Gigabyte Radeon HD 7950 WindForce 3 Graphics Card Review

So 580 is still better then the new AMD cards... hmm xD Wasn't the 580 suppose to be the top card for AMD for the 7000 series? Or?...
 
Why not comparing single GPU cards only? Or why didn't the author use a crossfire setup of the benchmarked card, eg. 7950? Because he needed Nvidia to top the chart! What a lame conduit for a respected site.
 
We didn't test in Crossfire because we didn't have a pair of cards.

The SLI results are there for your own reference and not once we hinted Nvidia was faster just because the dual GPU scores are also there. If you read the review you will see it's mostly praises for this card (which received a very positive score of 90 -- we don't just give those away).

If you extrapolate from our results, it's highly likely a pair of these boards will surpass the 580 SLI setup unless Crossfire acts up on a given game.
 
mosu, it doen't matter how hard you shill for AMD, Rory Read isn't hiring.

Nice review yet again. The Windforce cooling is definitely one of the better implementations around.
 
My mistake, I meant, wasn't the 7970 be the top for the 7000 series for AMD?
Arti-x
 
The first rule of any marketing lesson will tell you to put the product you sell in front of competition at any means.The right name for the review should be "Second Best from AMD Smashes Nvidia Top Of The Line " in every aspect from game results, power consumption and thermal characteristics-hence durability, not to mention having a better price.Now I can see for the first time how infinity equals zero by means of an aussie cook and Nvidia fan.
 
Would be interesting to see the benchmarks with the new drivers for both parties, particularly considering the nvidia perf improvements. Other note - didn't know 12.1 catalysts had a release that supported the HD7xxx series (the AMD page pointed to 11.12 RC11 for ages). Running 12.2 pre-certified here atm and no issues so far.
 
I kinda agree with mosu, not of the ranting part, but if you hadn`t had 2 cards for CF you shouldn`t have put the SLI scores, moreover you should have added a 580GTX 3gigas factory overclocked card for the reference.
 
Why not comparing single GPU cards only? Or why didn't the author use a crossfire setup of the benchmarked card, eg. 7950? Because he needed Nvidia to top the chart! What a lame conduit for a respected site.

Not sure how to respond really, I am disappointed that this comment actually came from a registered member. It’s the same mix of cards we have been using for the past 6 months, the 7950 has simply been added to the mix. If you want to mail me a second 7950 I can e-mail you my address and I promise to update the results asap.

mosu, it doen't matter how hard you shill for AMD, Rory Read isn't hiring.

Nice review yet again. The Windforce cooling is definitely one of the better implementations around.

Thanks dividebyzero and I agree the Windforce is a nice setup.

The first rule of any marketing lesson will tell you to put the product you sell in front of competition at any means.The right name for the review should be "Second Best from AMD Smashes Nvidia Top Of The Line " in every aspect from game results, power consumption and thermal characteristics-hence durability, not to mention having a better price.Now I can see for the first time how infinity equals zero by means of an aussie cook and Nvidia fan.
You are comparing Nvidia’s 15 month old GPUs to AMD’s month old GPUs. Maybe you should wait till later in the year before talking such rubbish.

Would be interesting to see the benchmarks with the new drivers for both parties, particularly considering the nvidia perf improvements. Other note - didn't know 12.1 catalysts had a release that supported the HD7xxx series (the AMD page pointed to 11.12 RC11 for ages). Running 12.2 pre-certified here atm and no issues so far.
There isn’t much difference in the games tested, I have already checked.

I kinda agree with mosu, not of the ranting part, but if you hadn`t had 2 cards for CF you shouldn`t have put the SLI scores, moreover you should have added a 580GTX 3gigas factory overclocked card for the reference.

The card is operating at default GTX 580 frequencies and I have no idea why you wouldn’t want to see as many configurations as possible.
 
