Would upgrading to a GPU with more ram stop stuttering?

[Wile E]

And absolutely nothing that you have said in any way translates to low cache cpus causing stutter. Sorry cadaveca, you are quite simply wrong in this argument, on ALL counts. It has been proven over and over again, even right here in these forums. Low cache cpus do not cause stutter, and they have little impact on real-world gaming performance. They hurt benchmarks, and that's about it.

 

[cadaveca]

LoL. You're taking it the wrong way then, because that's totally NOT what i am saying...merely that it has an impact on performance. The OP had enough mhz...enough vga...I've had an e7400, and in comparison to my E8400, feels quite slow.

Having gone through the Core2 gen since Conroe, I've seen how the added cache of the 45nm chips helps, too. Nevermind the perforamnce boost in memory they brought.

In the end, I'm not talking about anything other than ideals. A low cache cpu is NOT as ideal as one with larger cache, and this is becoming even more important as games get multi-threaded...they need the data for those threads in a very timely manner.

 

[Wile E]

Yeah, but the OP was about stutter, and you automatically jumped on CPU. Single core Athlons can game perfectly smoothly. Sure, they suffer in performance somewhat, but they are smooth. Nobody said cache makes no difference, we all just said it makes LITTLE difference, especially in gaming.

As for the guy with audio problems, check drivers. I'd bet money that it's a driver or software issue. Perhaps AV softare is interfering? And don't some games have issues with Xfire and other chat type programs as well? Any dual core is capable of processing the sound of a video game without stuttering, so the issue has to be software.

 

[phanbuey]

LoL. You're taking it the wrong way then, because that's totally NOT what i am saying...merely that it has an impact on performance. The OP had enough mhz...enough vga...I've had an e7400, and in comparison to my E8400, feels quite slow.

Having gone through the Core2 gen since Conroe, I've seen how the added cache of the 45nm chips helps, too. Nevermind the perforamnce boost in memory they brought.

In the end, I'm not talking about anything other than ideals. A low cache cpu is NOT as ideal as one with larger cache, and this is becoming even more important as games get multi-threaded...they need the data for those threads in a very timely manner.

but he's talking about noticeable stutter... not a low framerate as a cpu with low cahce would have...

I've played with CS:S and Single-core AMD semprons for a while.. and it wasnt until my first Athlon that I noticed it SMOOTHER than my sempron at the same clock. But the sempron didnt stutter... it lagged... it felt sluggish, but it didnt stutter.

I know the stutter he is talking about, and it was the same with my 750i chipset... i would have it too - 80 FPS and then a jerk - like the virus scanner just kicked in or something, only oddly, it would happen wayyy too often. The system was fast, had way more than enought GPU Vmem for 1680x1050 (2x gtx260's), but it would stutter.

As soon as I changed CPU's to the quadcore q9650... the system stuttered less, was less jerky, but still did it noticeably still. The problem was still there, but less pronounced. Once i got the i5 750, the stutter was completely gone.

I'm guessing that the inefficient 750/780 nv chipset had something to do with it. My farcry2 and Crysis warhead benches also improved by close to 40% with the same gfx combination. Cache is definitely great for 3d games, but im not sure that it has to do with the jerkiness of the system.

 

[cadaveca]

I'm guessing that the inefficient 750/780 nv chipset had something to do with it. My farcry2 and Crysis warhead benches also improved by close to 40% with the same gfx combination. Cache is definitely great for 3d games, but im not sure that it has to do with the jerkiness of the system.

nV didn't exactly have the best 775 chipsets. In fact, as a final chipset, it was embarassing..

But you know, bringing the memory controller out of the chipset, and right next to the cache of the cpu has enormous benefits when done right.

And as to why, really it's so important to performacne, and not just gaming, and would be why they had shared cache on Core2 chips. Keeping it this way in i7 was very important, as data going to the 4th core has much farther to travel than the 1st, as the QPI link is next to one core's cache only. With a shared cache, any core can read something pretty much as soon as it hits the L3, without waiting for a crossbar managing data to the cores from the chipset link(as in 775 quads)in the way.

I'm convinced that a shared cache and triple channel mem would have Phenom performing as well, if not better, than i7. But in the end, AMD's Phenom target market isn't the same as i7.

AMD semprons are kinda moot to this point, as memory is so close(since 754 AMD has had IMC, this is what made them so good far gaming back then, remember?), but with 775, and it being so far away, cache becomes oh so much more important, and why, currently, i7 is so much better than AMD for gaming...it's all about cache and memory control.

 

[rahulyo]

Is mobo cause stuttering/low performance problem ???

