|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Call of Duty 4: The next game in our benchmark list is the ever popular Call of Duty 4. Call of Duty 4 may not represent the best in DirectX 10 goodness, however the game is still visually stunning, and being as popular as it is, it is not hard to justify running it. To measure framerate in Call of Duty 4, I used the FRAPS benchmarking tool, which allows me to log framerate against time. I used the "Shock and Awe" single player mission, from the beginning until you land the chopper. I chose this mission due to the ease of replicating this section of the game over several tests. All settings are set to maximum quality unless otherwise noted.
The first test is essentially the best case scenario for Call of Duty 4; running at a fairly moderate resolution, no AA and Bilinear AF. We averaged 61.905 frames per second, with a minimum of 37 and a max of 88. This means the game should be fairly playable with few slowdowns at this setting.
For the next step up, we kept everything the same but increased the resolution to 1680x1050. Here, we got an average of 52.898 FPS, with a minimum of 25 and a max of 76. Performance has dropped a little bit, but still very playable.
Here we increased the AA to 4x, and maximized the AF as well, while keeping the 1680x1050 resolution. The average FPS dropped to 33.519, with a minimum of 14 and a maximum of 56 FPS. At this point, we are getting on the edge of playability.
We upped the resolution one more time, to 1920x1200, a common resolution for 24" LCD monitors. We turned the AA and the AF back down, and we got a result of 44.613 Average FPS, 16 minimum, and 69 maximum. With only a couple of slowdowns, this is still very playable.
Finally, we turned the screws, and maxed out resolution, AA and AF. We averaged 28.422 FPS on this run, with a minimum of 10 FPS and a max of 45. This is definitely out of the playable range for this card, and more for the 3870 or an X2 card. 3DMark:
3DMark03 First off, we have 3DMark03, with default settings on the stock clocks:
3DMark05 Next up is 3DMark05, also with default settings on the stock clocks:
3DMark06 Now, for the more recent 3DMark06, we did a lot more playing around, mainly to see what kind of gains overclocking gives us in different scenarios: Default Settings, stock clocks:
Default Settings, overclocked:
Here, we can see that overclocking got us about 760 points, or 7.5% increase in score. 1920x1200, 4xAA, 16xAF, Stock clocks:
1920x1200, 4xAA, 16xAF, Overclocked:
In this case, the overclocks gave us nearly 600 points, or an 11.2% increase in score. One of the most overlooked factors in reviews these days is the capability for decoding of HD content. Many mid-range cards utilize some sort of GPU acceleration for HD, which is a godsend for HTPC enthusiasts. For nVidia, the playback scheme is known as "PureVideo", and it is known as "AVIVO" for AMD based products. To test the 3850, I used an hour long video encoded with H.264 at 1080p resolution. The playback was absolutely flawless, and with the low temps and quiet fan, I can see this card making it into many HTPCs. Conclusion: The 3850 is in a very peculiar place in the market right now, with it being pretty much the same price as the 3870 Toxic edition. What it is going to end up coming down to is this: if you are looking purely for gaming performance, the 3870 is the better deal. On the other hand, if you are looking for a card that is cool and quiet for an HTPC type setup, with some gaming on the side, then the 3850 should be more your cup of tea. The 1GB flavor is currently only about $15 more than the 512MB version, so I don't see a good reason not to go for the 1GB.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
