Are AMD Current CPUs Not Enough For Solid 60 FPS Anymore?

[cadaveca]

LOL...it's not going to be any time soon, FYI. I'm currently testing off-line games, and coming up with a lsit of apps that are affected.

But I also have to look at alot of maps within each game...so this makes it alot more time consuming than jsut trying to get some FPS numbers.

Robert(the OP) highlighted this too, in GhostBusters...his vgas ahve enough grunt to push that game maxed out, but certai nscenarios lead to really poor performance, just simply due to the number of vgas he has(of course, the investigation revolves around finding out exactly when these limitations are exposed, and it may just tunr out to be bad game programming..and not the number of vgas...that's the question that needs to be answered.).

 

[CDdude55]

I don't think it's really "solved", and I think you should have been able to tell that my posts over the past 2 months have been leading up to exactly this..

Sorry for not keeping track of your posts throughout the past two months.

And thanks for saving the tech world as we know it with your hard hitting research and grab-the-bull-by-the-balls attitude. .

 

[cadaveca]

bleh..I don't think, really, that AMD has missed this. Nor do I think I'm doing anything special...I just think I have noticed something that many have not, and given the responses around these issues, I might be right there.

honestly though, my mentioning all these problems I've had, and how long they have been around for without suitable fixes, to me, more highlights what we should expect from the future gen. It's only because the 6-series is coming that I'm trying to make this stuff widely known...because we can use these things to judge how good the 6-series is...without including the competition in the mix.

I do not consider AMD and Nv to be competitors at this point...thier primary focus is too different.

I've mentioned before I expect "final hurrah" driver for the 5-series before the 6-series launches. So maybe alot of the outstanding issues will be fixed...but maybe, as I think, alot of these issues are pure-hardware-based, and cannot be solved by driver...time will tell.

And if they ARE hardware-based, they better damn well get fixed with these new chips.

Eyefinity launched...but you need 2x5870 for decent performance. And 2x5870 needs an overclocked cpu...that's a problem to me..to me, Eyefinity has launched.. but doesn't work.I can only consider it working once all of these issues have been addressed, and once it doesn't require you void your warranty to use it. Until that happens, and AMd sells and promotes tech that requires you overclock, then I have to agree with the OP and this thread title...that no AMD cpu is capable of pushing 60FPS.

 

[EastCoasthandle]

I'm not sure what direction this thread is in since the OP. However, if you are using CF or SLI with an AMD processor and you believe that you are not experiencing the level of smoothness and frame rate that others are getting there is one possible solution. Other then making sure you use RamMap to empty "System Working Set" and "Modified Page List" one can also use RadeonPro to increase Flip Queue Size from it's default of 3 to 5. Depending on the game, etc that may yield some improvement.

 

[Loosenut]

@ Cadaveca The more I read, the more I'm in awe of the burden you are voluntarily shouldering.

Lol, you should apply for a gov't. grant...

 

[cadaveca]

~snip~

Thanks again for that info..I'll check it out.


What troubles me about these fixes is that while I myself might be aware of them, I do not think most others are.

That said, needing to use outside software to get basic performance either says AMD, or Windows, sucks big time. And because I cannot personally prove who of the two is at fault here, I'm left ignoring fixes like this, as they are far to advanced for a normal user. Things like this should happen automatically, and be transparent to the end user. it's not us that needs this info...it's AMD.

@ Cadaveca The more I read, the more I'm in awe of the burden you are voluntarily shouldering.

Lol, you should apply for a gov't. grant...

I don't think this is such a big deal as you make it out to be. But thanks anyway...and really, I don't need money to do this...nor do I expect to get anything out of it.

This is just me geeking out on the stuff I love.

 

[SKD007]

People compare what's priced equally; why I often compare an i5 750 to 1055T. That's what's fair any way you cut it.

