Sign in with
Sign up | Sign in
Your question

7870 Metro 2033 68FPS wth....

Last response: in Graphics & Displays
Share
September 28, 2012 9:41:02 PM

I was on Youtube today watching reviews of video cards as I always do and I came across the Iceq 7870 review and at 1920x1200 at max settings that card was scoring around 68 FPS and with a 100 Mgz overclock 73 FPS. I recall just a few months ago SLI GTX 580's in reviews hitting only 55FPS now I am sure drivers have since flipped that number. Have we really come this far? who else thinks that is kinda nuts for a mid end card to hand with a GTX 670 in metro 2033 that is unheard of.
I am no amd fanboy in fact I like Nvidia more but holy hell if that didnt shock me a bit. :o  :o  :o 
a b U Graphics card
September 28, 2012 10:54:00 PM

What were the game settings? I bet the game settings were not maxed out!
a c 217 U Graphics card
September 28, 2012 11:06:51 PM

Where you benchmark also matters. If you use the games benchmark tool, that 7870 would be lucky to get 30 FPS. There is a good chance that 580 SLI review used the benchmark tool.
Related resources
a b U Graphics card
September 29, 2012 12:42:04 AM

mrlime09 said:
I was on Youtube today watching reviews of video cards as I always do and I came across the Iceq 7870 review and at 1920x1200 at max settings that card was scoring around 68 FPS and with a 100 Mgz overclock 73 FPS. I recall just a few months ago SLI GTX 580's in reviews hitting only 55FPS now I am sure drivers have since flipped that number. Have we really come this far? who else thinks that is kinda nuts for a mid end card to hand with a GTX 670 in metro 2033 that is unheard of.
I am no amd fanboy in fact I like Nvidia more but holy hell if that didnt shock me a bit. :o  :o  :o 


Metro 2033 likes GCN GPUs and memory bandwidth and regardless of that, chances are that the tests that you're referring to were not the same given that GTX 580 SLI is something like 70-80% faster, on average, than the 7870.
September 29, 2012 2:34:55 AM

The test was From Overclockers3d.com Tom is very trusted and it was live gameplay with the GTX 580's. The guy who said the 7870 was maxing out Metro 2033 at 1920x 1200 at max settings I am not sure who he is I do know that he use to work for tiger direct and seems to now have like a tech Vodcast show or something.
September 29, 2012 2:49:35 AM

foundout who the guy is He is that dude from Teksyndicate.com again another pretty trusted Website as they review everything from apple products to pc gear and what not.
a c 217 U Graphics card
September 29, 2012 3:03:32 AM

Where they benchmark (as in, what map and area) is going to have significant differences in performance. The exact settings used will also vary resulting in very different results.

Metro 2033 has a wide variety of areas with very different performances. Unless they use the same benchmark, you cannot compare them.
September 29, 2012 4:27:33 AM

Watched another Teksyndicate.com video card Review and he reviewed a GTX 680 4g card and with a stock clock he got 42.20 fps in Metro 2033 everything at Max and at 1920x1200, and with two in sli he 67.4fps. Here's the deal is GK104 even a high end GPU? I mean I know it spanks most games but a GTX at $499.99 for a basic model should never ever for any reason get beat by a card for half it's price aimed at a budget mid end market. http://www.youtube.com/watch?v=QIJIhv3mqNw&feature=plcp http://www.youtube.com/watch?v=UkOZhwutP-o&feature=plcp
a b U Graphics card
September 29, 2012 4:38:50 AM

I've said it before and I'll say it again. Metro 2033 likes AMD Radeon 7xxx cards. If you look at most other games, the 670 will beat the 7870 significantly except in some older and/or not very intensive games where you can load up the MSAA very high and the 7870's superior MSAA scaling might be enough to give it the lead.
September 29, 2012 4:51:34 AM

I am not a Metro 2033 Fan at all I find the game boring but If I paid $500 dollars or more like $550 for a 4gb GTX 680 and I couldn't max out a game like Metro 2033 at higher fps then 44fps I'd be pissed off. Something isn't right here I am starting to see with updated drivers the 7xxx Cards pull away from the 6xx Nvidia cards. I wonder if Nvidia did the right thing going with the GK104. Not every one overclocks but from what I have seen thus far with overclocks every 6xx card is getting spanked. mind you that you have to overclock the *** out of most 7xxx cards to match or beat 6xx cards but for me this is troubling.
September 29, 2012 4:56:24 AM

Like a 7870 overclocked can pull far ahead of a 66ti, a 7950 can pull ahead of 670, and the same goes for the 7970 it can pull ahead of a 680. To make matters worse most of these overclocks are easy to do and the fact that the 6xx cards for the most part do not overclock as high or as well as the 7xxx cards I kind wonder what the heck is going on.
a c 217 U Graphics card
September 29, 2012 5:02:57 AM

If you want to actually get a real comparison, get one that benchmarks all the cards in question, side by side. Example: http://www.tomshardware.com/reviews/radeon-hd-7870-revi...

