Powercolor Radeon HD 4850 512 MB

[wolf2009]

sorry, done .

[achilleask]

Originally Posted by Jelle Mees

My opinion on the latest TPU reviews:

Company of heroes, I don't know for sure but my guess is that this game did not sell well at all. I don't know a single gamer who owns this game and I am active on game-forums all over the planet and none of my friends ( Belgium, USA, Canada ) know or own this game. I know it's a good game and everyone that did buy it enjoyed it, but come on, that game is like two years old and I really think that nobody cares about that game anymore. So for me, a useless benchmark.

Far Cry, Cryisis got released and Far Cry 2 is on the way, and you guys still waist time bencharming in Far Cry? Come on, it was released more then 4 years ago. Yes it was a good game, but who's gonna buy a new GPU for Far cry???

Prey, beatifull game, awesome gameplay, but come on, who still plays it? I am pretty sure I finished that game with my 7600GT on 1600x1200.

Quake 4, again, nice game, but it's old and for some reason the multiplayer never got the attention it deserved.

Splinter Cell 3, I enjoyed it....more then 3 years ago.

3DMark03, 3dMark05, why?

Why don't I say anything about F.E.A.R, simply because the multiplayer is free ( and not dead yet ) and the game is one of the first multi-GPU optimized games.

Don't get me rong, I love the TPU reviews, I just wish that you guys tested in more games that are usefull for more gamers.

Couple suggestions
- World of Warcraft ( IT IS the most popular MMORPG on the planet and running it on 2560x1600 with high AA and AF settings is still not possible on all areas )
- Replace Far Cry with Far Cry 2 as soon as it gets released
- Stop running 3DMark03 and 3DMark05 and start running 3DMark Vantage ( better, stop running 3DMark )
- Trackmania Nations Forever, it's free, it's new, it's becoming more popular every day and you need powerfull hardware to run it maxed out
- America's Army 3.0 as soon as its out ( free online UnrealEngine-3 FPS-game )
- Age of Conan, this game is selling like crazy in Europe, better then GTA4
- Team Fortress 2
- Mass effect

I have only played Trackmania Nations Forever, so don't think I mention these games for my personal intrest.

Replacing old, dead games by new and/or popular games would only improve the TPU reviews that are already really great. And please don't start bitching at me just because I think it's better, it's just my opinion, just one guy's opinion. ( I got banned for posts like this one on Guru3D )

One more thing, about the overclocking part. I would love to see one more 3DMark chart where the OC results are included, just so that gamers know what performance impact to expect once overclocked. Pretty sure everyone would love to know if an overclocked 4850 can beat the 3870X2.

I think most people misunderstand the point of a benchmark. Benchmarking isn't a "promise" of how good a game will run on your VGA. It's a means for comparison. A quantification of performance.
If game A runs at 60fps on VGA X, and with the same settings at 100fps on VGA Y, then Y > X.

If we followed your reasoning, of testing VGAs on new games every time, then we would have VGA X running at 60fps on game A and VGA Y running at 50fps on game B. However Y might still be better than X, only B is a more demanding game than A, hence the performance drop. Changing games all the time doesn't help benchmarking. The applications displayed in the review are run for many VGAs, I agree, however using standard applications across multiple reviews provides a consistency in the process making results reusable and comparable between them.

Most games that are used for benchmarking are games that are very demanding upon their time of launch and can stress out a VGA with "cool features and graphical effects". Additionally, these games use a lot of the features we pay for on VGAs. The reason they remain as benchmarks for a year or two is to have a solid base of comparison between VGAs of a generation period of 1-2 years.

I repeat, benchmarking is for COMPARISON, not to tell you which card is best for which game. This is why 3Dmark is used. Following your logic, yes, 3Dmark is useless. I wouldn't care how high my VGA scores on 3Dmark. It's not like I'm going to be running it and looking at the "pritti piksharz". However I can see, for instance, that the VGA I intend to buy scores 3000 points, while an alternative scores 2800 points, helping me decide what the best VGA to buy is.

Running a benchmark on World of Warcraft is not only pointless but stupid. First of all, the graphics are not impressive (I agree it has some nice spell and environmental effects but nothing that will hog a good gaming PC's resources). Secondly, I have rarely seen a benchmark run on WoW. If I tell you WoW runs at 100fps on this VGA... so what? I'm pretty sure all VGAs that cost more than $60 can hit 100fps on WoW. It can not be used as a means for comparison, so it is not.