Sign in with
Sign up | Sign in
Your question
Solved

MSI GTX 660ti PE vs Sapphire HD7950 Vapor X

Last response: in Graphics & Displays
Share
December 12, 2012 5:53:44 AM

Hey everyone. After much research I have come down to 2 cards that I would like to buy, im from new zealand so the prices are higher.

The MSI GTX 660ti Power Edition which is $500
The Sapphire HD7950 Vapor X which is $470

From what Ive seen everyone has the "HD7950 hands down" response to the gtx660 ti and hd7950 and yet when I look at the benchmarks and youtube videos the gtx 660ti usually outperforms or loses in the odd case. Newegg customers seem to really love the msi gtx 660ti and too the sapphire hd7950 but more seem to have issues with the hd7950. I read the argument is that the hd7950 has more bus and video ram. It has more shader units and so forth yet the 660ti is on par with this card in performance.

I guess my question is simply which card is better without a biased opinion and from someone who knows what they are talking about or has seen both cards in action to shed some light on the matter because I see a card with more bus width more video ram high shader units etc that's on par with a card that is not and leads me to believe that the processor on the nvidia card is simply much faster and better quality. I don't claim to be an expert, I don't know a hell of a lot about graphics cards so don't attack me for something I have said that isn't correct I would just love a better understanding.

Cheers!
a b U Graphics card
December 12, 2012 6:10:52 AM

Buy 7950 .my opinion is msi made low cost products but quickly they raise complind ex : iam see many msi motherboards malfuntuning and after 10 to 12 months many mobo bios failed.then only u buy another mobo the custamer care is not curing this.
m
0
l
December 12, 2012 6:13:23 AM

thanks for your response. I have no concerns for the quality of MSI nor Sapphire I believe they are top notch, have been using an MSI GTX 460 cyclone and works like a dream.
m
0
l
Related resources
a b U Graphics card
December 12, 2012 6:14:09 AM

7950 is best choice!
m
0
l
December 12, 2012 6:19:08 AM

Yes but why? the only argument that gets put forward is the high bus width and video ram which doesn't seem to make a huge difference. I do see the hd7950 pull away at redulous high resolutions or multi monitor performance but other than that.
m
0
l
a b U Graphics card
December 12, 2012 6:19:12 AM

7950 is better than 660.
m
0
l
December 12, 2012 6:31:52 AM

It sounds like everybody has read a comment of someone saying hd7950 is better and followed it without question. Is there a comparison of the 2 cards with 12.11 drivers for amd against a non reference gtx 660ti? I really want to be convinced
m
0
l
December 12, 2012 6:47:09 AM

instantcoffeenz said:
It sounds like everybody has read a comment of someone saying hd7950 is better and followed it without question. Is there a comparison of the 2 cards with 12.11 drivers for amd against a non reference gtx 660ti? I really want to be convinced


I agree. I don't know why everyone automatically jumps on board with AMD. Although I do agree that MSI isn't an exceptional company. Look into this gpu. http://www.newegg.com/Product/Product.aspx?Item=N82E168...

I was going with that until I switched my build around a bit and went with the ASUS GTX680-DC2T-2GD5 2GB which is $540.
m
0
l
December 12, 2012 6:52:57 AM

Asus is really expensive in my country way overprices compared to everything. not that it matters either I really hate the look of their coolers. If they were well priced I would have considered them :)  thanks for your response! why is MSI bad? I really like the look of their cards , they are cool and quiet and from my experience it has been very good!.
m
0
l
a b U Graphics card
December 12, 2012 7:38:49 AM

9750 is amd card other companys buy directly from amd and change their original cover only not parts. but low cost cards are low quality parts so life is decreased cards.
m
0
l
a b U Graphics card
December 12, 2012 7:39:51 AM

I'd suggest you go with the msi 660 ti pe.The stock 660 ti performs in between the 7950 boost and 7950 boost.
http://www.techspot.com/review/615-far-cry-3-performanc...
The OC'd card from msi will perform much better.I'm using a msi 660 ti pe/oc myself and I love it.
m
0
l
December 12, 2012 7:50:56 AM

instantcoffeenz said:
It sounds like everybody has read a comment of someone saying hd7950 is better and followed it without question. Is there a comparison of the 2 cards with 12.11 drivers for amd against a non reference gtx 660ti? I really want to be convinced

http://techreport.com/review/23981/radeon-hd-7950-vs-ge...

there you go.Latest driver all nice and rosy.
m
0
l
a b U Graphics card
December 12, 2012 7:57:53 AM

^that website used 310.54 drivers.Look at the techspot review it uses the 310.70 beta drivers.

