Sign in with
Sign up | Sign in
Your question
Closed

Is it a good idea to crossfire 7870's?

Last response: in Graphics & Displays
Share
March 5, 2013 12:58:31 PM

Alright guys this is my setup:

Asrock Z77X4 8x 8x Crossfire capability
Intel i5-3570k can overclock up to 4.0ghz
16gb corsair RAM
and a 750W Gold Corsair AX750 (62 amps)

So I'm more than capable of getting a second one but i want to clarify a few things

1. I'm going to buy a 120hz benQ or Samsung (is this a good idea)?
2. I'm going crossfire because i can't sell my 7870 and at the time i purchased my card the 7970's were all $500+ i payed $250 for mine a few months back. So another 7870 seems better to me at the moment...

A few questions:
1. Can i easily disable/turn off crossfire in the CCC or something without taking the card out of the motherboard?
2. My current 7870 never goes over 50 degrees in my case, a second one shouldn't over heat right?
3. If I overclock my cpu to 4.0 ghz, will it be more sufficient or is keeping it at stock fine?
4. is my 8x 8x lane on my motherboard fine?

Thanks for the help :)  :bounce: 
a c 143 U Graphics card
a b K Overclocking
March 5, 2013 1:12:04 PM

Are you going 120Hz because of 3D?
1- Yes.
2- Yes.
3- OC your CPU as high as possible to avoid any bottlenecks, 3570K is great anyways just ensure proper cooling.
4- Yes, no problem at all.
Score
0
March 5, 2013 1:16:46 PM

ilysaml said:
Are you going 120Hz because of 3D?
1- Yes.
2- Yes.
3- OC your CPU as high as possible to avoid any bottlenecks, 3570K is great anyways just ensure proper cooling.
4- Yes, no problem at all.


I'm not planning on 3D gaming directly but maybe, i find those glasses terribly uncomfortable

My main reason is that i get terrible eye strain from my current 60hz monitor which many people have said that using 120hz really helps.

thanks for the reply.
Score
0
Related resources
a c 143 U Graphics card
a b K Overclocking
March 5, 2013 1:30:36 PM

I'm not familiar with what they say, I have very different aspects regarding this but since you're gonna just 2D on it, it's okay.
Score
0
March 5, 2013 2:09:15 PM

ilysaml said:
I'm not familiar with what they say, I have very different aspects regarding this but since you're gonna just 2D on it, it's okay.


Alright thanks for the input!
Score
0

Best solution

a c 216 U Graphics card
a b K Overclocking
March 5, 2013 2:37:34 PM

Krazeee said:
I'm not planning on 3D gaming directly but maybe, i find those glasses terribly uncomfortable

My main reason is that i get terrible eye strain from my current 60hz monitor which many people have said that using 120hz really helps.

thanks for the reply.


I am actually someone who gets nausea unless I have high FPS. I have to play a long time before eye strain kicks in. Both can be symptoms of simulator sickness.

Have you tested to see if 30 FPS makes your eye strain worse than 60 FPS? For me, I get nausea in 1st person games in a few minutes at 30 FPS, but can last quite a lot longer at 60 FPS. That made it clear to me that higher FPS did help, but not everyone is the same.

As far as the glasses go, I found the 3D Vision 1 glasses to be uncomfortable, but the 3D Vision 2 glasses are much improved.

I feel a little uncomfortable recommending crossfire atm.
http://techreport.com/blog/24415/a
Share
March 5, 2013 4:00:14 PM

bystander said:
I am actually someone who gets nausea unless I have high FPS. I have to play a long time before eye strain kicks in. Both can be symptoms of simulator sickness.

Have you tested to see if 30 FPS makes your eye strain worse than 60 FPS? For me, I get nausea in 1st person games in a few minutes at 30 FPS, but can last quite a lot longer at 60 FPS. That made it clear to me that higher FPS did help, but not everyone is the same.

