Just how bad is my E8600 holding back my GTX570SC?

B

Burke888

Gawd
Joined
Aug 1, 2004
Messages
613
Hey guys I was wondering what type of improvements I could see if I upgraded my 3.33Ghz E8600 coupled with 8GB of 400mhz RAM? I just purchased an eVGA GTX 570 Super Clocked card and didn't really see much FPS improvement over my XFX 5870.

I am thinking it could be my old CPU is holding back the new racehorse and wanted to ask what everyone else thought.
Thanks for the feedback.
 
Not too bad. The 85 and 8600 are the highest IPC chips of the Core2 line.

I think if you overclock it a bit over stock it should really pump some more fuel to the GPU so to say. Are you stock speed right now? I'd go 3.6+ ghz if you can pull it off.
 
Just running at the stock 3.33Ghz speeds right now on air. I've never even overclocked any of my components.
 
I'd say quite a bit. I was running a GTX 460 with a Q9550 (stock speeds as my motherboard was maxed out just to run the CPU), I upgraded from a GTX 275 at the time. I saw no improvement in Bad Company 2. As soon as I upgraded to the 2500k my fps sky rocketed. Granted some games don't need the requirements that BC2 has when it comes to CPU power, I can assure you that there is bottle-necking from your CPU.

Edit: And while I realize going from 275 to 460 isn't the hugest upgrade in the world, the fact of the matter was that I only got the improvement I wanted after the CPU upgrade.
 
your 5870 and 570 were being held back by your CPU setup. my e8400 bottlenecked my 4870x2, which is on par, or a bit faster then a 5870. my e8400 was OC'd to 4ghz too.
 
No fanboi-ism but you should have kept the 5870 and bought a better cpu or cooler + overclocked if this was the only issue, of course maybe you had other issues with the 5870.

Gtx570 is a sexy card though :)
 
John Stamos said:
I'd say quite a bit. I was running a GTX 460 with a Q9550 (stock speeds as my motherboard was maxed out just to run the CPU), I upgraded from a GTX 275 at the time. I saw no improvement in Bad Company 2. As soon as I upgraded to the 2500k my fps sky rocketed. Granted some games don't need the requirements that BC2 has when it comes to CPU power, I can assure you that there is bottle-necking from your CPU.

Edit: And while I realize going from 275 to 460 isn't the hugest upgrade in the world, the fact of the matter was that I only got the improvement I wanted after the CPU upgrade.
Click to expand...

...only because your CPU was stock. There is NO WAY a q9550 running at 3.6+ would bottleneck a 460. Even a 570 isn't really bogged down by a q9550. My system with a [email protected] and gtx570 runs great even when compared to my sig rig.
 
CruisD64 said:
...only because your CPU was stock. There is NO WAY a q9550 running at 3.6+ would bottleneck a 460. Even a 570 isn't really bogged down by a q9550. My system with a [email protected] and gtx570 runs great even when compared to my sig rig.
Click to expand...

very true, my HIGHLY overclockd gtx460 is the bottleneck for me.
 
CruisD64 said:
...only because your CPU was stock. There is NO WAY a q9550 running at 3.6+ would bottleneck a 460. Even a 570 isn't really bogged down by a q9550. My system with a [email protected] and gtx570 runs great even when compared to my sig rig.
Click to expand...

I know, but his argument was whether or not his E8600 was holding him back. I'd argue the Q9550 is reasonably equal to the E8600 in some situations. More so leaning towards the Q9550 with more and more quad core enabled games coming out. In which case my upgrade helped, therefore an upgrade for him would help as well. I was just providing an example of a similar CPU with a bottleneck situation.
 
John Stamos said:
I'd say quite a bit. I was running a GTX 460 with a Q9550 (stock speeds as my motherboard was maxed out just to run the CPU), I upgraded from a GTX 275 at the time. I saw no improvement in Bad Company 2. As soon as I upgraded to the 2500k my fps sky rocketed. Granted some games don't need the requirements that BC2 has when it comes to CPU power, I can assure you that there is bottle-necking from your CPU.

Edit: And while I realize going from 275 to 460 isn't the hugest upgrade in the world, the fact of the matter was that I only got the improvement I wanted after the CPU upgrade.
Click to expand...
a gtx275 to a gtx460 was not much of an upgrade at all as they are within 10% of each other.

as for the OP, your cpu was already limiting your 5870 by quite a bit so of course you saw no improvement going to a gtx570. you are lucky to get much more than half of what a high end card can do in more cpu intensive games.

and tbh a gtx570 is not much faster than a 5870 anyway. even with a good cpu that is a waste of money unless you think 15% is worth the trouble and cost.
 
