Jump to content

Frame Rates Issue

DevilishBooster
Go to solution Solved by tmlhalo,

I've been focusing on getting important files transferred and resetting system settings, but I was able to get one game installed that had significant issues with freezing/stuttering (SW-KOTOR). It's an older game so I felt it would be a good test as it should not have been freezing even on the WD Black2 drive. Initial results seem to show that the problem has been fixed, so I'm confident in saying that the issue was the HDD portion of the Black2. I will post a second update and (hopefully) mark this as solved once I am able to install a couple more games and run real tests. Here are the HDTune results I got immediately after getting the drive set up and the OS installed.

This is the Intel 530 Series m.2 that holds the OS and core programs.

attachicon.gif2015-01-13_2100.png

These are the two Crucial M550 SSDs holding the games/higher demand programs.

attachicon.gif2015-01-13_2105.pngattachicon.gif2015-01-13_2118.png

And this is a WD Black HDD I'm using for file storage since I don't need lightning fast access for Word and Music.

attachicon.gif2015-01-13_2123.png

Everything looks right by the charts. Access times are low. Transfer rate is sitting above 200 MB/s for the SSDs. As you can see in your 2.5 5400 rpm WD 640 GB drive it doesn't have those extreme dips like the WD Black^2 did. Some part of that Black^2 was defective. Glad to hear the SW KOTOR tests turned out good. If you have any free roam open world games they usually are the most demanding for dynamically loading parts of the map. Though 99.99999% positive the issue is fixed.

I'm also going to try logging the numbers doing the same tests to see if it logs a difference.

From what I can find online the R9 280 should be staying around 1 - 1.25v. The only thing I could find about adaptive power for AMD was an ultra low power state feature but that won't apply unless the display goes to sleep. If volts while the card is idle are between 1-1.25v then AMD might not have that technology. A quick glance will solve that. I'd also compare the idle amps and watts (vddc current / vddc power).  

Link to comment
Share on other sites

Link to post
Share on other sites

From what I can find online the R9 280 should be staying around 1 - 1.25v. The only thing I could find about adaptive power for AMD was an ultra low power state feature but that won't apply unless the display goes to sleep. If volts while the card is idle are between 1-1.25v then AMD might not have that technology. A quick glance will solve that. I'd also compare the idle amps and watts (vddc current / vddc power).  

I just ran a log while just doing menial browsing/WE tasks. My computer had only been running for about 5 minutes before I started the log, with no heavy graphics use.

From what I can tell it averaged:

~0.810V

0.3A with random spikes up to 1.0A-1.2A (and even a couple spikes up to 2.4A and even one up to 3.0A) and occasional drops to 0.0A

and 0.1W with randoms spikes ranging from 1.2W up to 3.0W. 

Idle GPU-Z Sensor Log.txt

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

what if you set the fan speed in the CCC to run at a constant 40%...??

 

are these games internet related?

Link to comment
Share on other sites

Link to post
Share on other sites

what if you set the fan speed in the CCC to run at a constant 40%...??

 

are these games internet related?

They are internet related in that they can be played online, but it is not a server ping lag issue, it is a local graphics issue. Thermal Throttling has already been ruled out because repeated tests have shown that the neither the CPU or GPU reach those temps.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

understand the temps you are getting, but those temps are only the temps reached by the processor ( on the card ) not the vrm area.

 

setting fan speed to run at a steady pace will not allow the card to reach peak heat before the fans spin up. doing this ( set fan speed ) will help keep things a more constant temperature.

 

just trying to test stuff here.

 

 the internet connection......... wifi by any chance?

Link to comment
Share on other sites

Link to post
Share on other sites

understand the temps you are getting, but those temps are only the temps reached by the processor ( on the card ) not the vrm area.

 

setting fan speed to run at a steady pace will not allow the card to reach peak heat before the fans spin up. doing this ( set fan speed ) will help keep things a more constant temperature.

 

just trying to test stuff here.

 

 the internet connection......... wifi by any chance?

Nope, direct wired. It's not an internet connection issue. It happens in every game that I play. Ghosts is the only game that I play online. In DA:O it happens regularly (along with all the other games that I listed in the original post), and I'm just playing the campaigns.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

use a graphics card utility and under clock the cards cpu and ram. the over clocked graphics card may be too close to it's limits causing these problems. under clocking could very well help. ( msi afterburner or similar )

Link to comment
Share on other sites

Link to post
Share on other sites

use a graphics card utility and under clock the cards cpu and ram. the over clocked graphics card may be too close to it's limits causing these problems. under clocking could very well help. ( msi afterburner or similar )

It's not OCd from the factory, it's just OC ready. Compared to the ASUS model it actually has a lower core GPU frequency and a lower boost clock frequency.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

use a graphics card utility and under clock the cards cpu and ram. the over clocked graphics card may be too close to it's limits causing these problems. under clocking could very well help. ( msi afterburner or similar )

Just out of curiosity, I set the GC fans and case fans all to 100% and tried running Dragon Age Origins again while logging with GPU-Z. I have two 120mm fans in the front of my case and the single rear 120mm fan. As shown in the logs, the GPU never got over 46C and I still had the issue of the frame rates freezing/stuttering at least 7 times in about 3-4 minutes of playing. The VDDC stays constant at about 1.25V, but by looking at the log data you can see that the Amps and Watts both dip down like in the other tests.