As far as the glasses go, I found the 3D Vision 1 glasses to be uncomfortable, but the 3D Vision 2 glasses are much improved.

I feel a little uncomfortable recommending crossfire atm.
http://techreport.com/blog/24415/a


I just fired up a game I've had many hours on, witcher 2, its pretty demanding i played at 30fps and i believe my eyes just died. I then played on 60fps and it was like usual. It hurt only after a while. And about the 3D, I'll have to check them out in store if I do decide to go for a kit.
Score
0
a c 216 U Graphics card
a b K Overclocking
March 5, 2013 4:02:10 PM

Krazeee said:
I just fired up a game I've had many hours on, witcher 2, its pretty demanding i played at 30fps and i believe my eyes just died. I then played on 60fps and it was like usual. It hurt only after a while. And about the 3D, I'll have to check them out in store if I do decide to go for a kit.


Cool, it sounds like the 120hz monitor and higher FPS will help you, like it does for me.
Score
0
March 5, 2013 4:07:53 PM

bystander said:
Cool, it sounds like the 120hz monitor and higher FPS will help you, like it does for me.


I sure hope so! thanks for the information. And also I've been searching and I'll probably get the 120hz monitor before i get to crossfire and on other forums people have said that when you use a 120hz monitor and you get 60fps, its choppy and not as good as getting 60fps on a 60hz, i dont know how true this is but it worries me because some games don't work with crossfire. Is this true or is it just random nonsense?

And I'm planning on the BenQ XL2420T, what do you have?
Score
0
a b U Graphics card
a b K Overclocking
March 5, 2013 5:22:49 PM

Krazeee said:
I sure hope so! thanks for the information. And also I've been searching and I'll probably get the 120hz monitor before i get to crossfire and on other forums people have said that when you use a 120hz monitor and you get 60fps, its choppy and not as good as getting 60fps on a 60hz, i dont know how true this is but it worries me because some games don't work with crossfire. Is this true or is it just random nonsense?

And I'm planning on the BenQ XL2420T, what do you have?


http://www.pcper.com/reviews/Graphics-Cards/NVIDIA-GeFo...

AMD Radeon has been outed. Their sloppy drivers, as you can see, just push out junk in CrossFire. Do not crossfire your 7870's, it seems the slower the card, the worse this effect. If you are sensitive to rough/slow framerates, you need to get away from AMD and now. This is not an easy problem, Nvidia has been working on this for years and seem to have it down pretty well. If you want a fast and smooth frame rate, sell your Radeon and get a couple 660TI's or 670's if you can afford them. Remember, you cannot go by simple framerate, which Radeon rules, you need smooth framerate, which Geforce rules.

I have no doubt AMD Radeons will overcome this, but it may be a year or more away.
Score
0
a b U Graphics card
March 5, 2013 6:19:41 PM

babernet_1 said:
http://www.pcper.com/reviews/Graphics-Cards/NVIDIA-GeFo...

AMD Radeon has been outed. Their sloppy drivers, as you can see, just push out junk in CrossFire. Do not crossfire your 7870's, it seems the slower the card, the worse this effect. If you are sensitive to rough/slow framerates, you need to get away from AMD and now. This is not an easy problem, Nvidia has been working on this for years and seem to have it down pretty well. If you want a fast and smooth frame rate, sell your Radeon and get a couple 660TI's or 670's if you can afford them. Remember, you cannot go by simple framerate, which Radeon rules, you need smooth framerate, which Geforce rules.

I have no doubt AMD Radeons will overcome this, but it may be a year or more away.


Nvidia Geforce has been outed. Their users' sloppy reasoning, as you can see, just push out junk on forums. Do not SLI two Nvidia cards, it seems that faster the card, the worse the negative effect on the gamer's ego. If you are sensitive to rough/slow GPU zealots, you need to get away from Nvidia and now. This is not an easy problem, Nvidia's been working on this for years amd still seem to have it up pretty well. If you want a fast and smooth framerate, sell your house and get a couple of NASA supercomputers if you can afford them. Remember, you cannot go by simple price to performance ratio, which Radeon rules, you need smooth high prices, which Geforce rules.