A single 580 performs roughly the same as a single 7870 in Metro 2033. You should also note how the 7870 only gets near the FPS you are talking about at medium settings.

Here are more:
http://www.tomshardware.com/reviews/geforce-gtx-660-ti-...
http://www.anandtech.com/bench/Product/517?vs=548
September 29, 2012 5:10:16 AM

The video I watched was posted on the 24th of this month with the latest drivers those reviews are out of date and no longer reflect the current performance.
a c 217 U Graphics card
September 29, 2012 5:12:18 AM

mrlime09 said:
The video I watched was posted on the 24th of this month with the latest drivers those reviews are out of date and no longer reflect the current performance.


You watched a single video that showed 1 card and tried to compare it to another video or benchmark that most likely used very different settings. Driver changes would not make such radical differences. You might see a 10% difference at most. The only way you can compare results is if you benchmark both in the same way with only the video card being different. These 2 benchmarks clearly were not compared in the same manner.
a b U Graphics card
September 29, 2012 5:14:01 AM

Anything with Catalyst 12.7 and up is probably pretty close to modern performance, but that's a fair point.
a b U Graphics card
September 29, 2012 5:14:10 AM

bystander said:
You watched a single video that showed 1 card and tried to compare it to another video or benchmark that most likely used very different settings. Driver changes would not make such radical differences. You might see a 10% difference at most. The only way you can compare results is if you benchmark both in the same way with only the video card being different. These 2 benchmarks clearly were not compared in the same manner.


Driver updates can and have made greater than 10% differences, but you're still undoubtedly correct about everything else that you said here.
a c 217 U Graphics card
September 29, 2012 5:16:02 AM

luciferano said:
Driver updates can and have made greater than 10% differences, but you're still undoubtedly correct about everything else that you said here.

Clearly not in this case:

a b U Graphics card
September 29, 2012 5:28:01 AM



Looking at averages of a single game doesn't prove much. For example, any game or games that wasn't properly supported by a previous driver and is supported by a new one can have a much more than 10% performance increase (something that I guarantee to have happened, especially with CAPs too, with several games over the last few months). If you compare Catalyst 12.8 or 12.9 to the older Catalyst drivers, then even the average performance increase might breach 10%. Besides, across a wide array of games, the average performance increase of 12.7 from the predecessors might be over 10%. Again, a single game being tested proves little.
a c 217 U Graphics card
September 29, 2012 5:32:56 AM

luciferano said:
Looking at averages of a single game doesn't prove much. For example, any game or games that wasn't properly supported by a previous driver and is supported by a new one can have a much more than 10% performance increase (something that I guarantee to have happened, especially with CAPs too, with several games over the last few months). If you compare Catalyst 12.8 or 12.9 to the older Catalyst drivers, then even the average performance increase might breach 10%. Besides, across a wide array of games, the average performance increase of 12.7 from the predecessors might be over 10%. Again, a single game being tested proves little.


We are talking about this one single game; Metro 2033. It shows almost no improvement using drivers 12.7. All benchmarks that compare the 7870 with a 580 (not even 580 sli), are showing them close and usually showing a single 580 as faster. This is contradictory to the OP's assertion that the 7870 was faster than 580's in SLI. As I've tried to show, with several benchmarks, is that you have to compare using the same settings. You'll notice every site has different averages with the same card as well, further showing that you can't compare results not compared the same.