EDIT: the conclusion is the same btw
m
0
l
December 12, 2012 8:11:31 AM

310.70 just boost FC3 performance. Those games tested in that review were already getting max performance from that driver.Maybe hitman a bit less.
m
0
l
a b U Graphics card
December 12, 2012 8:44:48 AM

True I guess.But the most surprising benchmark was the hitman absolution where a 7950 at 61fps suffered worse frame times than a 660 ti at 39fps.
m
0
l
a b U Graphics card
December 12, 2012 9:59:39 AM

The 7950 performs way better check out this benchmark on games and other stuffs... http://www.anandtech.com/bench/Product/550?vs=647
As you can see, the 660Ti suffer low FPS from higher resolution but perform better on smaller resolution like 1080p(1920x1080) other reason why you should get 7950 is that's it's only few dollars higher than 660Ti for a bit better performances...
m
0
l
December 12, 2012 10:34:45 AM

Anik8 said:
True I guess.But the most surprising benchmark was the hitman absolution where a 7950 at 61fps suffered worse frame times than a 660 ti at 39fps.

i really like the way techreports benchamrks.Really love there latency test.Which tells you which card is giving more stuttering and which is smooth.No other tech site does it.Which is a real shame.Since framerate alone can't give you that smooth feel.Some game even at 80 frames feels stuttery than some at 40FPS.
m
0
l
a c 144 U Graphics card
December 12, 2012 11:30:27 AM

that's why they say frame rates alone does not tell the story. hard ocp also mention about this smoothness in most of their SLI vs CFX review though they did not do detail works like tech report does. they simply write in their summary how they feel about it when they do the test and did not have real data to back it up. some people suggesting they represent the data like tech report did but they said they can't just carbon copy the method use by tech report.
m
0
l
a b U Graphics card
December 12, 2012 12:12:34 PM

legendkiller said:
The 7950 performs way better check out this benchmark on games and other stuffs... http://www.anandtech.com/bench/Product/550?vs=647
As you can see, the 660Ti suffer low FPS from higher resolution but perform better on smaller resolution like 1080p(1920x1080) other reason why you should get 7950 is that's it's only few dollars higher than 660Ti for a bit better performances...

First of all we're talking about performance through recent drivers.The anandtech bench is months old and even in there if you go downwards, the 660 ti wins.

@cooldudesubho
yea they've doing a nice job in their style of reviewing GPUs.
m
0
l
a b U Graphics card
December 12, 2012 12:44:08 PM

I want to put my 2 cents here. The 2 mentioned cards perform very similarly in almost all game benchmarks. The difference (aside from the architectural differences) is that the 660Ti has 192bit bus BUT has a higher memory speed with less memory amount (2gb). 7950 has 384bit bus BUT has a lower memory speed WITH more amount of memory (3gb). The ultimate (potential) performance difference come from the overclock-ability of 7950. The only weaker point of this card is its lower memory speed compared to 660ti. 660ti's is its 192bit bus. Why do they say "7950 hands down"? Because a 7950 will reach very high bandwidth (twice the amount of 660Ti's i'd dare to say) with memory overclocking (which is not so hard to do) and will destroy a 660Ti in bandwidth intensive stuff (at least in theory)
m
0
l
a b U Graphics card
December 12, 2012 1:18:53 PM