After everyone sorts out standard settings I'd recommend retesting with slight setting changes. See if a performance hit exists for one person but not another then compare the setups.

ya u r right if the price is the same.. but does the x4 phenom II cost the same as i7 920 ??.. i didnt know that... hmmm

 

[SKD007]

Protip: Open up notepad and write your midway-through-thread reply so you don't forget the points in your mind currently, then finish reading before actually replying. Not saying it will apply here, just less likely to say something that doesn't fit doing it that way.

The way u talk.. :shadedshu u make people feel bad

 

[erocker]

If people want to come up with proof or results for this topic, fine. Make a thread and post them. Otherwise this thread is going nowhere fast. Stay on topic.

 

[LAN_deRf_HA]

The way u talk.. :shadedshu u make people feel bad

I thought it was nice practical advice, and the "not saying it will apply here" was meant to emphasize it was more general advice, less attack on the poster. ON topic: I don't think we have it together enough as a community to organize this properly.

 

[Robert-The-Rambler]

Thanks Erocker and a Clarification

If people want to come up with proof or results for this topic, fine. Make a thread and post them. Otherwise this thread is going nowhere fast. Stay on topic.

I just want to restate that what I am really looking for is a list of games where say at 1080p it is impossible for AMD processors on any platform, even the fastest current ones at stock everything, even when your single GPU or multiple GPU setup is capable of that performance, to simply maintain solid/locked 60 frames per second and maybe even at times performance is terrible. I want people to get an accurate idea of whether buying certain AMD products are worth it and if you want to get certain games what to actually expect. Why?

Because so often benches are done that don't give a true picture of what is going to happen when you actually game. Resident Evil 5 is one where the bench is way more stressful than the game especially the fixed one. What if there was some other way of testing real world performance without identical short term demo runs? Maybe a reviewer should actually play entire levels and track that performance on a graph and report variances in frame rates with Vsync on. It doesn't have to be identical runs to get clarity if the test is long enough. Who cares about max frames if you are getting valleys on different platforms that alter the gameplay experience?

So often reviews of new gear are done on overclocked I7s to remove bottlenecks. That is fine and all but what about others with lesser gear. Perhaps reviews should be done on the best that AMD has stock vs the best that Intel has at stock and then with the methodology I mention track the performance giving a very clear picture of what to expect when you actually play the game. We do buy this stuff to actually play games. Surely we have the time to actually play the games and report general performance in charts utilizing programs such as FRAPS to tell us that maybe Ghostbusters or whatever it may be might really suck on this type of system whatever it may be. Or on an I7 we saw almost a flatline through the entire Times Square Level making it a much better choice for this title for example.

I don't care about benchmarks so much anymore. I want raw data that actually illustrates real world gaming conditions and gives the real story of how your system is going to play any particular game. I don't want to necessary turn this into an epic ramble but I've been pondering this for a while. Something keeps gnawing at my mind about standard deviations in data and what it all means and how it relates to the variance from the mean/average. If we lock the FPS at 60 and track performance while we play a game for an extended period the best systems are going to have the lowest standard deviation or variance from the mean. We can track that by simply saving the results using a program such as FRAPS, can't we? That is what I want to see. Real games played by real people with real data that tells you in a general sense exactly what to expect when you plan on purchasing a game, not just a representation of the game, but the actual game.

Thinking about breasts is a whole lot simpler.

 

[cadaveca]

I think the issue at hand is that performance can change so much every month with each driver release, it's really hard to keep it up to date, and would really require many people working all month long, IMHO.

It would be nice if they could develop a nice little extensionof STEAM, say, that actually kept track of exactly what you are looking for...and I think Valve themselves actually do this with Source-based games...I know they do collect quite a large amount of data from players actually playing the game, down to things like where people died, ammo used, etc...and all other things that they use for load-balancing.

 

[Mussels]

its got little to do with drivers, video cards, or CPU's... modern games just often have shitty coding, and slow down at various points of the game no matter your hardware.