As far as the idea of drivers improving games more than 10%, that can happen in rare exceptions, but normally only because there were bugs that had to be fixed. Metro 2033 was not one of those games.
a c 186 U Graphics card
September 29, 2012 5:40:20 AM

The problem is no proof. There is not enough reliable professional reviews showing what the new drivers have actually done.So there will always be people that question it until this has been done.By many professional reliable review sites.There here no way of proving it until then sadly.
September 29, 2012 5:47:10 AM

Screw it I am going to have to bite the bullet and just buy the 7870 iceq and show you guys real results. To be honest I have a hell of are hard time even thinking this card can max out Metro 2033 at max at 1920x1200. I am going to turn of HT on my 990x and overclock it to 4.2GHz as so the card has ample cpu power behind it record a game with fraps and see what happens. I am buying the card one the first I should have it by the 3rd or 4th let's see what happens.
September 29, 2012 5:51:07 AM

I mean you guys can watch the videoas well. I honestly doubt that dude has anything to gain about lying about a Iceq 7870. for the record that guy can and doe's play with everything new and old. but anyways I will bite the bullet on this one as it is only $220 dollars.
a c 217 U Graphics card
September 29, 2012 5:55:16 AM

bigcyco1 said:
The problem is no proof. There is not enough reliable professional reviews showing what the new drivers have actually done.So there will always be people that question it until this has been done.By many professional reliable review sites.There here no way of proving it until then sadly.


I posted 2 sites that compared them.
a b U Graphics card
September 29, 2012 5:55:36 AM

mrlime09 said:
I mean you guys can watch the videoas well. I honestly doubt that dude has anything to gain about lying about a Iceq 7870. for the record that guy can and doe's play with everything new and old. but anyways I will bite the bullet on this one as it is only $220 dollars.


I didn't watch both videos fully yet, but did he actually claim that he ran the exact same test with the exact same hardware other than the graphics card with both the 7870 and 680 reviews?
a c 186 U Graphics card
September 29, 2012 5:56:56 AM

bystander said:
I posted 2 sites that compared them.
Indeed you did i missed that my bad
a c 217 U Graphics card
September 29, 2012 5:57:10 AM

mrlime09 said:
Screw it I am going to have to bite the bullet and just buy the 7870 iceq and show you guys real results. To be honest I have a hell of are hard time even thinking this card can max out Metro 2033 at max at 1920x1200. I am going to turn of HT on my 990x and overclock it to 4.2GHz as so the card has ample cpu power behind it record a game with fraps and see what happens. I am buying the card one the first I should have it by the 3rd or 4th let's see what happens.


The argument was not that it cannot play at "max" settings (as long as you don't include DoF and PhysX), the argument was it is not faster than two 580's in SLI.

You likely will run into areas where a single 7870 will struggle still. Assuming 1080p or higher resolution.
September 29, 2012 6:20:46 AM

I still doubt that guy lied about the benchmarks Perhaps I can get an aswer out of him what his system setup was and how long did he play the game. I go with what I see and at the moment he indeed claim to have gotten 68fps with MAX everything in Metro 2033 with a 7870. I have no reason to think the guy is lying and or claimed something that isn't so. He did say MAX settings that would assume everything is cranked up. He also claimed a GTX 680 only got 42.20fps in Metro 2033. Once again I will bite the bullet on this and find out and post results. My system should be able to max out the FPS of what that 7870 can push, so it will be cut and dry.
a c 217 U Graphics card
September 29, 2012 6:27:47 AM

mrlime09 said:
I still doubt that guy lied about the benchmarks Perhaps I can get an aswer out of him what his system setup was and how long did he play the game. I go with what I see and at the moment he indeed claim to have gotten 68fps with MAX everything in Metro 2033 with a 7870. I have no reason to think the guy is lying and or claimed something that isn't so. He did say MAX settings that would assume everything is cranked up. He also claimed a GTX 680 only got 42.20fps in Metro 2033. Once again I will bite the bullet on this and find out and post results. My system should be able to max out the FPS of what that 7870 can push, so it will be cut and dry.


I never said he lied. I said that you can't compare one persons benchmark results to another random benchmark.

The 580's in SLI likely used different settings. A lot of people say Max settings, but for some people that includes PhysX and DoF, and for others it doesn't. For some benchmarks, they compare in the game in some random tunnels which aren't very demanding. Others will compare inside the outposts or the benchmark tool, which are FAR more demanding. Some include AAA, some MSAAx4. Then you have factory OC'ed cards, and manually OC'ed cards, compared to factory clocked cards.

There are just so many ways the benchmarks can be different without lying.
a c 186 U Graphics card
September 29, 2012 6:34:24 AM

bystander said:
I never said he lied. I said that you can't compare one persons benchmark results to another random benchmark.

The 580's in SLI likely used different settings. A lot of people say Max settings, but for some people that includes PhysX and DoF, and for others it doesn't. For some benchmarks, they compare in the game in some random tunnels which aren't very demanding. Others will compare inside the outposts or the benchmark tool, which are FAR more demanding. Some include AAA, some MSAAx4. Then you have factory OC'ed cards, and manually OC'ed cards, compared to factory clocked cards.