The temps are very impressive... sadly the price is not. In the UK this card retails for £400 and that is just too much to spend even on a high end card.
 
slh28 said:
The temps are very impressive... sadly the price is not. In the UK this card retails for £400 and that is just too much to spend even on a high end card.

I was just checking and £400! wow!

Never mind, won't be getting one of these in a hurry :)

BTW, good review, ignore the haters, Fact of the matter is i still got all the information I needed to make a decision on this and make an opinion on its performance. Overall good performance but just a little too pricey...
 
Steve said:
Why not comparing single GPU cards only? Or why didn't the author use a crossfire setup of the benchmarked card, eg. 7950? Because he needed Nvidia to top the chart! What a lame conduit for a respected site.

Not sure how to respond really, I am disappointed that this comment actually came from a registered member. It’s the same mix of cards we have been using for the past 6 months, the 7950 has simply been added to the mix. If you want to mail me a second 7950 I can e-mail you my address and I promise to update the results asap.

mosu, it doen't matter how hard you shill for AMD, Rory Read isn't hiring.

Nice review yet again. The Windforce cooling is definitely one of the better implementations around.

Thanks dividebyzero and I agree the Windforce is a nice setup.

The first rule of any marketing lesson will tell you to put the product you sell in front of competition at any means.The right name for the review should be "Second Best from AMD Smashes Nvidia Top Of The Line " in every aspect from game results, power consumption and thermal characteristics-hence durability, not to mention having a better price.Now I can see for the first time how infinity equals zero by means of an aussie cook and Nvidia fan.
You are comparing Nvidia’s 15 month old GPUs to AMD’s month old GPUs. Maybe you should wait till later in the year before talking such rubbish.

Would be interesting to see the benchmarks with the new drivers for both parties, particularly considering the nvidia perf improvements. Other note - didn't know 12.1 catalysts had a release that supported the HD7xxx series (the AMD page pointed to 11.12 RC11 for ages). Running 12.2 pre-certified here atm and no issues so far.
There isn’t much difference in the games tested, I have already checked.

I kinda agree with mosu, not of the ranting part, but if you hadn`t had 2 cards for CF you shouldn`t have put the SLI scores, moreover you should have added a 580GTX 3gigas factory overclocked card for the reference.

The card is operating at default GTX 580 frequencies and I have no idea why you wouldn’t want to see as many configurations as possible.

15 months later or not, simple fact is that RIGHT NOW the 7950 > GTX 580 relative to current pricing. And this is coming from a GTX 580 owner myself.

Of course though we'll see how the next Nvidia cards stack up. But I really don't know if they will get those low power consumption numbers of the 7900 series.
 
15 months later or not, simple fact is that RIGHT NOW the 7950 > GTX 580 relative to current pricing. And this is coming from a GTX 580 owner myself.

Of course though we'll see how the next Nvidia cards stack up. But I really don't know if they will get those low power consumption numbers of the 7900 series.

I'm not debating that at all, did anyone read the conclusion? I'm just saying why abuse the older Nvidia range so much when ... well its so old?
 
It seems that pointing something true upsets a lot of people.Never mind, I do not consider I'm ranting anything as long what I've said it's real.By the way, in Romania the Gigabyte Radeon HD 7950 WindForce 3GB DDR5 384-bit GV-R795WF3-3GD costs 475 US $ before tax.In other words, the 580 card is second iteration Kepler and has under a year of availability.The only way to be truly objective is comparing apples to apples.not launching dates or anything else than unbiased testing suites.
 
Nice review guys. I was looking at getting one of the Sapphire OC edition 7950 and give the old 6850 to the wife, but this looks like a great model too from GB.
 
Why not comparing single GPU cards only? Or why didn't the author use a crossfire setup of the benchmarked card, eg. 7950? Because he needed Nvidia to top the chart! What a lame conduit for a respected site.

All I could say was "wow.. guess you really can't please everyone."