Wow you just went too much carried away there.First of all when it comes to memory speed in particular amd couldn't give their cards much headroom and hence they were forced to release their cards much below the 6GHz data rate.Hence it is not much overclock-able either without increasing any voltages.
nvidia released their kepler cards at 6Ghz data-rate and they have provided much headroom in there too such that they've even shown how synchronous these card's data rate is with diagrams showing profound data streams.
People can overclock memory speeds of 680s,670s,660 ti-s and 660s to higher than 7.2Ghz easily without increasing any voltage.Even I've tried and successfully breached that mark.
m
0
l
a b U Graphics card
December 12, 2012 1:34:49 PM

Anik8 said:
Wow you just went too much carried away there.First of all when it comes to memory speed in particular amd couldn't give their cards much headroom and hence they were forced to release their cards much below the 6GHz data rate.Hence it is not much overclock-able either without increasing any voltages.
nvidia released their kepler cards at 6Ghz data-rate and they have provided much headroom in there too such that they've even shown how synchronous these card's data rate is with diagrams showing profound data streams.
People can overclock memory speeds of 680s,670s,660 ti-s and 660s to higher than 7.2Ghz easily without increasing any voltage.Even I've tried and successfully breached that mark.

I already mentioned what you say...anyway, here is what i mean (produced bandwidth):

Radeon - (384 / 8) * 1200 * 4 = 230,400MB/s
GeForce - (192 / 8) * 1500 * 4 = 144,000MB/s

Overclocked (+250 Mhz):
Radeon- (384 / 8) * 1450 * 4 = 268,800MB/s
GeForce - (192 / 8) * 1750 * 4 = 168,000MB/s

You can't alter bus width, but you can overclock memory speed. As the higher overclocking is involved, the gap will be higher. All in all, these are numbers on paper and may not represent the final outcome. Like i mentioned before, this is my 2 cents of theoretical information that i see about the difference between 2 architectures.

EDIT: LOL sorry i miscalculated the radeon cause a 7950 has 1250mhz default not 1200. And also a 7950 can reach high overclocks without valtage-tweaking too. Many people can achieve around 1500mhz memory clocks without altering voltage.
m
0
l
a c 144 U Graphics card
December 12, 2012 2:00:59 PM

bandwidth intensive stuff? can you give some examples to that?
m
0
l
a b U Graphics card
December 12, 2012 2:09:19 PM

renz496 said:
bandwidth intensive stuff? can you give some examples to that?

High MSAA for example 8x and probably very high resolutions. Though these are theoretical information what i share, i already provided the calculations. The 2 companies follow different approaches in their designs. AMD put higher bus but lower memory, Nvidia put lower bus but higher speed. At last, game engines, optimizations, detail levels etc will differ a lot and those differences i talked about MAY/MAY NOT show in real life.
m
0
l
a c 144 U Graphics card
December 12, 2012 2:23:04 PM

it does show in that situation (high level of MSAA or high resolution) in some games. but whenever there is talk about 7k series vs 600 series in AA aspect some people jump to conclusion that nvidia 600 series can't do AA at all. this review did the test how well GTX660Ti handle AA:

http://www.hardocp.com/article/2012/08/27/nvidia_geforc...

Quote:
The Radeon HD 7950 has a 384-bit memory bus with 240GB/sec of memory bandwidth, while the GALAXY GTX 660 Ti GC has a 192-bit bus with 144GB/sec of memory bandwidth, yet we sometimes see the GALAXY GTX 660 Ti GC video card performing faster, but the 660 Ti does have a GPU clock advantage. We don't see that 66% advantage in memory bandwidth the Radeon HD 7950 has in real-world gaming with high setting AA configurations. This means there are other factors besides the width of the memory bus and the bandwidth that affect performance between these two cards; namely GPU clock and architecture advantages.


m
0
l
a b U Graphics card
December 12, 2012 2:44:48 PM

Well, MSAA is not the only AA tech around, and honestly i know of a couple of better AA techs such as FXAA and SSAA. The thing is, MSAA provides a good balance of performance and visual smoothness. Since MSAA is only applied at the edges of the objects (not the whole object area) it is kind of lightweight technique which provides the needed eye candy by not sacrificing a huge performance. Oh well, at the end 7950 is slightly faster in many scenarios. At the same rice point i'd prefer 7950
m
0
l
a b U Graphics card
December 12, 2012 3:08:54 PM

MSAA, great! that's the only thing you can point out.In overall framerates the 660 ti performs in between a 7950 and a 7950 boost.In frametimes and latencies 660 ti is races ahead.
m
0
l
a b U Graphics card
December 12, 2012 3:50:51 PM

Anik8 said:
MSAA, great! that's the only thing you can point out.In overall framerates the 660 ti performs in between a 7950 and a 7950 boost.In frametimes and latencies 660 ti is races ahead.

