facebook rss twitter

Review: MSI R7950 Twin Frozr 3GD5/OC

by Tarinder Sandhu on 16 May 2012, 07:42 4.0

Tags: MSI

Quick Link: HEXUS.net/qabgpb

Add to My Vault: x

Final thoughts and rating

Folks with the ability to spend £300-plus on their next graphics card need to take a good look at the GeForce GTX 670 from NVIDIA and Radeon HD 7950 from AMD. These two premium GPUs offer solid performance and features at a price, albeit steep, that's at least 15 per cent lower than the best single-GPU cards'.

Recent price cuts have pushed Radeon HD 7950 pricing down to £300. Spend an extra £20 and the MSI R7950 Twin Frozr 3GD5/OC can be yours. Equipped with a better cooler and shipping with a 10 per cent hike in the core clock, there's plenty of headroom in the GPU, as shown by our overclocking tests.

Benchmark numbers also show the card to perform well at a 1080p resolution though, with our maximum in-game quality settings, not so great when tasked to run three screens. But factor in the quiet operation and class-matching power consumption and we'd only just give the nod to a regular GTX 670, especially if you don't mind tinkering with a little overclocking on this Radeon HD 7950 GPU.

MSI's done a credible job with the Frozr III version of the HD 7950 3GB card. The £320 price, which is bang-on GTX 670 money, makes it, we believe, worthy of being on any sensible enthusiast's shortlist. Recommended for its cool, quiet operation and enormous overclocking potential.

The Good

Outstanding overclocking potential
Excellent thermal characteristics
Keeps relatively quiet under load

The Bad

GTX 670 looms large
Conservative out-of-the-box overclocking

HEXUS Rating

4/5
MSI R7950 Twin Frozr 3GD5/OC

HEXUS Awards


MSI R7950 Twin Frozr 3GD5/OC

HEXUS Where2Buy

The reviewed card is available from here.

HEXUS Right2Reply

At HEXUS, we invite the companies whose products we test to comment on our articles. If any company representatives for the products reviewed choose to respond, we'll publish their commentary here verbatim.



HEXUS Forums :: 6 Comments

Login with Forum Account

Don't have an account? Register today!
Why use MSAA at these resolutions ? With it turned off the framerates would surely be much higher, and gives the end user a much more useful benchmark.
Remember just how wide the viewing area on three screens is. The pixel density is no smaller than running on one 1,920x1,080 screen, and we'd always have high AA/AF for that. Of course you can reduce the AA/AF to an extent where the cards produce much higher framerates and you could keep reducing them for, say, a Radeon HD 7850 or Radeon HD 7770.

The point here is to see how these super-powerful, expensive video cards react when the same quality settings are applied to a single screen (24in, 1,920x1,080), a super-high-resolution screen (30in, 2,560x1,600) and super-wide screens (72in, 5,760x1,080). Putting the same IQ load on cards provides clear continuity between resolution results and indicates that a very high-quality three-screen gaming experience is just outside the remit of a single HD 7950 or GTX 670 card. But the settings also provide an adequate test for two-card HD 7950 or GTX 670, as you will see in an upcoming article.
Another good review Tarinder, I'm especially liking the FPS/Time graphs, very useful for the more detailed differences between how the cards handle different games.
Oh yeah I understand the need for it but in a way it's kind of an unfair test, as the extra VRAM puts the AMD at an unfair advantage, considering how much extra MSAA uses, when FXAA does as good of a job with less of a penalty.

Or am I wrong ? From what I remember MSAA uses a lot of VRAM because it generates higher res textures to overlay, so with the 3 screens the VRAM needed is massive, and would equal more of a performance hit than the calculation itself.

That's kind of what I remember anyway, wouldn't be the first time I'm wrong :D
Tunnah
Oh yeah I understand the need for it but in a way it's kind of an unfair test, as the extra VRAM puts the AMD at an unfair advantage…D

AMD is giving you a 3GB buffer for the same price as NVIDIA provides a 2GB buffer on the GTX 600-series. I personally consider this a good thing from AMD - let's forget about the the design choices, bus width, etc. - and, taking it to an extreme, would prefer NVIDIA provide 4GB of GDDR5 memory at the same price as 2GB.