I have no doubt Nvidia Geforcers will overcome this, but it may be a year or more away.
Score
0
a c 143 U Graphics card
a b K Overclocking
March 5, 2013 6:33:44 PM

babernet_1 said:
http://www.pcper.com/reviews/Graphics-Cards/NVIDIA-GeFo...

AMD Radeon has been outed. Their sloppy drivers, as you can see, just push out junk in CrossFire. Do not crossfire your 7870's, it seems the slower the card, the worse this effect. If you are sensitive to rough/slow framerates, you need to get away from AMD and now. This is not an easy problem, Nvidia has been working on this for years and seem to have it down pretty well. If you want a fast and smooth frame rate, sell your Radeon and get a couple 660TI's or 670's if you can afford them. Remember, you cannot go by simple framerate, which Radeon rules, you need smooth framerate, which Geforce rules.

I have no doubt AMD Radeons will overcome this, but it may be a year or more away.

Unless they unveil their testing methodology and everyone else test it and prove the inaccuracy of FRAPs, nothing can be taken into account. They said "CrossFire doesn't look faster than a single card" with the frame capture thing, they tested on BF3, how about throwing in Metro 2033 and comparing the results?
Score
0
a c 216 U Graphics card
a b K Overclocking
March 5, 2013 7:30:00 PM

ilysaml said:
Unless they unveil their testing methodology and everyone else test it and prove the inaccuracy of FRAPs, nothing can be taken into account. They said "CrossFire doesn't look faster than a single card" with the frame capture thing, they tested on BF3, how about throwing in Metro 2033 and comparing the results?


For years, you do see review sites mentioning that even though the FPS looked good, some games would not look good. That isn't a new revelation. They also did give out their methodology. They have a special card that takes takes video output, connected to another computer I believe, and save it to analyze. I would not dismiss it like it didn't exist. If you watched the video they gave to show you, it is pretty clear, and actually brought to my attention what microstutter actually looks like, and it is something I've noticed for years, I just never attributed it to the video card. I thought it was due to reading data about a new area. I still get it with my Nvidia card, but it is less often than with my previous AMD setup.

Anyways, I do not believe that 60 FPS on 60hz is better than 60hz on 120hz. At least not with v-sync on. However, your target FPS, if your goal is to reduce eye strain, is going to be 80+. At least that is when I no longer get nausea.
Score
0
a c 143 U Graphics card
a b K Overclocking
March 6, 2013 12:08:08 AM

I'll look deeply into that real time frame rendering thing, I'll need to see much more articles about that covering the topic, but what I couldn't understand, why only Crossfire? What nVidia have been doing not to fall into this? Is it related to V-Synch modes? Or is it about AFR? And how about higher refresh rates that will trap most of the FPS in shooting games? How does it look on 120Hz monitor?
Lots of questions going in mind, so I'll wait to gather all of these together to understand it quite well.

For the eye strain matter, lots of arguments going on this, it was clear for me to get eye strain from flickering especially on CRT monitors, that's why these monitors were great at 75-85 Hz. Also I get quickly tired from looking at a closed distance as the eyes get focused on the same distance without getting exercises. Going with LED wasn't good for me, i reduce the brightness and contrast about 25 degrees below or i just feel the sun is front of me. At 60 Hz, the game is very smooth and my eyes are very comfortable.

Aside from the OP, have you seen the TressFX demo? Any word about availability?
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 12:19:51 AM

ilysaml said:
I'll look deeply into that real time frame rendering thing, I'll need to see much more articles about that covering the topic, but what I couldn't understand, why only Crossfire? What nVidia have been doing not to fall into this? Is it related to V-Synch modes? Or is it about AFR? And how about higher refresh rates that will trap most of the FPS in shooting games? How does it look on 120Hz monitor?
Lots of questions going in mind, so I'll wait to gather all of these together to understand it quite well.