7950 and 7950 boost is exactly the same card, only slight overclock difference and frametime issue can be solved by a 3rd party solution software, which gives BETTER result than Nvidia's on-board solution (there was even an article with tests in toms hardware). 660ti and 7950 perform similarly, give or take, apart from the difference i mentioned above, if you care to read again
m
0
l
a b U Graphics card
December 12, 2012 4:09:53 PM

lol dear frametimes and latencies cannot be fixed by 3rd party solutions,it depends on the functioning of the chip.do you have any idea on how GPUs work.Moreover this statement of yours says it that you haven't even read the contents of the techreport article.you are just one of those fanboys that end up making biased claims about the camp they support without any conclusive proof or evidences.
That slight overclock difference is enough to provide extra fps in benchmarks.It is why stock cards are recommended for benchmarks,someone can easily throw a overclocked 660 ti and it will get past the 7950 boost as well.Throughout this thread several people has shown fps benchmarks,frametimes-latencies ones and all of them suggests the stock 660 ti is the better option than a 7950.
m
0
l
December 12, 2012 4:13:10 PM

cooldudesubho said:
i really like the way techreports benchamrks.Really love there latency test.Which tells you which card is giving more stuttering and which is smooth.No other tech site does it.Which is a real shame.Since framerate alone can't give you that smooth feel.Some game even at 80 frames feels stuttery than some at 40FPS.
I know! thats another thing I have been thinking about, the stuttering and overall smoothness of the gameplay. higher fps isn't always better and that is 1 thing I did notice when moving to nvidia is less stuttering in games and overall smoothness improved. I think im going to play it save spend the little bit extra and go with the MSI N660 Ti PE. Still not 100% sure on that but about 99..9% sure lol
m
0
l
a b U Graphics card
December 12, 2012 4:32:14 PM

Anik8 said:
lol dear frametimes and latencies cannot be fixed by 3rd party solutions,it depends on the functioning of the chip.do you have any idea on how GPUs work.Moreover this statement of yours says it that you haven't even read the contents of the techreport article.you are just one of those fanboys that end up making biased claims about the camp they support without any conclusive proof or evidences.
That slight overclock difference is enough to provide extra fps in benchmarks.It is why stock cards are recommended for benchmarks,someone can easily throw a overclocked 660 ti and it will get past the 7950 boost as well.Throughout this thread several people has shown fps benchmarks,frametimes-latencies ones and all of them suggests the stock 660 ti is the better option than a 7950.


http://forums.guru3d.com/showthread.php?t=322031

Either you are biased or i am :)  This is the feature list of Radeon Pro, the program mentioned. This will teach you more about GPUs and software features that might improve the user experience. A feature of this software:

- New: Dynamic Framerate Control (DFC) support. This feature acts like a frame rate limiter with smoothness control, just set a frame rate target and RadeonPro will try to keep it as close as possible while maintaining frame rendering times close to each other to avoid stuttering.

Also, if you didn't know, 7950 Boost IS a reference design which isn't factory overclocked by 3rd party company.
m
0
l
a b U Graphics card
December 12, 2012 4:48:57 PM

dude...! framerate limiter is no big deal kepler cards also support it and evga precision is used to apply it ever since the 680 was launched.The point is the kepler GPUs performed smoothly even without any framerate limiter.Moreover I may already use v-sync to reduce screen tearing and how much more do I need to sacrifice on the performance to enjoy smooth gaming?? tell me.With kepler cards its not even that much necessary.
Plus I never mentioned the 7950 boost is a 3rd party solution but there is also another lesser clocked reference 7950 out there and I said the stock 660 ti performs in between them.You brought up the 3rd party term in the first place to flee from the harsh truth of the latency tests conducted by techreport.
m
0
l
a c 87 U Graphics card
December 12, 2012 9:20:03 PM