Last edited:
You are running a cpu that three years ago cost half of what your video card goes for now.
 
I think there's too many blanket statements being used. Your bottleneck will also depend on the resolution you're using not just the games you're playing. If you wanted DX11 then you didn't waste your money (and those of us with a G92 can't do DX11 at all so it's an upgrade even at similar framerates). I bet the 570SC's minimum framerate is higher than your 5870's so you will probably notice smoother gameplay more often than you'll notice a CPU bottleneck.
 
mope54 said:
I think there's too many blanket statements being used. Your bottleneck will also depend on the resolution you're using not just the games you're playing. If you wanted DX11 then you didn't waste your money (and those of us with a G92 can't do DX11 at all so it's an upgrade even at similar framerates). I bet the 570SC's minimum framerate is higher than your 5870's so you will probably notice smoother gameplay more often than you'll notice a CPU bottleneck.
Click to expand...
a 5870 is DX 11 also so yes going from a 5870 to a gtx570 is silly. and with his cpu he will never see the minuscule 15% difference between those cards anyway. and there is about zero chance his minimum framerate went up since his cpu would be the limitation there too.
 
I'm gonna give you the simple answer....

For yesterdays games that most reviewers are still using to review modern hardware, some will be fine and a hand full will hold back a gtx570.
For TODAYS modern quad core optimized games, your cpu is holding the gtx570 back a hell of alot.
For tomorrows games its just gonna get worse.

Its not about resolution, its about the games that use 4 cores, and today, and even more tomorrow ,4 cores are mandatory to get the best performance from your gpu.

For reference here is a list of quad core optimized games made last year MARCH 2010....imagine how many there are now?
http://www.grandtheftpc.com/2010/03/list-of-quad-core-optimized-games.html
 
Last edited:
cannondale06 said:
a 5870 is DX 11 also so yes going from a 5870 to a gtx570 is silly. and with his cpu he will never see the minuscule 15% difference between those cards anyway. and there is about zero chance his minimum framerate went up since his cpu would be the limitation there too.
Click to expand...
At the top end, comparing 100 fps to 120 fps doesn't seem like a big deal but a game will run noticeably smoother with a minimum framerate of 44 instead of 36.

While a 5870 is DX11 everything I read says that the 5xxx series can't handle the tesselation like the 570 can. Are you pointing out that the 5870 card is DX11 because you believe all DX11 cards are equal performers?

happymedium said:
I'm gonna give you the simple answer....

For yesterdays games that most reviewers are still using to review modern hardware, some will be fine and a hand full will hold back a gtx570.
For TODAYS modern quad core optimized games, your cpu is holding the gtx570 back a hell of alot.
For tomorrows games its just gonna get worse.

Its not about resolution, its about the games that use 4 cores, and today, and even more tomorrow ,4 cores are mandatory to get the best performance from your gpu.
Click to expand...
You have to ask him what games he's playing and which resolution he's running at because most of us around here have backlogs of games approaching 40+ games. I don't know which games everyone is anticipating are pushing the envelope of PC's. It's rare to have a game released that was developed PC-centric and most aren't taxing hardware from last generation unless you are running insane resolutions and/or multiple monitors.
 
Last edited:
mope54 said:
At the top end, comparing 100 fps to 120 fps doesn't seem like a big deal but a game will run noticeably smoother with a minimum framerate of 44 instead of 36.

While a 5870 is DX11 everything I read says that the 5xxx series can't handle the tesselation like the 570 can. Are you pointing out that the 5870 card is DX11 because you believe all DX11 cards are equal performers?
Click to expand...
you said DX11 as if the 5870 did not have it. you are living in fantasy land if you think his min framerates went up going to a gtx570 with his dual core cpu. do you even even have a Core 2 duo? well I do and have used different cards in this system and know first hand. not to mention I have the ability to look at reviews.

if he had upgraded from a much slower gpu to a gtx570 then you might have a point. however he was already very cpu limited with the 5870 he already had. and being realistic he is probably at 1920 at best and could be running an even lower res for all we know.
 
John Stamos said:
I know, but his argument was whether or not his E8600 was holding him back. I'd argue the Q9550 is reasonably equal to the E8600 in some situations. More so leaning towards the Q9550 with more and more quad core enabled games coming out. In which case my upgrade helped, therefore an upgrade for him would help as well. I was just providing an example of a similar CPU with a bottleneck situation.
Click to expand...

needmorecarnitine said:
you should list which games you are playing to get better answers.
Click to expand...