Not everyone is a fanboy and I was able to quickly look at the charts and discern that the 7950 is generally faster than the 580 (a bit of a surprise to me really.) If the pricing were fixed (it's clearly very broken) then it would be even more impressive. This iteration of the 7950 seems like a stopgap between the reference 7950 and 7970 for a couple dollars less making it kind of meh.

FYI - I think anyone seriously looking to purchase cards in this performance range can look at charts and understand them well enough to see which card performs best with the applications or application types they may use. When reading a review I want it compared to as many other offerings as possible so in my opinion this comparison was great. Thanks for the review Steve.

edit: captainawesome - you're not allowed to use the word cranky in your posts... it's too confusing for me ;)
 
LNCPapa said:
FYI - I think anyone seriously looking to purchase cards in this performance range can look at charts and understand them well enough to see which card performs best with the applications or application types they may use. When reading a review I want it compared to as many other offerings as possible so in my opinion this comparison was great. Thanks for the review Steve.
Thank you, Papa.
 
mosu,

Are you an AMD shill or something?

"The only way to be truly objective is comparing apples to apples.not launching dates or anything
else than unbiased testing suites."

1) The article clearly showed that HD7950 is better than GTX580;
2) The article even highlighted that HD7950 by Gigabyte is a better buy than a reference HD7970 due to factory pre-overclock, additional overclocking with very noise levels courtesy of 3x Windforce heatsink, and due to lower price and power consumption.

Other results, such as GTX580 SLI, were included with HD6970 CF, HD6990, GTX590, etc. Using your retarded logic, the review should remove all Crossfire and SLI configurations too?

Also, everyone understands that it's "fair" to compare cards that are currently for sale. But HD7950 should easily beat GTX580 considering the latter is 15 months old. In technology, we expect faster performance at the same price or much lower price with similar performance. That's the point Steve was trying to make.

The only true way to gauge how good HD7950 really is to see what the competitor has. Technically, yes HD7950 is better than GTX580, but those cards are really meant to be competitors to Kepler. If HD7950 is faster than GTX670 Ti, then yes it's a great card. But if 2 months later NV launches a $299 GTX670Ti that beats an HD7950, then HD7950 @ $490 will look like the worst buy in a LONG time.

Right now, this review shows us where the new 7900 series stands and provides us an idea whether or not it's worth it to buy a 7900 series with an aftermarket cooler such as what Gigabyte has given us. But don't pretend to be oblivious to the fact that Kepler IS coming.

Your logic is so flawed in fact, it's like comparing the new Lamborghini Aventador and claiming it destroys the Ferrari 599, totally oblivious that the fact that the new Ferrari 620 GT's launch is imminent.

Only a fanboy, or a person who can drop $500 on GPUs every 6 months would jump at the HD7950 without seeing what NV has to offer. But hey, if 7900 series drops $100-200 in the next 2 months, don't come crying here.
 
i've got one of those bad *** HD 7950 from Sapphire.. although mine is the reference design, not the OC version. the thing is a beast. i am doing 1100mhz core and 1450 mem all with just 1.15 core voltage. i don't know how Sapphire did it... but all there cards voltages are lower than the AMD reference spec, even their OC version is too. i'm scared to push more because the skimpy reference cooler that's on right now might not be enough. i am really eager and can't wait to get the Zalman's new 79XX VF3000 cooler to come out. i wanna see how much more this GPU can push. as it stands at the moment, 1100/1450, that's 37% more performance than the stock card. can't beat that. the VRM module also seem to run surprisingly cooler than i thought. VRM is one those things people tend to overlook, but can easily fry your card if you let it cook up to 120c. luckily those HD 7950's usually hover around 85c~90c at the most, plenty cool for a VRM module, consider there's no heatsink what so ever covering them on the reference design.
 
LNCPapa said:
FYI - I think anyone seriously looking to purchase cards in this performance range can look at charts and understand them well enough to see which card performs best with the applications or application types they may use. When reading a review I want it compared to as many other offerings as possible so in my opinion this comparison was great. Thanks for the review Steve.

+1
 
Back