Kepler cards do not support it in the same way as RadeonPro. Look up Tom's recent article on the Radeon 7990 versus the GTX 690 for a little more information on it. RadeonPro is not a mere frame rate limiter, it controls frame-by-frame output better. RadeonPro doesn't necessarily sacrifice performance either. Tom's Radeon 7990 versus GTX 690 article showed this quite clearly.

V-Sync is improves screen tearing at the cost of high input lag. Besides that, AMD supports it too and the best type of V-Sync, virtual V-Sync, is supported by every computer with an Intel i5 or i7 on a Z77 or Z68 motherboard regardless of what graphics card(s) they have.

The GTX 660 Ti, with current drivers, can't even beat the Radeon 7870 on average. Even the 800MHz reference Radeon 7950 (that other 7950 reference) bests the GTX 660 Ti on average (only by a little, but still) and the Radeon 7950 Boost is fast enough to compete with the GTX 670, a considerably better card than the GTX 660 Ti (especially for overclocking, something that the Radeon 7950 can still best even the GTX 680 in nine times out of ten).

I often talk about Tech Report's testing methods quite often... I support them. They don't hide things such as micro-stutter and variable frame rates in the sub second time frames. However, you do not understand what it is nor how it works. It most certainly is affected by software. For example, the single most impacting pieces of software for a graphics card are the drivers. These pieces of software can make huge differences. Catalyst beta 12.11 is not nearly the newest AMD driver. There have been seven releases since it, a few of which were significant improvements over the first 12.11 driver (more so than even the first 12.11 driver compared to 12.8). That Tech Report test was done without RadeonPro as well as comparing a mere very slightly overclocked 7950 Boost model to by far the best GTX 660 Ti model of them all, not exactly a fair comparison. Furthermore, it only tested a few games, so it can't really be used as a concrete average anyway.


You're all making many mistakes about memory and such. First off, the calculated memory bandwidth simply isn't the real memory bandwidth. It is merely the theoretical memory bandwidth. Nvidia's memory controllers are slightly more efficient per theoretical memory bandwidth than AMD's, but AMD's higher width is still enough to give them much more bandwidth desptie the lower frequencies. AMD's memory can overclock about as well as Nvidia's in frequency on average andthat gives them a huge advantage in memory bandwidth, part of AMD's advantage with MSAA (which generally is the best AA because it is among the most effective types, can be done in almost all games, and doesn't kill performance) and other memory-bandwidth intensive features.

Increasing texture quality and such also increases the need for memory bandwidth. This is show easily in Tom's review of the GTX 650 Ti where the GTX 650 Ti had a roughly 45% advantage over the GTX 650 on average with low texture settings, yet that advantage dropped to about 37% with higher texture settings. You may argue that it's not a large difference and you may be right in that, but that is a nearly ten percent difference. There was no such large discrepancy on the AMD cards.

I'll also point out how most reviews of the Nvidia cards on the internet don't use new AMD drivers (even for their time) while they almost purely use new Nvidia drivers, an obvious attempt to make AMD look worse than they are. Many reviews also go out of their way to use unrealistic settings to make Nvidia look better (and often fail anyway, but that's not the point).

Somewhat basic summary: Radeon 7950 Boost competes with GTX 670, not the GTX 660 Ti, in gaming performance with current drivers.
Even worse for the MSI GTX 660 Ti PE recommendation, it goes without saying that such a recommendation is not the best because MSI PE cards are well-known for their inferior reliability to even reference Nvidia models due to over-voltage without improving card component quality to compensate for that over-voltage.