I'd say if you're playing a modern game like Bad Company 2 that utilizes more than 2 cores then the simple answer is yes, your current CPU is bottlenecking your GPU.

Not all games will have their physics, sound, networking, etc. in different threads - Games coded like this won't show an advantage between dual and quad core CPUs.

If possible list the games you currently play and are looking into purchasing in the future.
 
RoscoP said:
I'd say if you're playing a modern game like Bad Company 2 that utilizes more than 2 cores then the simple answer is yes, your current CPU is bottlenecking your GPU.

Not all games will have their physics, sound, networking, etc. in different threads - Games coded like this won't show an advantage between dual and quad core CPUs.

If possible list the games you currently play and are looking into purchasing in the future.
Click to expand...
its not just about games that use more than 2 cores. I wish everybody would stop forgetting that newer cpus like Sandy Bridge are much faster clock for clock and tear a Core 2 duo to pieces even in games like SC 2 that only uses 2 cores.
 
@ mope54...I think you missed my edit...

"For reference here is a list of quad core optimized games made last year MARCH 2010....imagine how many there are now?
http://www.grandtheftpc.com/2010/03/...zed-games.html

I went from a e8400 @ 3.6 to a q9550 @ 3.6. Guess where my minimum fps went ? they went up at least 25% and more depending on the game and yes thats with the same gpu with both cpu's.
 
happymedium said:
@ mope54...I think you missed my edit...

"For reference here is a list of quad core optimized games made last year MARCH 2010....imagine how many there are now?
http://www.grandtheftpc.com/2010/03/...zed-games.html

I went from a e8400 @ 3.6 to a q9550 @ 3.6. Guess where my minimum fps went ? they went up at least 25% and more depending on the game and yes thats with the same gpu with both cpu's.
Click to expand...
Yes, I missed your edit. I have a quad core, as well. But the point remains that there is no blanket statement to cover him because what he experiences will be dependent upon the games he plays and the resolution he plays at. My point is that it's not necessarily a given that he wasted money by upgrading to the 570, I haven't said he won't experience a bottleneck at any time.
 
Any Triple / Quad Core CPU 3Ghz or better, with DDR3 1333MHz or better, and with a 64-Bit O.S. = Your Graphics Card bottleneck = GONE!

Before :mad:
After :) = YOU :p
 
Hmmm so it sure seems like my processor is holding me back. I totally cannot run Crysis 2 with the new DX11 patch, I get really bad FPS even with my GTX 570 SuperClocked
 
this is an impossible question to answer accurately without knowing what the native resolution of your display is.
 
christianw said:
this is an impossible question to answer accurately without knowing what the native resolution of your display is.
Click to expand...
I thought max resolution depends on the monitor's dot pitch. Do CRTs have a "native" resolution?
 
TheQuestian said:
Pretty interesting article posted on Tom's. Not sure what your opinion is of the site, but I found the data pretty useful.

Finding CPU/GPU Bottlenecks
Click to expand...
the article is useless for determining how many cores you need on cpus with a different architecture. for example an E8600 oced to 4.0 would still be slower than an i5 at 3.0 in nearly every case. i7/i5 cpus are faster because they also have higher ipc not just more cores.
 
cannondale06 said:
the article is useless for determining how many cores you need on cpus with a different architecture. for example an E8600 oced to 4.0 would still be slower than an i5 at 3.0 in nearly every case. i7/i5 cpus are faster because they also have higher ipc not just more cores.
Click to expand...
Haha, okay, sorry! Thought it might at least help determine which applications are more GPU- vs. CPU-dependent. I will go back to my room, now. :cool:
 
TheQuestian said:
Haha, okay, sorry! Thought it might at least help determine which applications are more GPU- vs. CPU-dependent. I will go back to my room, now. :cool:
Click to expand...
sorry I was not trying to be rude to you at all. its really those tech sites that do those articles that bother me. it just leads people to very inaccurate conclusions on how much cpu power and cores are needed. it would be like me shutting off one core of my E8500 and coming to the conclusion that any single core cpu can handle a game because my one core did.
 