 

[TheMailMan78]

its got little to do with drivers, video cards, or CPU's... modern games just often have shitty coding, and slow down at various points of the game no matter your hardware.

That pretty much summed it up Robert. I hate to say Mussels is right but he is. Games nowadays are coded for consoles first and PC second and depending on the developer and who his "sponsors" are some hardware will ALWAYS perform better then others. But with a simple driver update that could change. This is the nature of PC gaming.

In the end raw power will never be enough. Crossfire rigs have ALWAYS had problems scaling. This is nothing new. Its all in the coding. With that being said I personally run a single ATI GPU with an AMD CPU get 60 FPS in most of my games as a minimum and I have seen i7 rigs get a lot less due to user error. Every system is different. Every user is different. Just learn to use what you have the BEST you can.

 

[LifeOnMars]

I'm looking to build a gaming rig in the next month or so, 1680x1050 res monitor. I was set on getting a nice hexacore AMD setup as I have a fairly decent budget, this thread however has me worried as I like to run my games smooth as silk. Is it wise to stick with an intel i7 or what is new on the horizon from both companies, any pointers would be great?

 

[TheMailMan78]

I'm looking to build a gaming rig in the next month or so, 1680x1050 res monitor. I was set on getting a nice hexacore AMD setup as I have a fairly decent budget, this thread however has me worried as I like to run my games smooth as silk. Is it wise to stick with an intel i7 or what is new on the horizon from both companies, any pointers would be great?

Dude this thread in short is just smoke and mirrors. With that being said I personally would wait a month or two if you can. New architecture is right around the corner. I mean within months. Wait and see.

 

[CDdude55]

I'm looking to build a gaming rig in the next month or so, 1680x1050 res monitor. I was set on getting a nice hexacore AMD setup as I have a fairly decent budget, this thread however has me worried as I like to run my games smooth as silk.

This thread is the last thing you should be reading in determining what rig to go for.

And i agree with MailMan, we should be seeing a new architecture coming out pretty soon, so i think it would be best to wait for that.

 

[Steevo]

My 4850 and quad was rocking GTA4 just fine, not at 60 all the time with the settings I had, but it was still running awesome compared to others.

Don't worry about running a AMD system, if it were that much of a bottleneck no one of us would run them, but we do. I love mine still, I have enough power to run a TV with hardware accelerated Netflix while I game on my monitor on most games.

 

[Dent1]

I'm looking to build a gaming rig in the next month or so, 1680x1050 res monitor. I was set on getting a nice hexacore AMD setup as I have a fairly decent budget, this thread however has me worried as I like to run my games smooth as silk. Is it wise to stick with an intel i7 or what is new on the horizon from both companies, any pointers would be great?

The performance issues being discussed in this thread are specific to Ghostbusters an Dragon Age Origins only, it has no bearing on any other game and from what I’ve read these two isolated games are to do with a bug or bad coding? Which is causing poor scaling on crossfire with the AMD X4s. TheMailMan has an X6 and a non-crossfire setup and he is getting solid FPS, unlike the OP.

Look at this logically, search Google and you'd see plenty of forums with people complaining that their i7 or SLI set up is getting poor frame rates in a specific game, but it wouldn’t stop you from buying a i7 would it? Poor performance in a isolated game has no reflection on the i7 or Phenom IIs performance.

All this “wait for AMD’s next architecture” stuff is nonsense, do not avoid AMD, avoid Ghostbusters and Dragon Age Origins.

Anyways, this isn’t something that concerns you as 1680x1050 is considered to be low resolution, the OP has a 30" screen @ 2560x1600

 

[erocker]

The performance issues being discussed in this thread are specific to Ghostbusters an Dragon Age Origins only,

I can tell you right now, with my CPU at stock, I get a constant 60fps in Ghostbusters. In Dragon Age I am also consistently over 60fps with the exception of right before the end battle where there are 100's of soldiers on screen and I dip into the 50's. It's not even a part where you have to battle as you're just walking towards a gate. I'd be interested to see an i5 750 during this scene.