For the eye strain matter, lots of arguments going on this, it was clear for me to get eye strain from flickering especially on CRT monitors, that's why these monitors were great at 75-85 Hz. Also I get quickly tired from looking at a closed distance as the eyes get focused on the same distance without getting exercises. Going with LED wasn't good for me, i reduce the brightness and contrast about 25 degrees below or i just feel the sun is front of me. At 60 Hz, the game is very smooth and my eyes are very comfortable.

Aside from the OP, have you seen the TressFX demo? Any word about availability?


I'm pretty sure the reason behind the microstutter, in AMD's case, is due to AFR and not doing anything to space out frames.

Since both cards are rendering their own frames, if nothing is done to offset their displayed times, they could send two frames to the frame buffer at the same time, completely nullifying one frame, or nearly nullifying them in the examples they showed. To fix the problem, they need to have their software to attempt to space out the frames, but they also need to do this in a way that keeps the displayed image to match up to the action as well. I'm fairly sure that v-sync would also stop it, but no one ever benchmarks with v-sync.

The CRT problem is something almost everyone experiences, but what we have is related to simulator sickness. It doesn't have to be a CRT for us to have issues. Most people don't have this issue, but there are a number of people who do and don't know it as well. The Army did a study once, with their flight simulators, and found close to half their participants experience varying degrees of symptoms. It is clear that FPS helps me, and the OP, as he did test it at 30 FPS and 60 FPS and found 30 FPS made eye strain unbearable almost instantly. I'm pretty sure the issue is related to latency, as watching someone else play at 30 FPS has no adverse affects, it is only when I control the action with a mouse in 1st person view. I believe it is a similar issue to what causes motion sickness. The mind expect to see one thing due to your hand movements, but it is delayed, and the brain does not like that. This may not be solution for everyone, but it is for me, and it appears to also help the OP, though he has not gone above 60 FPS to know where the cut off point is for improvement.
Score
0
a c 143 U Graphics card
a b K Overclocking
March 6, 2013 12:38:37 AM

I'll wait and see if Anandtech or HardOCP covers this well as they give sometime pretty simple clarifications, HardOCP compares the CF performance from time to time, there was no word about the frame thing.
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 12:56:07 AM

A lot of people are now doing the frame latency tests, but you'll notice most always clarify that SLI and CF are likely inaccurate lately. The part of this new test that will take a while to catch on is that it requires special hardware to read in the frames. It really is the best method, at least if it works the way they intended (the video appeared to look legit to me), but I don't know how easy it will be for other sites to get the hardware for this testing.
Score
0
a c 143 U Graphics card
a b K Overclocking
March 6, 2013 1:54:37 AM

You still haven't told me about the new TressFX, is it in-game engine or driver related or what, I still haven't got the game.
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 1:59:15 AM

I have never seen it, but I will likely, once I get the new Tomb Raider. It is a new feature to make hair much more real. It appears to be a PhysX competitor by the look of it, or at least it may be going in that direction.

http://1985fm.com/news/2013/3/5/tomb-raider-pc-ultra-vs...
Score
0
a c 143 U Graphics card
a b K Overclocking
March 6, 2013 2:13:40 AM

Yes it is, but Physx is more intended for lots of effects while TressFX doesn't, the AMD blog stated some stuff about being it well-developed on GCN based cards, I wanna see a comparison on both cards as guys in the geforce forums reported some glitches when running it on their cards in SLI, another word about it is that it's very taxing, a youtube guy said it takes 20-25 FPS down. Anyways it looks fantastic on the TR trailer.