Last edited:
cannondale06 said:
sorry I was not trying to be rude to you at all. its really those tech sites that do those articles that bother me. it just leads people to very inaccurate conclusions on how much cpu power and cores are needed. it would be like me shutting off one core of my E8500 and coming to the conclusion that any single core cpu can handle a game because my one core did.
Click to expand...
Oh, no worries, man. Glad to have a second opinion on it to keep a balanced discussion.
 
just my 02 generally speaking the following is true

if you want to know where your GPU bottleneck is, keep increasing the resolution until performance drops off (you might have to keep adjusting graphics settings until you find the curve as you might have your GPU able to keep up until you hit the high settings) IE frams stay the same at 720p up to max res on your display of lets say 1080p on meduim and start to fall off after 720p on high.

if you want to know where your CPU bottleneck is keep increasing the graphical quality until the frame rate begings to drop off. (typically done on the highest display resolution setting)

It will take some working as you might have to adjust both settings to map out where the bottle necks are

Putting a 6970 in a rig with a C2D e8500 = CPU bottleneck as the 6970 will be able to run the game at any single resolution up to the 6790 res limit at the same frame rate (eg 640*480 up to 2560*1600)

The best way to find a CPU bottleneck is to use a card that you KNOW is capable of high res gaming with all the details cranked up

the best way to find a GPU bottleneck is to use a cpu that you KNOW is capable of delivering a crapload of data to a GPU (fastest i7 for example)

putting a 4870 in a rig with a C2D e8500 = GPU bottleneck at anything above 1650*1080 for your average game as frames will definately start to drop off. I know that on my righ Mafia 2 stutters on highest settings at 1680*1050 when scenes with large amts of debris are rendered.
 
That is not going to work if you do so in single place.
Each game will have areas that are CPU bottlenecked and GPU bottlenecked.

So you would at least need to compare whole levels.
 
Do you really think that resolution doesn't effect the load on the cpu? GTA4, Oblivion, New Vegas, FO3, and FSX are prime examples of games where upping the resolution will up the load on the cpu. Believe it or now BFBC2 with a good video card can also do the same.
 
Michaelius said:
That is not going to work if you do so in single place.
Each game will have areas that are CPU bottlenecked and GPU bottlenecked.

So you would at least need to compare whole levels.
Click to expand...
yeah you are 100% correct there. for example Splinter Cell Conviction had some parts where my E8500 could not even get but around 30 fps. I thought it was my gpu but lowering the res to 1280x720 gave the same framerate. other parts of the game my gpu was the limitation.
 
cannondale06 said:
and tbh a gtx570 is not much faster than a 5870 anyway. even with a good cpu that is a waste of money unless you think 15% is worth the trouble and cost.
Click to expand...

Wrong! An EVGA 570SC is WAY faster than a 5870 by any measure given a fast enough enough CPU (i7 3.6GHz and up). (Keep in mind the 570SC is actually on the 580 performance level and can be overclocked from there)

Fact is: The E8600 is a serious bottleneck to the Evga 570SC card.
 
BBA said:
Wrong! An EVGA 570SC is WAY faster than a 5870 by any measure given a fast enough enough CPU (i7 3.6GHz and up). (Keep in mind the 570SC is actually on the 580 performance level and can be overclocked from there)

Fact is: The E8600 is a serious bottleneck to the Evga 570SC card.
Click to expand...

a 5870 delvers 84% of the performance of a gtx570 so no I am not wrong. http://www.techpowerup.com/reviews/ASUS/GeForce_GTX_570_Direct_Cu_II/23.html

a 570 is SMALL upgrade when you already have a 5870. and the 5870 can be oced too so just because that 570 is factory oced does not change the fact that it is not much of an upgrade at all. and of course with his cpu there is probably zero difference between a gtx570 and 5870.
 
Last edited:
BababooeyHTJ said:
Do you really think that resolution doesn't effect the load on the cpu? GTA4, Oblivion, New Vegas, FO3, and FSX are prime examples of games where upping the resolution will up the load on the cpu. Believe it or now BFBC2 with a good video card can also do the same.
Click to expand...

uhm not really as the fillrate on these new cards is in the multi Gigapixel range. games are hardly fill rate limited anymore.even when you get to ultra high resolutions. Cards now adays are limited in compute power.

there is not a cpu available today that can expose the fillrate limitions of a 6970 or a 580...
 
cannondale06 said:
a 5870 is DX 11 also so yes going from a 5870 to a gtx570 is silly. and with his cpu he will never see the minuscule 15% difference between those cards anyway. and there is about zero chance his minimum framerate went up since his cpu would be the limitation there too.
Click to expand...

This is why people with litlle technical knowledge should read/ask about their "issue", i mean this prob is glaringly obvious it's the cpu - a 5s google would tell you that.

I'm kinda gutted for the op that he had to shell out for a 570/sell 5870 when it was never the underlying issue :(, bad luck to you - but at least when you overclock that cpu you should notice incredible gains in cpu limited games.
 
You must log in or register to reply here.
Back
Top