There are just so many ways the benchmarks can be different without lying.
+1
September 29, 2012 7:32:41 AM

you know what I agree with you bystander. I am just confused now as to if the GK104 is really a high end GPU now. If Logan played through the same areas of the game and scored a higher score with a lesser card then to me that speaks wonders. I really have my doubts but like I said I am going to buy one and figure this out myself.
a b U Graphics card
September 29, 2012 7:41:16 AM

The GTX 670 competes with the Radeon 7950 and 7970. GK104 is high-end. Nvidia abandoned hot-clocking with the Kepler generation in addition to cutting down the architecture and the die shrink, so they managed to change relative GPU positions in the markets. GK110 (there never even was a GK100, probably for this reason) was sent only to the high-end compute-oriented professional markets where the huge profits margins would offset the cost of making the cards based off of the huge GPU and Nvidia bumped up each successive GPU up in position. GK104 has triple the CUDA cores of the GTX 580's GF110 (1.75 times if you account for hot clocking) and it performs where it should, roughly on-par with the former dual-GPU flagship (GTX 480 performed roughly on-par with the GTX 295, etc.). GK104 is high end.
September 29, 2012 1:17:20 PM

All I have to do now is test this for myself since I have fraps and the ever boring Metro 2033. I will post 2 YouTube videos of me playing straight from my pc and a second over the shoulder video. I think Logan meant over all performance of both cards in that game and that a 7870 for whatever reason does indeed beat a 4gb GTX 680 in Metro 2033 with MAX settings which everything is on ultra. I think I am going to prove something but I am not sure what. I will find out Logan's Msaa settings ect.
a c 217 U Graphics card
September 29, 2012 2:52:19 PM

mrlime09 said:
All I have to do now is test this for myself since I have fraps and the ever boring Metro 2033. I will post 2 YouTube videos of me playing straight from my pc and a second over the shoulder video. I think Logan meant over all performance of both cards in that game and that a 7870 for whatever reason does indeed beat a 4gb GTX 680 in Metro 2033 with MAX settings which everything is on ultra. I think I am going to prove something but I am not sure what. I will find out Logan's Msaa settings ect.


Don't forget that Windforce was clocked to 1200 Mhz compared to its factory OC of 1000 Mhz, compared to 900ish Mhz of a Stock 7870.

With over a 30% OC, of course it moves up to the charts in power, but here is the thing, you can OC other cards too. If you compare a 7870 at 1200 Mhz, it is going to compare to a 680 at stock.
September 29, 2012 3:22:00 PM

mrlime09 said:
you know what I agree with you bystander. I am just confused now as to if the GK104 is really a high end GPU now. If Logan played through the same areas of the game and scored a higher score with a lesser card then to me that speaks wonders. I really have my doubts but like I said I am going to buy one and figure this out myself.

WTF are you talking about?OP is yet to link any video showing whatever benchmark he is talking about.
September 29, 2012 4:10:24 PM

Calm down there little guy things will be ok. Most people with rather low IQ's tend to start getting angry over nothing. I did post a video watch it ok little guy and take a break go outside an get some sun no need to get all crazy.
September 29, 2012 5:18:28 PM

bystander said:
Don't forget that Windforce was clocked to 1200 Mhz compared to its factory OC of 1000 Mhz, compared to 900ish Mhz of a Stock 7870.

With over a 30% OC, of course it moves up to the charts in power, but here is the thing, you can OC other cards too. If you compare a 7870 at 1200 Mhz, it is going to compare to a 680 at stock.



your showing your true colors with that statement 900ish? here is a ref card http://www.newegg.com/Product/Product.aspx?Item=N82E168... they are all clocked at 1000mhz stock. http://www.anandtech.com/show/5625/amd-radeon-hd-7870-g...

All I wanna do is see if this can be done or not so I am and I will find out. also just so you know the amd drivers have been 12.8 for a little while now and not 12.7
a b U Graphics card
September 29, 2012 7:28:37 PM

bystander said:
Don't forget that Windforce was clocked to 1200 Mhz compared to its factory OC of 1000 Mhz, compared to 900ish Mhz of a Stock 7870.

With over a 30% OC, of course it moves up to the charts in power, but here is the thing, you can OC other cards too. If you compare a 7870 at 1200 Mhz, it is going to compare to a 680 at stock.


1.00GHz is the stock reference frequency of a 7870. Good 7870 models can overclock probably a little beyond a stock reference GTX 680.
a c 217 U Graphics card
September 29, 2012 11:04:06 PM

luciferano said:
1.00GHz is the stock reference frequency of a 7870. Good 7870 models can overclock probably a little beyond a stock reference GTX 680.