 

[LoneReaction]

I know the stutter he is talking about, and it was the same with my 750i chipset... i would have it too - 80 FPS and then a jerk - like the virus scanner just kicked in or something, only oddly, it would happen wayyy too often. The system was fast, had way more than enought GPU Vmem for 1680x1050 (2x gtx260's), but it would stutter.

I feel kinda bad that I caused members to have tension in this thread =X

Thinking about it more, the stutter I experienced back in my E7200 setup, with any game, is what the quote above describes.

However, with CPU intensive (was using 100% on E7200) games like dragon age, it's more like the game skips a frame every second, or 2 seconds. Consistently. I think it would be impossible to capture this on fraps (because running it changes things) or on a digicam, unless it's a high speed one.

It could be the software that cause this, but I just remembered that I've been with XP, Vista, and Win 7 on the E7200, and all of them behaved identically.

 

[Mussels]

LoL. You're taking it the wrong way then, because that's totally NOT what i am saying...merely that it has an impact on performance. The OP had enough mhz...enough vga...I've had an e7400, and in comparison to my E8400, feels quite slow.

Having gone through the Core2 gen since Conroe, I've seen how the added cache of the 45nm chips helps, too. Nevermind the perforamnce boost in memory they brought.

In the end, I'm not talking about anything other than ideals. A low cache cpu is NOT as ideal as one with larger cache, and this is becoming even more important as games get multi-threaded...they need the data for those threads in a very timely manner.

and i've been on core 2 since allendale. if those CPU's caused stuttering, everyone would have been far worse off on the older CPU's they upgraded from...

you've made an assumption, and we all know its wrong. simple as that. cache is NOT the cause. if you have shit onboard sound that depends on the CPU, maybe its upgrading to the faster CPU that helps? maybe a soundcard would do the same thing?

OCing, faster per clock (cache), better sound card... ya know, theres more than one solution here.

 

[cadaveca]

Then why does he no longer have any issues? And ya gotta drop the whoel soudncard thing..wasn't me suggesting changing a soundcard.

The answer "software" doesn't do it for me.

And like I said in the quote, he's pushing his system to the limit. We are talking about a specific situation, not "in general". Could have just been the overclock itself causing the problem, and I'm more inclined to go with that than the "unknown software" cause.

It would be intersting to see if he's installed all the same software, and when he does, if it causes stutter...because if it WASN'T the cpu, than he'd be subject to the same problem.

 

[Mussels]

uhhh... we're saying a soundcard might have fixed it. Not dropping it, cause its probably a working solution as opposed to just the CPU one.

If you hadnt realized we were mentioning that as an alternative to a CPU with more cache, you arent reading our posts...

 

[cadaveca]

No, I'm reading, and when 2mb of cache equates 200mhz of performance, you're simply arguing to be argumentative, it seems!

The difference is, you're looking just at performance via mhz, and in the end, agreeing with me, in a way. Performance does not have to be had by mhz alone...cache is just another option, similar to a sound card. But as others have mentioned, in Vista and Win7, all sound uses cpu time, so replacing a soundcard is kinda moot, unless you got a board with a bad codec, and today codecs are pretty damn good.

It could be the software that cause this, but I just remembered that I've been with XP, Vista, and Win 7 on the E7200, and all of them behaved identically.

And there ya go. 3 OSes, same issue, gone with cpu upgrade.


In the end, it's not mhz he was lacking, and this is my point. So we gotta look elsewhere, and I said cache, plain and simple. Buying another cpu with 2MB would have had the same issues. Personally, I think a 45nm Core2 would have done just fine, but maybe not.

 

[DaedalusHelios]

I'm convinced that a shared cache and triple channel mem would have Phenom performing as well, if not better, than i7. But in the end, AMD's Phenom target market isn't the same as i7.

That makes no sense whatsoever. i7's perform almost the exact same when they are run in dual channel. I ran one of my i7's with dual channel memory for a while until I found a good tri-channel kit for it. It had the same PPD then as it does now, when at the same clocks. Phenom II would have to have better performance numbers to be able to take advantage of triple channel memory and have it make a difference. AMD did the right thing keeping it dual channel as it simply isn't fast enough to take advantage of triple channel. They have to walk before they can run. i7 is even far from taking full advantage from triple channel.

I am just stating it from experience and benchmarks.

AMD is a generation behind Intel. Nvidia is a generation behind ATi. I have faith in all these four companies because it is all we have got to choose from in such a limited market.

 

[Mussels]

triple channel would only really help if you had all 8 threads hammering away at something like video encoding - its intels future proofing, not something that really benefits us right now.

 

[cadaveca]

That makes no sense whatsoever.

I was refering to bandwidth only.

 

[Wile E]

No, I'm reading, and when 2mb of cache equates 200mhz of performance, you're simply arguing to be argumentative, it seems!