Back to the frame rendering, when I watched the video I only noticed the screen tearing which is a well-known phenomenon and its' cause.
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 2:26:34 AM

ilysaml said:
Yes it is, but Physx is more intended for lots of effects while TressFX doesn't, the AMD blog stated some stuff about being it well-developed on GCN based cards, I wanna see a comparison on both cards as guys in the geforce forums reported some glitches when running it on their cards in SLI, another word about it is that it's very taxing, a youtube guy said it takes 20-25 FPS down. Anyways it looks fantastic on the TR trailer.

Back to the frame rendering, when I watched the video I only noticed the screen tearing which is a well-known phenomenon and its' cause.

I didn't really notice it much in the Dirt 3 video, but it is quite apparent in the Dishonored video. It sort of skips, as well as tears. Then they slow it down and show you why.

As far as the TressFX issue with Nvidia cards, I found on another forum, that some guys were having a good experience if they turned off Tessellation. I don't know if it is an issue when both are used together, or if Tessellation was the culprit on its own. But guys with a single 680 were getting 40+ FPS with it on.
Score
0
a c 143 U Graphics card
a b K Overclocking
March 6, 2013 2:55:26 AM

I'm waiting for the next generation of CPUs whether AMD or Intel and then I'll upgrade my HD 6950 it doesn't cut it anymore in maxed settings, in HA, I cranked down AA to 2X with everything else on Ultra and the game looked really so brilliant but it doesn't just happen in most of the titles.
Score
0
March 6, 2013 11:19:32 AM

I'm not convinced on this whole issue. Thousands of popular and well knowing people have crossfire and they test it. Why wouldn't it be more known?
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 1:46:53 PM

Krazeee said:
I'm not convinced on this whole issue. Thousands of popular and well knowing people have crossfire and they test it. Why wouldn't it be more known?

It is quite known, but reviewers have to be objective. Most don't tell you what looks better unless it is bad. Otherwise they have to give raw data and show you what it says. Until now, no one has given raw objective data that support what many people have noticed, because everyone was using FPS as the only metric for measuring.

Here are a couple videos with side by side comparisons. They aren't broken down to be analyzed, so you'd never have this shown in a benchmark. The 2nd one hides what video card is what. The differences between the 2nd video is most apparent during the turns.

http://www.youtube.com/watch?v=NLcq6IQz-sM&feature=play...

http://www.youtube.com/watch?v=0Md1Xpwq1X0&feature=play...
(They tell you which card is which as well as FPS if you watch the link given at the end of the 2nd video).

EDIT: If what you see doesn't bother you, then crossfire might not bother you as much. That said, those videos were with single cards.
Score
0
March 6, 2013 5:24:05 PM

bystander said:
It is quite known, but reviewers have to be objective. Most don't tell you what looks better unless it is bad. Otherwise they have to give raw data and show you what it says. Until now, no one has given raw objective data that support what many people have noticed, because everyone was using FPS as the only metric for measuring.

Here are a couple videos with side by side comparisons. They aren't broken down to be analyzed, so you'd never have this shown in a benchmark. The 2nd one hides what video card is what. The differences between the 2nd video is most apparent during the turns.

http://www.youtube.com/watch?v=NLcq6IQz-sM&feature=play...

http://www.youtube.com/watch?v=0Md1Xpwq1X0&feature=play...
(They tell you which card is which as well as FPS if you watch the link given at the end of the 2nd video).

EDIT: If what you see doesn't bother you, then crossfire might not bother you as much. That said, those videos were with single cards.

not sure if you can tell micro stuttering from a youtube video...

But what i understand is that both AMD and Nvidia have this problem and it's more noticable on SLI/CROSSFIRE... But, people on other posts said that it only happens in some games. And limiting the fps to 60 also helps get rid of it. Even if i have a 120hz monitor, getting a 60fps would be better than 60hz 60fps and its only for certain games?
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 5:46:00 PM

Krazeee said:
not sure if you can tell micro stuttering from a youtube video...