As a recommendation, Gigabyte's Radeon 7950 3x WindForce is easily the best card for overclocking in its price range. I last saw it around $300-$308.
m
0
l
December 13, 2012 2:43:26 AM

blazorthon said:
Kepler cards do not support it in the same way as RadeonPro. Look up Tom's recent article on the Radeon 7990 versus the GTX 690 for a little more information on it. RadeonPro is not a mere frame rate limiter, it controls frame-by-frame output better. RadeonPro doesn't necessarily sacrifice performance either. Tom's Radeon 7990 versus GTX 690 article showed this quite clearly.

V-Sync is improves screen tearing at the cost of high input lag. Besides that, AMD supports it too and the best type of V-Sync, virtual V-Sync, is supported by every computer with an Intel i5 or i7 on a Z77 or Z68 motherboard regardless of what graphics card(s) they have.

The GTX 660 Ti, with current drivers, can't even beat the Radeon 7870 on average. Even the 800MHz reference Radeon 7950 (that other 7950 reference) bests the GTX 660 Ti on average (only by a little, but still) and the Radeon 7950 Boost is fast enough to compete with the GTX 670, a considerably better card than the GTX 660 Ti (especially for overclocking, something that the Radeon 7950 can still best even the GTX 680 in nine times out of ten).

I often talk about Tech Report's testing methods quite often... I support them. They don't hide things such as micro-stutter and variable frame rates in the sub second time frames. However, you do not understand what it is nor how it works. It most certainly is affected by software. For example, the single most impacting pieces of software for a graphics card are the drivers. These pieces of software can make huge differences. Catalyst beta 12.11 is not nearly the newest AMD driver. There have been seven releases since it, a few of which were significant improvements over the first 12.11 driver (more so than even the first 12.11 driver compared to 12.8). That Tech Report test was done without RadeonPro as well as comparing a mere very slightly overclocked 7950 Boost model to by far the best GTX 660 Ti model of them all, not exactly a fair comparison. Furthermore, it only tested a few games, so it can't really be used as a concrete average anyway.


You're all making many mistakes about memory and such. First off, the calculated memory bandwidth simply isn't the real memory bandwidth. It is merely the theoretical memory bandwidth. Nvidia's memory controllers are slightly more efficient per theoretical memory bandwidth than AMD's, but AMD's higher width is still enough to give them much more bandwidth desptie the lower frequencies. AMD's memory can overclock about as well as Nvidia's in frequency on average andthat gives them a huge advantage in memory bandwidth, part of AMD's advantage with MSAA (which generally is the best AA because it is among the most effective types, can be done in almost all games, and doesn't kill performance) and other memory-bandwidth intensive features.

Increasing texture quality and such also increases the need for memory bandwidth. This is show easily in Tom's review of the GTX 650 Ti where the GTX 650 Ti had a roughly 45% advantage over the GTX 650 on average with low texture settings, yet that advantage dropped to about 37% with higher texture settings. You may argue that it's not a large difference and you may be right in that, but that is a nearly ten percent difference. There was no such large discrepancy on the AMD cards.

I'll also point out how most reviews of the Nvidia cards on the internet don't use new AMD drivers (even for their time) while they almost purely use new Nvidia drivers, an obvious attempt to make AMD look worse than they are. Many reviews also go out of their way to use unrealistic settings to make Nvidia look better (and often fail anyway, but that's not the point).

Somewhat basic summary: Radeon 7950 Boost competes with GTX 670, not the GTX 660 Ti, in gaming performance with current drivers.
Even worse for the MSI GTX 660 Ti PE recommendation, it goes without saying that such a recommendation is not the best because MSI PE cards are well-known for their inferior reliability to even reference Nvidia models due to over-voltage without improving card component quality to compensate for that over-voltage.

