In short I have upgraded from a 5850 to 7950 and see no real improvement.
More telling is if I run the F1 2012 benchmark (all ultra, 8x AA), I get 52 fps average with 40fps minimum, which is quite a bit under what the 7950 is capable of.
Fair enough if its new processor time but everywhere I read my current CPU should be up to it.
Is there an easy way to test different parts to figure exactly where the bottleneck is?
Thanks
I would say its the 2 x HDDs whats causing the bottleneck effect either get a single Hdd and use the others as extra storage in non raid mode this way it will make that much difference....
I would look at SSD this may help
also change M/B and CPU to a upgrade like socket 1155 as that M/B & CPU is quite oldish now for most games and stuff
Go ivy bridge if you can afford it if not then go Sandy bridge
I think that the mobo/CPU combination that you are running would be the main reason you are experiencing low frame rates, try overclocking the CPU to see if that produces a better benchmark result.Originally Posted by c12038
Er.. can you explain what you mean? I'm stumped by this reply!
OP, change resolution and graphical settings - if lowering the settings *doesn't* improve the framerate then you are CPU bottlenecked.
Ok I did 2 tests, both on F1 again.
Ultra settings 800x600 = the same
Ultra Low setting 800x600 = 60 fps (v-synced)
I guess the ultra low load the processor less (particles effects etc)
So I guess its cpu/board upgrade time.
Had a quick look and it seems i5 2500k or i5 3570k is the way to go.
As above could raid 0 be causing this, I don't have a spare to test with right now?
Thanks for the help so far.
Short answer: Not a chance
Long answer: Depending on your hardware, a software RAID 0 has a minor increase on CPU load. If the CPU is already the bottleneck, then any additional load, be that RAID, sound channels, network chip etc. cuts into already precious CPU time, and so can have a minor effect on framerate if you were thrashing the hard drive at the same time as playing. So yes, it might have an effect on framerate, but it's not the cause, the lack of CPU would be the cause.
Didnt think of that, I guess it wont be making a huge difference.
Hdd on raid consumes cpu processing time
as indicated here :-
Indicates the proportion of CPU cycles spent waiting for IO (disk or network) events. If you experience large IO Wait proportions, it can indicate that your disks are causing a performance bottleneck.
This is known as being IO Bound. To solve this issue, first attempt to optimize your applications disk usage (e.g. caching, delayed writes). From that point, you may need to look at improving disk performance directly through faster drives and/or RAID arrays. (More on Disk IO)
System
Indicates the proportion of CPU cycles spent performing kernel-level processing. Generally you will find only a small proportion of your CPU cycles are spent on system tasks, hence if you see spikes it could indicate a problem. Software RAID setups undergoing rebuild can cause high system CPU consumption.
User
Indicates the proportion of CPU cycles spent performing user instigated processing. This is where you should see the bulk of your CPU cycles consumed; it includes activities such as web serving, application execution, and every other process not owned by the kernel.
Great explanation, but totally wrong in terms of the context. It will only be IO Bound during the IO operations, which at worst would mean the OP gets a lower frame rate during the loading screens, I can't see that being an issue as he will not be able to commence play until the level has fully loaded by which time the IO operations would have ceased.
Having run raid 0 on rigs with 2 drives, 4 drives, 2 drives with an SSD cache drive and 4 drives with an SSD cache I can pretty much confirm that I have yet to experience a noticeable drop in frame rates, with the exception of the rig running 4 drives and an SSD cache drive, which for some reason was performing better with around a 5% higher frame rate then when it was set up with 1 single drive running with an SSD cache drive connected, but I think that had more to do with issues with the Windows installation then the CPU or drive configuration.
It's not just wrong in terms of the context, it's wrong in terms of effect on frame rate. IO bound means that your application has to wait for slow IO before continuing. This will not usually effect framerate so cannot account for the fps 'issue' flagged by the OP.
As I explained in my post a few up, the only way RAID can effect framerate is due to being CPU bound. If you were IO bound then the CPU load would actually reduce, and attempting to solve that might increase CPU load!
Furthermore, if you *are* IO bound, changing to one HDD would likely make the problem worse! So I'm sorry, but there is some very fishy advice being given.
Percy1983, you're right - storage won't make any noticeable difference to your issue. While you might be able to slightly reduce CPU load by reducing the effect of additional things like sound processing, it's not going to be enough to remove the CPU bottleneck. Improving the CPU (which unfortunately necessitates a change in motherboard etc.) will not only allow your 7950 to fly, it'll give enough headroom for all these other tasks as well.
I sort of wish it was right in a way, then instead of spending £600+ on the latest top end GPU I could just spend £150 on a new SSD and my frame rates would rocket.
Are the new intel chips that much better. I would have thought a Q9550 @ 3.53 GHz would still be plenty. If it were me I would chuck in another HDD and install windows to just rule that out.
Don't buy any hardware just yet. I have zero experience with the game in question. And we don't have access to your pc first hand so let's remove some other potential issues
Do you have another game, or a benchmark suite you can run to compare your pc to other rigs first? Maybe the game you play is badly suffering for many people?
Next up is "what's also on your pc running at same time" for example do you have an old version of some dire anti virus running, or multiple other apps hogging your pc that you don't remember are on it
I have an xfi so audio isnt the issue. I am running avast anti virus.
If I open task manager I am idling at 0% CPU with ocasional 1% blips.
What would be a useful benchmark to run?
3d Mark run tests on GPU, CPU and GPU + CPU together so you could try the free version and see how that goes, if you register you can upload your scores and then put a link on this thread to show how your PC coped, like this. If you pay attention to the Physics and combined tests this is where the CPU and CPU+GPU get tested, IIRC.
If you have GTA IV you could take a look at the benchmark utility that's included, if not then you could install the Steam demos for Mafia 2 and Just Cause 2 to see what they achieve as well.
Both Batman Arkham Asylum and Arkham City have benchmark tools, but I'm not sure if they have playable demo versions with the same tool available.
Just ran the 3d mark basic test and got:
3DMark Score
P7590
Graphics Score
9111
Physics Score
5140
Combined Score
4941
Graphics Test 1
40.1 FPS
Graphics Test 2
45.0 FPS
Graphics Test 3
57.39 FPS
Graphics Test 4
27.48 FPS
Physics Test
16.32 FPS
Combined Test
22.98 FPS
Thats with the GPU overclocked as per system specs.
Not sure if thats good or bad.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
LinkBack URL
About LinkBacks
Reply With Quote