But what i understand is that both AMD and Nvidia have this problem and it's more noticable on SLI/CROSSFIRE... But, people on other posts said that it only happens in some games. And limiting the fps to 60 also helps get rid of it. Even if i have a 120hz monitor, getting a 60fps would be better than 60hz 60fps and its only for certain games?


I don't think it would be better, but it shouldn't be worse. The reason people say 60FPS looks choppy after they get a 120hz monitor is because after you get used to higher FPS, lower FPS is more noticeable.
Score
0
a c 271 U Graphics card
a b K Overclocking
March 6, 2013 5:48:06 PM

Krazeee said:
not sure if you can tell micro stuttering from a youtube video...

But what i understand is that both AMD and Nvidia have this problem and it's more noticable on SLI/CROSSFIRE... But, people on other posts said that it only happens in some games. And limiting the fps to 60 also helps get rid of it. Even if i have a 120hz monitor, getting a 60fps would be better than 60hz 60fps and its only for certain games?

I've seen people claiming microstutter on all sorts of configurations, single and dual cards.
Score
0
March 6, 2013 7:59:01 PM

Mousemonkey said:
I've seen people claiming microstutter on all sorts of configurations, single and dual cards.


I guess I shouldn't worry about it then? Because If i did worried then i would never even have a graphics card in the first place ^-*
Score
0
a c 216 U Graphics card
a b K Overclocking
March 6, 2013 8:09:20 PM

Krazeee said:
I guess I shouldn't worry about it then? Because If i did worried then i would never even have a graphics card in the first place ^-*


Since you are selling your current card, looking at buying an Nvidia 3D Vision 2 monitor, and all data points to fewer problems with SLI, I would go with Nvidia cards if I were you.
Score
0
March 7, 2013 9:02:17 AM

bystander said:
Since you are selling your current card, looking at buying an Nvidia 3D Vision 2 monitor, and all data points to fewer problems with SLI, I would go with Nvidia cards if I were you.


... I never said i'm selling my current card, however, the BenQ is a 3D vision monitor which i don't plan on using so yeah. And Nvidia is the same damn thing as AMD, nobody cares except the fanboys who want to promote a corporation just because they feel apart of something.

Look at it this way

Radeon 7870 1x the performance --- + Can use my current PSU, can crossfire later

Radeon 7970 1.25-1.5x the performance ---... If i do this i need to sell my current one (money loss) and i wont have the ability to crossfire it in the future since my psu is only 750w... this is more hassle for me since i need to sell my current card (which i cannot do where I live without a ton of issues, more issues than cfx a 7870...)

2XRadeon 7870 = 1-1.8x the performance --- I can get almost double the frames on quite a few games, on some i'll have issues so i could just turn off CFX for that specific game and in the end im a happy camper... + this can take advantage of the 120hz monitor...

Am I looking at it the right way?
Score
0
March 7, 2013 11:24:11 AM

BigMack70 said:
I just gotta throw one monkey wrench in here...

My experience with stutter, and the experience of others, is that at least with Crossfire it is completely removable with a simple framerate limiter. Not a perfect solution, because if your GPU setup can't output a framerate equal to or higher than the refreshrate of your monitor you will need to limit your framerate below that, but it more or less "solves" crossfire stuttering.

You need to be aware that with crossfire, sooner or later you are going to need to troubleshoot things. But in my experience with 7970 CF (which I assume will translate to 78xx CF), there are no problems with it that extend beyond slight annoyance if you are willing to troubleshoot.

I do not personally have experience with SLI but from forums, reviews, etc - it seems like the same problems exist there, just with either less frequency or intensity.


Thanks for the info, I do appreciate it. I guess I'll be going for a 7870 Crossfire. People say that non powerful cards dont crossfire well but the 7870 is also very powerful. It hopefully wont run into much issues...

Time to pick my monitor now, Thanks everyone... I don't know who to make as a best answer...
Score
0
March 7, 2013 11:26:00 AM

Best answer selected by krazeee.
Score
0
!