As a recommendation, Gigabyte's Radeon 7950 3x WindForce is easily the best card for overclocking in its price range. I last saw it around $300-$308.


see thats a load of crap im sorry but im sick of everyone putting these stupid claims forward when time and time again the performance benchmarks with or without the updates from amd show the 660ti in similar frame rates as the hd7950 not the hd7950 against gtx 670. this is exactly what im talking about, you put all this forward and do not show me the performance evidence to say that it is true. also the MSI cards are very high quality! + the triple voltage just allows you to overclock the card higher doesn't it? correct me if im wrong on that one. Maybe you've had an unlucky experience with MSI and now you hate them. btw according to linus tech tips the msi n660ti pe is the 660ti to get because of the high overclock ability. looking at the happy customers reviews on newegg for the msi gtx660ti compared to the hd7950's the 660ti appears to have a lot less issues with the card. Not only this, more of the better games favor nvidia cards. lastly, from what I see for the hd7950 to be actually worth it over the 660ti you have to overclock it very high. and what kind of settings are we talking about to make the radeon's look bad?.
m
0
l
a b U Graphics card
December 13, 2012 2:47:42 AM

I would choose a 7950 over a 660ti all day, every day........ and in fact I would choose a 7950 over a 670 as well!
m
0
l
December 13, 2012 2:50:28 AM

instantcoffeenz said:
see thats a load of crap im sorry but im sick of everyone putting these stupid claims forward when time and time again the performance benchmarks with or without the updates from amd show the 660ti in similar frame rates as the hd7950 not the hd7950 against gtx 670. this is exactly what im talking about, you put all this forward and do not show me the performance evidence to say that it is true. also the MSI cards are very high quality! + the triple voltage just allows you to overclock the card higher doesn't it? correct me if im wrong on that one. Maybe you've had an unlucky experience with MSI and now you hate them. btw according to linus tech tips the msi n660ti pe is the 660ti to get because of the high overclock ability. looking at the happy customers reviews on newegg for the msi gtx660ti compared to the hd7950's the 660ti appears to have a lot less issues with the card. Not only this, more of the better games favor nvidia cards. lastly, from what I see for the hd7950 to be actually worth it over the 660ti you have to overclock it very high. and what kind of settings are we talking about to make the radeon's look bad?.

Dude really stop posting threads like this in forum.Forums are full of fanboys.Just buy the card you think is good.There are plenty of benchmark.If you want more honest reviews read newegg customers reviews.
m
0
l
December 13, 2012 3:04:02 AM

cooldudesubho said:
Dude really stop posting threads like this in forum.Forums are full of fanboys.Just buy the card you think is good.There are plenty of benchmark.If you want more honest reviews read newegg customers reviews.
yeh I think your right. all I wanted was to find out why everyone everyone has the "7950 hands down " saying when most benchmarks , newegg reviews, and youtube gameplay videos say otherwise. all I got was more of the same "HD7950 is better!".
m
0
l

Best solution

a c 144 U Graphics card
December 13, 2012 3:27:58 AM

relax guys :) .

@ instantcoffeenz

this review from tech report might interest you:

http://techreport.com/review/23981/radeon-hd-7950-vs-ge...

http://techreport.com/review/24022/does-the-radeon-hd-7...

http://techreport.com/review/24051/geforce-versus-radeo...

personally i take that both card are good. in your case some people will choose 7950 because it is a bit cheaper and it has 1GB more VRAM than GTX660 Ti which is matter if the person consider to jump into multi monitor gaming.

Share
December 13, 2012 3:32:50 AM

yup have read that review. Have read pretty much all of them I'm only playing at 1680x1050 but soon I will play at 1920x1080p single monitor so im going to go with the graphics card that will give me the best performance with that which is the MSI N660 Ti Power Edition. Thanks for all your help guys you made it very clear on what the better card is!
m
0
l
December 13, 2012 3:33:42 AM

Best answer selected by instantcoffeenz.
m
0
l
December 13, 2012 5:48:11 AM

Hey i was the first one to link you to that review.i deserve best answer.You naughty boy.Good luck with your card.Hope you have intel iseries CPU though.
m
0
l
a c 144 U Graphics card
December 13, 2012 6:06:26 AM

lol. don't get mad dude :kaola: 
m
0
l
December 13, 2012 6:25:22 AM

cooldudesubho said:
Hey i was the first one to link you to that review.i deserve best answer.You naughty boy.Good luck with your card.Hope you have intel iseries CPU though.
oh sorry mate! much appreciated for your answer my friend! yes I do I have the intel i5 3570k and from what I see this card shines on this cpu aswell!
m
0
l
!