smooth frame rates, but the game was entirely playable. However, some people may find the lowered graphical settings and sometimes chunky frame rates unacceptable. For them, a newer video card is probably the only solution, especially if they want to use the dual display feature. A faster CPU may help, but none were available to test.
Ballistics Report
Results Summary - Fig. 5
Pros
• At moderate to high settings, its performance in modern games should be acceptable to most people.
• X800XL owners can play current titles without the pain of upgrading OR enduring terrible frame rates. Enjoy it while it lasts.
• As 2+ year old video hardware goes, you could do a lot worse than the X800XL.
Cons
• There are much faster video solutions available in the budget and mainstream markets for those looking to upgrade.
• There’s no reason to own one of these unless you bought it a year or two ago. The X800XL lacks important features such as support for Shader Model 3.0 and DirectX 10.
Verdict
Based on these findings, many X800XL owners can consider themselves lucky. If they can do without maximized video options, high screen resolutions, or other performance-sapping features, there’s no reason for them to feel compelled to upgrade at this time. As DirectX 10-enabled games and hardware become commonplace, however, it will become increasingly difficult to justify not upgrading. Currently, DirectX 10 is a new and largely unsupported feature, and the X800XL seems to have enough horsepower to run the latest games acceptably. Cheapskates can rejoice. For now.
Links
Article - Stalking through the Paces
» http://firingsquad.com/matrix/blog.asp/11053/369/Stalking_through_the_Paces
Article - The War Widens: Supreme Commander Dual Display
» http://firingsquad.com/matrix/blog.asp/11053/274/The_War_Widens:_Supreme_Commander_Dual_Display
Fraps Homepage
» http://www.fraps.com/
ATITool Homepage
» http://www.techpowerup.com/atitool/ |
| 
| 18 User Comment(s) • 4 root comment(s) |
WallShadows (2349) Apr 07, 2007 - 10:59 pm
| I would just like to point out that the threshold between 256 to 512MB is often negligible when the price is put into perspective.
True you will often see a performance gain, but the gain is often a minimal lead which comes at the cost of an extra amount of $$, and even smaller if you intend to overclock.
Otherwise, I found it a very fruitful article as I was once a X800XL card owner. I think it's nice to see people bring back aged techs since not everyone can afford the new things offered today, which is a major draw back to sites which conduct game reviews and assume most people own high end systems...» Login to reply to this DaugWok (85) Apr 08, 2007 - 09:10 am | Edited on Apr 08, 2007 - 11:10 am
| Absolutely. I didn't knock the card much for only having 256MB, because I don't think it's a critical reason to upgrade. If the card only had 128MB of memory, however, my conclusion probably would have been much different.
I thought about analyzing video memory usage in each game to determine how much 256MB hurt the X800XL, if any. I know Oblivion has a command to monitor that, but I don't know about the other games. Maybe I'll come across an application that does so, or fraps will gain that feature.
Anyway, thank you for both your criticism and praise. Writing this article helped me quite a bit, because I've been putting off upgrading (until Intel's price cuts and sub-$300 DX10 cards, you see) and it made me feel better about what I have.» Login to reply to this |

|



DaugWok (85) Apr 02, 2007 - 03:19 pm
| | Phew. That was a lot of work. Suggestions are highly, highly encouraged. » Login to reply to this  maj.havok (3) Jul 06, 2007 - 11:38 am
| | Thank you, for all this information. I currently have a Ati x850xt 256mb card, have had it for almost a year and am just getting used to the fact that I can still(with games released within the last year) max most settings up without a hiccup. Being always a generation or two behind, usually because of cost, this has been the first card I have been able to consitantly do that with. With new DX10 cards coming, I may have been tempted to upgrade, but now I have the opportunity to wait til the prices drop a little, or the 3.0 pixel shader really comes into effect as it did with the recent Lost Planet demo(wasn't able to play). » Login to reply to this DaugWok (85) Jul 30, 2007 - 06:34 pm
| | Hey, somebody posted a comment in the last 3 months! :) I still have and use the x800xl I reviewed for this article. I thought about getting an 8800GTS 640MB recently, but none of the games I want to use it with are out yet. I guess we're both still waiting for now... » Login to reply to this |

|

Martimus (158) Apr 03, 2007 - 04:52 pm | Edited on Apr 03, 2007 - 04:58 pm
| I couldn't find where you listed what settings you enabled in-game. I happen to have that card and I am interested to see how these games do with it. Other than when you mentioned if dynamic lighting was enabled, I can't tell what video settings you enabled in each game - not even the resolution.
Please add that information to the review and I will score it then. Thanks for writing this, it is very useful to me, and it seems pretty well written!
edit: I now see the settings in the pictures. Considering that the pictures are off to the side and out of the way, it may still be a good idea to write out the primary display settings in the main part of the article. That would still be a good idea even if the pictures were embedded in the article, because then you could organize how the information is portrayed and make a better "apples to apples" comparison.» Login to reply to this DaugWok (85) Apr 03, 2007 - 05:34 pm
| | I was trying to avoid adding that information as text to the article's body because it's pretty dry stuff and screenshots are more precise. As an alternative, I bolded the figure references in each game's section. Does that help, or do you still want to see the settings as text? » Login to reply to this 


Martimus (158) Apr 04, 2007 - 11:02 am | Edited on Apr 04, 2007 - 11:06 am
| The reason I suggest typing out the primary settings you used is two-fold. First, you don't force the reader to search for these settings, which breaks up the flow of the article, and may even keep the reader from finishing it. Second, the settings are displayed in completely different formats in each picture, which makes it even more difficult to compare what settings are used between the games. If you had actually written out those settings, you could have standardized the way they were portrayed and made the flow of information to the reader easier to understand. You could still include the screen shots of the settings if you wanted to show exactly how each of the games was set up, but I would like to see the major points such as screen resolution, AA, AF, and Dynamic Lighting settings shown in a standard shared format for each game that was tested. It could be done in a single sentence (Use the same general sentence structure for each), or a short table under each game. As a reader who plans to use the information written in this review, this was the main issue that I had while reading it.
edit: Of course if you used the same settings in each game, you could just write what settings you used in the beginning, and just comment on changes made in the standard settings that you defined when you review each of the games.» Login to reply to this |

|

|

|


» Note: You need to be logged in to write a comment!Login here, or if you don't have an account with FiringSquad, register here, it's FREE! |

My Media-Blog categories
No categories created yet.
|