 

[Dent1]

I can tell you right now, with my CPU at stock, I get a constant 60fps in Ghostbusters. In Dragon Age I am also consistently over 60fps with the exception of right before the end battle where there are 100's of soldiers on screen and I dip into the 50's

For the benefit of others can you confirm that you run a AMD processor?

Well, it's in my signature...

Ok.

 

[erocker]

Well, it's in my signature...

 

[cadaveca]

For the benefit of others can you confirm that you run a AMD processor?

He has no issues, becuase he's got an overclocked cpu. Which WASN'T Robert's question...he was asking for cpus running at stock, to which everyone here, except me, does not apply.

I just want to restate that what I am really looking for is a list of games where say at 1080p it is impossible for AMD processors on any platform, even the fastest current ones at stock everything, even when your single GPU or multiple GPU setup is capable of that performance, to simply maintain solid/locked 60 frames per second and maybe even at times performance is terrible.

It's a specific question, with a specific answer.


No, AMD cpus, at stock, are NOT good enough for 60FPS, when running more than one cypress card. Yes, they are fine, if you run a single Cypress-based card, or anything that performs less than that.

The question doesn't ask...HOW you get better performance, and it definately DOESN'T include overclocked cpus. I wish everyone would really STOP saying that they are fine, when overclocked...NO SHIT. That WAS NOT Robert's question. And it definately DOESN'T include JUST Dragon Age, and Ghostbusters.

Use one video card, and yes, things are generally fine...with a stock cpu, 3.0ghz and higher. Like Mussels metioned...if it isn't, then the issue isn't really anything other than bad programming.

But, my point in all of this, is that no matter the programming, 2 or more Cypress based cards NEED more than a stock AMD cpu. As such, if you are using multiple Cypress-bsed cards, you are better off with either an overclocked AMD cpu, or an overclocked i7 cpu(and i7 has the edge). And through my testing, this is due to a driver issue, to which both AMD and Intel cpus run into a bottleneck, and why that bottleneck occurs, isn't exactly obvious. Hence me doing the testing I am right now....I am just interested to find out, whether it's the NB that needs overclocking...the cpu speed...maybe ram...I don't know.

 

[erocker]

He has no issues, becuase he's got an overclocked cpu. Which WASN'T Robert's question...he was asking for cpus running at stock, to which everyone here, except me, does not apply.



It's a specific question, with a specific answer.


No, AMD cpus, at stock, are NOT good enough for 60FPS, when running more than one cypress card. Yes, they are fine, if you run a single Cypress-based card, or anything that performs less than that.

The question doesn't ask...HOW you get better performance, and it definately DOESN'T include overclocked cpus. I wish everyone would really STOP saying that they are fine, when overclocked...NO SHIT. That WAS NOT Robert's question. And it definately DOESN'T include JUST Dragon Age, and Ghostbusters.

Sucks I'm just on one card now but, I think you can leave you CPU stock and just up the CPU/NB frequency which I think is where the real bottleneck is. When Dirt 2 came out I know positively that I was running my CPU stock with CrossFire and I would get occasional slowdowns. Adjusting the CPU/NB from 2000 to 2400 fixed it.

 

[Dent1]

He has no issues, becuase he's got an overclocked cpu. Which WASN'T Robert's question...he was asking for cpus running at stock, to which everyone here, except me, does not apply.

Didn’t TheMailMan claim solid FPS running at stock?

The premise of the post was to confort LifeOnMars whom is looking at the AMD X6. I was relating his situation to TheMailMan whom is already owns AMD X6 and has no issues.

And it definitely DOESN'T include JUST Dragon Age, and Ghostbusters.

Maybe not but Dragon Age and Ghostbusters are the two games in question in this thread. To add some balance maybe we should list all the games which Intel CPUs perform arguably inadequate in?