Sapphire HD 7950 Vapor-X 3 GB

[INSTG8R]

So you are saying, we are the only ones who tested it, and you like the results of people better who use invalid test methods. testing whole system's power consumption with a $10 power meter is not really reliable

edit: i'm crawling out of bed to restest idle just for you

That's Dedicaton

 

[Jacob XP]

W1zzard, actually I like your and Kitguru's test quite a lot. Great read. I have no idea what instruments the various websites are using to test. I assumed they know what they're doing and have the right tools. I might be wrong there.

I have done some basic calculations and the higher idle power consumption Techpowerup.com is reporting would cost me a little less than 10 Euro extra a year. Not a dealbreaker although the money adds up when I keep the card for a couple of years.

I have also thought about buying a smaller card like a Radeon 7770, 7850 or 7870. Or a Geforce 660. If I chose one of those I could save a few euro's more. But I decided to spend some of my hard earned money on a Sapphire Vapor-X 7950 even if I'm not that big gamer. I want a good and quiet card and the offer I got was so good that I could not say no

Thanks a lot for doing a second test on the power consumption. You are crazy in a good way for taking time in doing this. I think I will do some experiments if I can get a stable card while undervolting it a bit. I'm looking for forward for next weekend. Right now I have a PC based on a 7 year old AMD Athlon 64 with 1 GB RAM and a 5670 graphics card !!! The new graphics card together with a new Ivy Bridge setup will be a HUGE change

 

[W1zzard]

I assumed they know what they're doing and have the right tools

not speaking of anyone specifically, but almost no reviewer i know has has had scientific training. testing tools range from mythbusters pseudo-scientific to better than what hardware manufacturers use. and then you still need to use them correctly. so yes, question the testing, and yes, question tpu, too

regarding noise levels, if that is your main priority then you must look at the asus 660 ti dc II. it's the best high performance low noise card on the market today, no other card comes even close.

again, not saying the sapphire card is a bad choice, you might not even hear the difference, especially when coming from an older noisy system

 

[INSTG8R]

W1zzard, actually I like your and Kitguru's test quite a lot. Great read. I have no idea what instruments the various websites are using to test. I assumed they know what they're doing and have the right tools. I might be wrong there.

I have done some basic calculations and the higher idle power consumption Techpowerup.com is reporting would cost me a little less than 10 Euro extra a year. Not a dealbreaker although the money adds up when I keep the card for a couple of years.

I have also thought about buying a smaller card like a Radeon 7770, 7850 or 7870. Or a Geforce 660. If I chose one of those I could save a few euro's more. But I decided to spend some of my hard earned money on a Sapphire Vapor-X 7950 even if I'm not that big gamer. I want a good and quiet card and the offer I got was so good that I could not say no

Thanks a lot for doing a second test on the power consumption. You are crazy in a good way for taking time in doing this. I think I will do some experiments if I can get a stable card while undervolting it a bit. I'm looking for forward for next weekend. Right now I have a PC based on a 7 year old AMD Athlon 64 with 1 GB RAM and a 5670 graphics card !!! The new graphics card together with a new Ivy Bridge setup will be a HUGE change

I wouldn't be nickle and diming power consumption with the AMD's That Zerocore is some slick tech.

I have Aida64 monitoring "Everything" on my G15's LCD. I of course have my GPU temp and Fan Speed on my "Main Screen" I STILL think it's SUPER COOL BEANS!! When the cards info DISAPPEARS off my screen. ZEROCORE!! The Card literally shiut right down, Fans and all.

W1z I don't think your monitoring that are you? It's certainly not the Idle figures is it?

 

[W1zzard]

ZeroCore is about 1 Watt for all cards. Great technology. It has a few minor gotchas like sometimes it doesn't engage when stuff is drawing to the screen, but if you let your PC running overnight without actively using it, it will make a huge difference. Better choice is to just turn off the PC, but lately I'm too lazy for that, too.

Idle in our testing not only reflects 100% idle, but also most desktop work like office, browsing etc.

 

[INSTG8R]

ZeroCore is about 1 Watt for all cards. Great technology. It has a few minor gotchas like sometimes it doesn't engage when stuff is drawing to the screen, but if you let your PC running overnight without actively using it, it will make a huge difference. Better choice is to just turn off the PC, but lately I'm too lazy for that, too.

Idle in our testing not only reflects 100% idle, but also most desktop work like office, browsing etc.

