Sapphire HD6950 2GB DiRT3 Toxic Edition Review

airman - August 11, 2011
» Discuss this article (14)

Lowest Prices

Testing:

Testing of the Sapphire HD6950 2GB DiRT3 Toxic Edition will consist of running it and comparison cards through the OverclockersClub.com suite of games and synthetic benchmarks. This will test the performance against many popular competitors. Comparisons will be made to cards of equal and greater capabilities to show where they fall on the performance ladder. The games used are some of today's newest and most popular titles to give you an idea of how the cards perform relative to each other.

The system specifications will remain the same throughout the testing. No adjustment will be made to the respective control panels during the testing, with the exception of the 3DMark Vantage testing, where PhysX will be disabled in the NVIDIA control panel if applicable. I will test the cards at stock speeds, then overclocked in order to see the effects of any increases in clock speed. The cards are placed in order from highest to lowest performing in the graphs to show where the cards fall by comparison. The drivers used are the 11.5 Catalyst drivers for AMD-based cards and the 275.27 for NVIDIA-based cards.

Processor: Intel Core I7 920 200x18 3.6GHz
Cooling: Noctua NH-U12P SE 1366
Motherboard: MSI X58 Platinum SLI
Memory: Mushkin 998995 Blackline PC312800 9-9-9-24 1600MHz
Video Card: Sapphire HD6950 2GB DiRT3 Toxic Edition
Power Supply: Mushkin 1000 watt Joule Modular power supply
Hard Drive: 1 x Seagate 1TB SATA
Optical Drive: Sony DVD-RW
OS: Windows 7 Professional 64-bit
Case: Corsair Graphite Series 600T

Comparison Video Cards:

Overclocking:

Sapphire HD6950 2GB DiRT3 Toxic Edition | GPU - 1010 MHz | Memory - 1500 MHz

Overclocking the Sapphire HD6950 DiRT3 Toxic Edition is very straight forward with the included Sapphire TriXX software. This software allows for overclocking of the core along with the memory with an almost unlimited range, and on top of that, the ability to adjust voltage as well. What feels like eons ago, in order to increase the voltage on a GPU, a hard volt-mod using potentiometers was required to establish higher voltages that are applied to the card. Now, it's as simple as moving a slider. Of course, adjusting voltages can be very dangerous so I would definitely leave that setting alone if you don't know what you're getting yourself into. Upping the voltage on anything generally allows for more stability at higher speeds, but it can also RUIN those overclocks due to the extra heat that's produced.

Using Sapphire's TriXX utility, I achieved a core clock of 1010 MHz and a memory clock of 1500 MHz. This equates to an almost 27% increase over the stock 800MHz and a 15% overclock over Sapphire's Toxic clocks. The memory overclocked very well and landed at a 20% increase over stock clocks and around 15% over the Toxic clocks. Some impressive numbers compared to overclocks of similar HD6950s have been found with this card here, allowing this card to produce similar results to its big sister, the HD6970, which sports a hundred extra shader cores. I seemed to have been a lucky contender in the quest for speed, as I was able to flash this card with the Sapphire HD6970 BIOS and unlock the extra cores and the higher stock memory speed. So, I went from an out-of-the-box speed of 880/1300 to 880/1375 and from 1408 shader cores to 1536. That's roughly a 10% increase in shader cores, and if results correlate, I should see some similar results. Below is a picture of the overclocked speeds I was able to obtain on both the stock and unlocked BIOS. The next picture shows the 1536 cores as opposed to 1408. Underneath these two pictures is a screenshot of what I gained as a result of this mod plus the original overclocking. My original 3DMark11 overclocked "Performance" score was 5645. With the extra shader cores, I sported over a 10% increase of performance and ended up just a hair over 6200. For a sub-$300 card, that's excellent!

Maximum Clock Speeds:

Testing for the maximum clock speed consists of looping Crysis Warhead and Unigine 2.5 for 30 minutes each to see where the clock speeds will fail when pushed. If the clock speed adjustment fails, then the clock speeds and tests are rerun until they pass the full hour of testing.