The difference is, you're looking just at performance via mhz, and in the end, agreeing with me, in a way. Performance does not have to be had by mhz alone...cache is just another option, similar to a sound card. But as others have mentioned, in Vista and Win7, all sound uses cpu time, so replacing a soundcard is kinda moot, unless you got a board with a bad codec, and today codecs are pretty damn good.



And there ya go. 3 OSes, same issue, gone with cpu upgrade.


In the end, it's not mhz he was lacking, and this is my point. So we gotta look elsewhere, and I said cache, plain and simple. Buying another cpu with 2MB would have had the same issues. Personally, I think a 45nm Core2 would have done just fine, but maybe not.

No, a Bad OC can cause issues. Bad chipset drivers can cause issues that will show up across all OSes. A bad BIOS revision can cause it. Bad graphics driver, bad sound driver, etc., etc. Still does not prove anything.

There are mountains of evidence that proves low cache cpus can game perfectly well, without stutter or poor performance, a ton of which is right on this site. Hell, DaMulta even showed us an older Athlon Single core with single channel ram was perfectly capable of gaming on an 8800GT, albiet at a reduced framerate, but smooth as butter.

Sorry, you are still flat out wrong, and barking up the wrong tree. It's simply not the cpu cache. We would be hearing about a hell of a lot more issues if low-cache cpus caused stuttering.

 

[DaedalusHelios]

No, a Bad OC can cause issues. Bad chipset drivers can cause issues that will show up across all OSes. A bad BIOS revision can cause it. Bad graphics driver, bad sound driver, etc., etc. Still does not prove anything.

There are mountains of evidence that proves low cache cpus can game perfectly well, without stutter or poor performance, a ton of which is right on this site. Hell, DaMulta even showed us an older Athlon Single core with single channel ram was perfectly capable of gaming on an 8800GT, albiet at a reduced framerate, but smooth as butter.

Sorry, you are still flat out wrong, and barking up the wrong tree. It's simply not the cpu cache. We would be hearing about a hell of a lot more issues if low-cache cpus caused stuttering.

I agree. I just see cache size comparisons are only valid on the same architecture too. Cache makes a CPU faster per clock when compared with another CPU of the same architecture. But it doesn't mean it necessarily could be bad enough to cause stuttering. That would have to be one incredibly slow CPU IMO. So I suppose it is possible in an extreme situation. Like a gtx 295 being driven by an early sempron somehow. But not in the OP's situation. I remember a micro stutter in early 9800gx2 drivers making me get a little nauseated but that was the launch driver's fault in a multi-GPU configuration.

 

[Mussels]

one cause for stutter, the original CPU had bad thermal paste and overheated/throttled a lot - new CPU may appear to have fixed the issue, but it could have been as simple as that.

its a possibility, but its far more likely than cache being a magic bullet.

 

[hellrazor]

I agree. I just see cache size comparisons are only valid on the same architecture too. Cache makes a CPU faster per clock when compared with another CPU of the same architecture. But it doesn't mean it necessarily could be bad enough to cause stuttering. That would have to be one incredibly slow CPU IMO. So I suppose it is possible in an extreme situation. Like a gtx 295 being driven by an early sempron somehow. But not in the OP's situation. I remember a micro stutter in early 9800gx2 drivers making me get a little nauseated but that was the launch driver's fault in a multi-GPU configuration.

Wait, he did get all new drivers (and game patches) before buying a new CPU, right? I'm not reading through 7 pages of people complaining about Rapelay just to figure it out (Oh, I mentioned it, please don't take my posting privs). It could have been a bad driver taking way too much in a certain situation (textures being loaded, a decent change in poly count, a special effect, etc.) and the CPU could have just added bandwidth between the GPU and itself, causing a smaller percentage of it to be taken up during the certain situation, leaving more to be used by normal stuff (animation, effects, lighting, etc.).

I'm not taking sides here, just a thought.

 

[TheMailMan78]

No, a Bad OC can cause issues. Bad chipset drivers can cause issues that will show up across all OSes. A bad BIOS revision can cause it. Bad graphics driver, bad sound driver, etc., etc. Still does not prove anything.

There are mountains of evidence that proves low cache cpus can game perfectly well, without stutter or poor performance, a ton of which is right on this site. Hell, DaMulta even showed us an older Athlon Single core with single channel ram was perfectly capable of gaming on an 8800GT, albiet at a reduced framerate, but smooth as butter.

Sorry, you are still flat out wrong, and barking up the wrong tree. It's simply not the cpu cache. We would be hearing about a hell of a lot more issues if low-cache cpus caused stuttering.

I used to game just fine on a 4200x2. Hell I played all the way through Fallout 3 with that thing.