Yeah that's what I thought. Yep it still impresses me. Can't say I have had any glitches I can think of, But yeah I go thru phases were it gets left on for a week. Then I may only use
it couple hours the whole next week

It's really an underrated/under advertised feature. I mean sure it pushes a few more watts under load than it's NV competitor but when it can literally just go to sleep and use next to nothing Nvidia has nothing to counter that with and those couple extra watts aren't as significant if you say tallied up consumption for the cards for say a month of being left on 24/7 and getting regular use.

 

[Casecutter]

Since we're all baring our souls here…

Here's one thing I'd like all reviews to capture nowadays when it come to power consumption. I'd like to see the average watts of the gaming run when actually playing, or doing the FpS runs. While each card runs that game the watt's are collected, averaged and presented. While then that can along with each/all the titles be averaged, providing a more tangible power consumption during all gaming.

With all the dynamic clock and boost being exploited it's hard with one title or synthetic stress to paint a clear picture of performance to watt. I've seen that when done like that there's not a clear winner (brand) provides any huge advantage anymore in the competing class.

More Data… mo Data.. give me data! Never anything wrong with that!

 

[W1zzard]

Here's one thing I'd like all reviews to capture nowadays when it come to power consumption. I'd like to see the average watts of the gaming run when actually playing, or doing the FpS runs. While each card runs that game the watt's are collected, averaged and presented. While then that can along with each/all the titles be averaged, providing a more tangible power consumption during all gaming.

Would that give a correct picture considering that some cards will run a given bench at more or less FPS? Some benches run constant time, some run longer on a slower card. Thoughts?

I think you are asking for energy (Joules) consumes. Watts are defines as Joules per second, add up and multiple by bench runtime -> Joules. But same thing

Edit: Or are you looking for power consumption graph over time? Will be kinda difficult to present with all our benches at 4 resolutions

thats gpu voltage for all tests at just 1 resolution, image that 4x as crowded.. will be hard to see much

 

[Casecutter]

I might say this could just be a separate review that breaks-out just a few cards to see how relative the power consumption of these new dynamic/boost clock react. Just a few of the higher end cards (670/680 and 7950/7970) and then at just 1920x, 2560x, and 5760x.

Would that give a correct picture considering that some cards will run a given bench at more or less FPS?

I would think it would isn't that a true performance to power consumption? How is Perf/watt you shown now calculated? Is it just "Summary FPS" versus your "Peak" (Crysis 2 at 1920x1200)? If that's the case your Perf/watt is not a not at all relative now-a-days. Or that Perf/watt just what the card(s) running (Cyrsis 2 @ 1920x) which provides just one true collaborating value.

Some benches run constant time, some run longer on a slower card. Thoughts?

Yes straight up FPS doesn’t really take into consideration in time/duration aspect? If a card takes 6 more seconds, but is way faster is that a manifested by FpS… not always. I suppose that’s why other review’s plot the frame times, latency, and 99th percentile.

Or are you looking for power consumption graph over time? Will be kinda difficult to present with all our benches at 4 resolutions

Nope, but can’t that graph be output as the average maybe better the "mean" and that would give a relative number against another card ran in the same test?

 

[nanonimo]

hello, finally i created an account in this fabulous page. I read your page daily. I decided to buy this card a few weeks ago becouse it was the best option for the money i could spend in my area. I don't know if i'm doing something wrong with it, today i switched to the second bios to rise the clocks speed to 950 mhz, but it still has the clock speed at 850 mhz and it should has 950 mhz by default while gaming, doesn't it?

Thank you, and sorry for my english i'm from spain

 

[tt_martin]

@nanonimo, everything is ok. 850Mhz is your core clock, but 950Mhz is Boost Clock. It enables when it's necessary. Check GPU-Z while gaming.

 

[nanonimo]

@nanonimo, everything is ok. 850Mhz is your core clock, but 950Mhz is Boost Clock. It enables when it's necessary. Check GPU-Z while gaming.

Ok thank you! this are the images of the gpu-Z, are they correct? i never saw the clock speed at 950, so it means the gpu has never needed higher clock speed... i play at full hd at maximoun detail.

 

[loc]

Just installed my 7950 vapor-x and have to say I am disappointed at the noise and the heat. Unigine Heaven run raises the temps to 85C when the whole system crashes and fan blows at 70% which makes really nasty noise.. not sure whats up with the card. I've not even used the overclock button.

edit: I found the culprit. When the amd's gpu boost was working it also increased the gpu voltage up to 1.25v which caused heating up. As that happened card also hit the power limit and lowered the gpu clocks as well resulting worse fps and massive heat. I turned on the cards gpu boost button which essentially disables the amd gpu boots and increases the clocks to fixed 950MHz. After this the gpu speed remains constantly at 950MHz with 1.056v and temperatures below 65C with auto fan. Seems amd's driver based gpu boost needs still some work.