If that is the case, it probably won't OC as well as I thought, rather than having the OC give you a 30% boost as I initially thought, the OC will only boost it by about 20%. The reference 680 has the same stock clocks, and also can hit well over 1200 mhz.

The reason people are starting to get edgy is that you are not taking the facts that have been shown, and continue to go on and on about some comment said in a video review. The fact is, at stock settings, the 7870 is a little slower than a single 580.
a b U Graphics card
September 29, 2012 11:43:54 PM

bystander said:
If that is the case, it probably won't OC as well as I thought, rather than having the OC give you a 30% boost as I initially thought, the OC will only boost it by about 20%. The reference 680 has the same stock clocks, and also can hit well over 1200 mhz.

The reason people are starting to get edgy is that you are not taking the facts that have been shown, and continue to go on and on about some comment said in a video review. The fact is, at stock settings, the 7870 is a little slower than a single 580.


Is that second paragraph targeted at OP or me? I ask because you didn't specify and it doesn't make sense if its targeted at me. Also, you're forgetting about memory overclocking. It is not irrelevant, it makes a difference and should be accounted for too.
a c 217 U Graphics card
September 30, 2012 12:08:22 AM

luciferano said:
Is that second paragraph targeted at OP or me? I ask because you didn't specify and it doesn't make sense if its targeted at me. Also, you're forgetting about memory overclocking. It is not irrelevant, it makes a difference and should be accounted for too.


All extra factory OCing done just goes further to show that OCing is the only part responsible for closer to 580/680 performance. It also means that those cards, when OC'ed will have more OC head room beyond that of the 7870's with factory OCing.

I was referring to the OP, but you seem to be wanting to support his claims, but more reluctant. Look at all the benchmarks. A single 7870 does not come close to 580's in SLI. With enough OCing, which that video commentator did mention he increased the clocks to 1200 mhz, it was able to get near the 680, which I do believe can be achieved, but add those OC's to the 680, which are also capable of 1200mhz, and also start at 1000mhz, you will see the differences remain pretty close to that of a stock 7870. And don't forget, now that you bring up memory, the 680's are hugely underclocked on the memory side compared to AMD memory clocks. They OC by +500 Mhz on many 680's.
a c 217 U Graphics card
September 30, 2012 12:50:54 AM

Let's just put this to rest with these benchmarks of the Gigabyte HD 7870 OC edition Windforce, the card the OP is talking about. In the chart, probably due to the huge name, it is listed as "GBT HD 7870 2GB OC", it's highlighted, so it's hard to miss.

Here is the article, so you can read through it yourself, to see it is the Windforce, even though it is not listed in the title.
http://www.guru3d.com/articles_pages/gigabyte_radeon_hd...
Head on over to the next page where we'll meet and greet Pitcairn, aka the Radeon HD 7800 series, and specifically the GV-R787OC-2GD aka Gigabyte Radeon HD 7870 OC WindForce 3x edition.



You can also get an idea of how two 580's in SLI would do by looking at the 590 near the top of the chart, as the 590 is two 580's in SLI on a single PCB, but slightly underclocked, so it's actually slower.
a b U Graphics card
September 30, 2012 1:05:09 AM

bystander said:
All extra factory OCing done just goes further to show that OCing is the only part responsible for closer to 580/680 performance. It also means that those cards, when OC'ed will have more OC head room beyond that of the 7870's with factory OCing.

I was referring to the OP, but you seem to be wanting to support his claims, but more reluctant. Look at all the benchmarks. A single 7870 does not come close to 580's in SLI. With enough OCing, which that video commentator did mention he increased the clocks to 1200 mhz, it was able to get near the 680, which I do believe can be achieved, but add those OC's to the 680, which are also capable of 1200mhz, and also start at 1000mhz, you will see the differences remain pretty close to that of a stock 7870. And don't forget, now that you bring up memory, the 680's are hugely underclocked on the memory side compared to AMD memory clocks. They OC by +500 Mhz on many 680's.


I've been telling OP this whole time that a 7870 does not come near GTX 580 SLI. At no point did I say that OP was correct about this and I could just go up to one of my previous posts here where I specifically stated that a 7870 (on average) performs around where the 580 does and that an overclocked 7870 can reach for a little above stock reference 680 performance.