 

[billcat479]

Billcat- that was too long of a post to read, but on a point- performance numbers will always have variables whether they are .0001% or higher. What is posted is the average a user would get from such a system. I understand most don't read the entire review but it would be recommended to because system spec details, that is where the variables come in I recall the test bed was a core i7 920 but it switched to a core i7 2600 several months back.

To you and everyone, I really am very sorry about my long posts. I can't seem to help it though. I got a fixation that I have to put everything up there so everyone knows why I have this or that view on a topic.
Again, very very sorry I am a radical poster.. I think it was because my brain is a bit out of whack after I got hit by a car and got a lot of resulting brain damage so I'm a flake at times.
It also blew my IQ from the 150's to the 130's so to me I'm stupid. I used to work for Intel at their Fab4 plant in Beaverton, Or. and now I eat oxycodon (for those that don't know it's a very strong narcotic) for horrible headaches and just play games and forget everything. I'm a lot like that guy in that House show. A missrable SOB.
The only fun part of forgetting all the time is I get to re-read the lord of the rings hundreds of times and enjoy it as much each time. I think I have read Ringworld 2 or 3 hundred times as I just love that story too.. hehehe.... Weird huh? Or the Foundation series or the Lensman series ect ect...............

 

[mediasorcerer]

I had a few issues with the ghz bios for 7950 [it worked but ran hotter and bsod occasionally], using original atm.
Great review and thankyou!!!
It really seems like the results are often dependent on which brand [nv or amd] the game is optomised for[far as i can see], but those are some great results in metro and sniper, glad i got one even though its ref, it was cheap, and i don't think having 3gb mem and 384bit bus hurts either.

 

[mediasorcerer]

To you and everyone, I really am very sorry about my long posts. I can't seem to help it though. I got a fixation that I have to put everything up there so everyone knows why I have this or that view on a topic.
Again, very very sorry I am a radical poster.. I think it was because my brain is a bit out of whack after I got hit by a car and got a lot of resulting brain damage so I'm a flake at times.
It also blew my IQ from the 150's to the 130's so to me I'm stupid. I used to work for Intel at their Fab4 plant in Beaverton, Or. and now I eat oxycodon (for those that don't know it's a very strong narcotic) for horrible headaches and just play games and forget everything. I'm a lot like that guy in that House show. A missrable SOB.
The only fun part of forgetting all the time is I get to re-read the lord of the rings hundreds of times and enjoy it as much each time. I think I have read Ringworld 2 or 3 hundred times as I just love that story too.. hehehe.... Weird huh? Or the Foundation series or the Lensman series ect ect...............

Forgive for off topic here pls,

You sound like a clever fella, hope you get over the oxy shit, no offence, but your probably smarter than that aye?

Since you posted this, i'm going to offer you my 2 bob advice , get a project going, show what you can do?
p.s. azimov is gr8.

 

[SonDa5]

This card design has been changed since this review came out.

 

[SonDa5]

I think I just found a major weakness for over clocking with the heat sink design.

Memory makes contact directly with the main Dual X heat sink.


http://tpucdn.com/reviews/Sapphire/HD_7950_Vapor-X/images/cooler2_small.jpg


I think this design has some win an lose at the same time. This heat sink will get soaked with heat when over clocking the memory and gpu core with this design.

 

[SPiDERR]

Software for overclocking

Can someone recommend some software for overclocking (overclocking and testing of stability)? I recently bought the Vapor-X 7950 OC and I am preparing my self for Crysis 3. I tried to overclock via CCC but the game was pretty much unstable. The system crashes messed up my mainboard BIOS.
So, please, can someone give me some suggestion how to safely overclock this GPU? THX

 

[meran]

Can someone recommend some software for overclocking (overclocking and testing of stability)? I recently bought the Vapor-X 7950 OC and I am preparing my self for Crysis 3. I tried to overclock via CCC but the game was pretty much unstable. The system crashes messed up my mainboard BIOS.
So, please, can someone give me some suggestion how to safely overclock this GPU? THX

HI i wanted to buy this card but i saw some reviews suggesting high vrm temps because of bad contact cooling http://www.xtremesystems.org/forums/showthread.php?284714-Sapphire-HD7950-3GB-Vapor-X-Review

i need someone to confirm