You're insane if you're telling me that Kepelr has underclocked memory compared to AMD. No reference Radeon 7xxx card has 1500MHz memory except for the Radeon 7970 GHz Edition, but almost every Kepler GTX card has ~1500MHz memory.

At no point did I claim that a 680 can't be overclocked let alone overclocked far better than a 7870. I specifically stated that an overclocked 7870 can keep up with a stock reference 680, not an overclocked 680.

Also, in games such as Metr0 2033 and a lot of older games where you pump up the MSAA to 8x, a 7870 can in fact compete well with the 680 and the 670 because it has greater MSAA efficiency. On a general rule, the 670/680 are faster and instead compete with the 79xx cards and I never said otherwise, I have already stated that this is true, but the 7870 can in some situations compete quite well with the 670 and 680.
a c 186 U Graphics card
September 30, 2012 1:10:59 AM

Lol!You guys crack me up if overclocked blah blah blah it's only possible not a guarantee thing if your so sure of it good for you do you want a star, badge, hug maybe or what's the deal?
a b U Graphics card
September 30, 2012 1:12:29 AM

bystander said:
Let's just put this to rest with these benchmarks of the Gigabyte HD 7870 OC edition Windforce, the card the OP is talking about. In the chart, probably due to the huge name, it is listed as "GBT HD 7870 2GB OC", it's highlighted, so it's hard to miss.

Here is the article, so you can read through it yourself, to see it is the Windforce, even though it is not listed in the title.
http://www.guru3d.com/articles_pages/gigabyte_radeon_hd...
Head on over to the next page where we'll meet and greet Pitcairn, aka the Radeon HD 7800 series, and specifically the GV-R787OC-2GD aka Gigabyte Radeon HD 7870 OC WindForce 3x edition.

http://www.guru3d.com/miraserver/images/2012/GV-R787OC-2GD/Untitled-13.png

You can also get an idea of how two 580's in SLI would do by looking at the 590 near the top of the chart, as the 590 is two 580's in SLI on a single PCB, but slightly underclocked, so it's actually slower.


The 590 is not just slightly underclocked compared to the 580. It is quite significantly underclocked.
a c 217 U Graphics card
September 30, 2012 1:37:32 AM

luciferano said:
The 590 is not just slightly underclocked compared to the 580. It is quite significantly underclocked.


What one person considers significant is not always considers significant to another. I was trying not to over sell the difference so he doesn't go ballistic on me, if he doesn't consider it significant. Either way, it doesn't come close.

I also was not trying to argue with you, I only was thinking you may agree with the OP, because at every time, you are trying to poke holes into my posts. Look at the benchmarks, they speak for themselves.

As far as the memory thing goes. I only know the 680's and 670's are significantly underclocked to what they can achieve. I also know that in the past, the AMD GDDR5 memory doesn't OC nearly as well. Whether it's due to poor cooling, or a lock of VRM's, I'm not sure. Maybe the 7000 series is different than the 6000 series in that regard, but I haven't seen people OCing their memory by 500 with stock voltages.
a b U Graphics card
September 30, 2012 1:44:32 AM

bystander said:
What one person considers significant is not always considers significant to another. I was trying not to over sell the difference so he doesn't go ballistic on me, if he doesn't consider it significant. Either way, it doesn't come close.

I also was not trying to argue with you, I only was thinking you may agree with the OP, because at every time, you are trying to poke holes into my posts. Look at the benchmarks, they speak for themselves.

As far as the memory thing goes. I only know the 680's and 670's are significantly underclocked to what they can achieve. I also know that in the past, the AMD GDDR5 memory doesn't OC nearly as well. Whether it's due to poor cooling, or a lock of VRM's, I'm not sure. Maybe the 7000 series is different than the 6000 series in that regard, but I haven't seen people OCing their memory by 500 with stock voltages.


I poke holes where you make mistakes just as I'd expect to be done to me when I make mistakes. I don't poke them where you are correct, which is the majority of what you say.
a c 217 U Graphics card
September 30, 2012 1:46:30 AM

luciferano said:
I poke holes where you make mistakes just as I'd expect to be done to me when I make mistakes. I don't poke them where you are correct, which is the majority of what you say.

Cool. Initially I thought so, but started to wonder.
September 30, 2012 4:48:01 AM

Dude's all I said is I wanna find out for myself with my set of test as I do not really think that kinda score with a 7870 can ever really happen. I repeat I plan to test this out as it is hard for me to think Logan would claim the 7870 got 68 fps in Metro 2033 with max settings for no reason. I can see all the charts and numbers in the world but until I test for myself I can't be sure of anything.
!