DAO Max Fans GPU-Z Sensor Log.txt

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

ever check the memory for errors? or did you ever try running 1 stick at a time and then try them each in different slots?

 

and what happens if you set the ram to run at 1333?

Link to comment
Share on other sites

Link to post
Share on other sites

Just out of curiosity, I set the GC fans and case fans all to 100% and tried running Dragon Age Origins again while logging with GPU-Z. I have two 120mm fans in the front of my case and the single rear 120mm fan. As shown in the logs, the GPU never got over 46C and I still had the issue of the frame rates freezing/stuttering at least 7 times in about 3-4 minutes of playing. The VDDC stays constant at about 1.25V, but by looking at the log data you can see that the Amps and Watts both dip down like in the other tests.

attachicon.gifDAO Max Fans GPU-Z Sensor Log.txt

Something seems up with the power delivery. Even for Idle. We'll see if we can do a few things to fix it if not then might need to RMA the gpu and power supply. Unless you have one of those tools that can measure the voltage the psu is putting out specifically, I don't have any ideas at the moment from discerning between the psu and gpu vrms. MSI Afterburner will let you increase the power draw of the card.  I'd try increasing the power limit to 110% then 120%. See if there is any difference on a couple more trial runs.

Link to comment
Share on other sites

Link to post
Share on other sites

ever check the memory for errors? or did you ever try running 1 stick at a time and then try them each in different slots?

 

and what happens if you set the ram to run at 1333?

Won't running the RAM at a lower speed than the factory settings increase the latency, and therefore increase graphics lag?

Edit: Yes, I have run RAM error tests and everything is running fine.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

Something seems up with the power delivery. Even for Idle. We'll see if we can do a few things to fix it if not then might need to RMA the gpu and power supply. Unless you have one of those tools that can measure the voltage the psu is putting out specifically, I don't have any ideas at the moment from discerning between the psu and gpu vrms. MSI Afterburner will let you increase the power draw of the card.  I'd try increasing the power limit to 110% then 120%. See if there is any difference on a couple more trial runs.

Ok, that's where I am now thinking the issue is. I will see what I can do about increasing the power limit and get back to you. Thanks for all your help so far!

EDIT: I do have a multi-meter as well as other electronics/circuit testing tools, so I'll look up some info and see if I can test the PSU.

EDIT-EDIT: My multi-meter is not digital, so I guess I can't use it for testing the PSU because it can potentially damage the computer circuits. I'm going to see if I can borrow one from a friend.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

Something seems up with the power delivery. Even for Idle. We'll see if we can do a few things to fix it if not then might need to RMA the gpu and power supply. Unless you have one of those tools that can measure the voltage the psu is putting out specifically, I don't have any ideas at the moment from discerning between the psu and gpu vrms. MSI Afterburner will let you increase the power draw of the card.  I'd try increasing the power limit to 110% then 120%. See if there is any difference on a couple more trial runs.

According to my mobo utility, the Voltage coming from the PSU is fine, so it must be something with the power circuit in the GC.

post-63986-0-23318400-1420260223_thumb.p

Using MSI AB I upped the power limit to 110% and 120% as you suggested and it didn't help. In fact, it almost seemed to make it a little worse. Looking at the data logs the VDDC Amps in both test seems to have fluctuated even more, while the VDDC Watts continued with their fluctuation and random drops to ~8.0W.

MSIAB 110 DAO GPU-Z Sensor Log.txt

MSIAB 120 DAO GPU-Z Sensor Log.txt

Any thoughts?

I am going to email Sapphire tonight to see what they say.

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

According to my mobo utility, the Voltage coming from the PSU is fine, so it must be something with the power circuit in the GC.

attachicon.gif2015-01-02_2240.png

Using MSI AB I upped the power limit to 110% and 120% as you suggested and it didn't help. In fact, it almost seemed to make it a little worse. Looking at the data logs the VDDC Amps in both test seems to have fluctuated even more, while the VDDC Watts continued with their fluctuation and random drops to ~8.0W.

attachicon.gifMSIAB 110 DAO GPU-Z Sensor Log.txt

attachicon.gifMSIAB 120 DAO GPU-Z Sensor Log.txt

Any thoughts?

I am going to email Sapphire tonight to see what they say.

Well, it is probably the vrms on the GPU, hopefully Sapphire can replace or shed some light on this. I included a screen shot of my GPU with BF3. The VDDC is pretty much a straight line. Even though it is Nvidia instead of AMD I can't see it being too different. GPUz doesn't show the amps and watts for mine.

post-165866-0-03606400-1420261793.png

Just to toss out the last bit of software related stuff. In case it is one of those hardware issues that can be fixed by software. You've installed everything related to the motherboard drivers? 

http://www.asrock.com/mb/AMD/Fatal1ty%20990FX%20Killer/?cat=Download&os=Win8a64

Updated the BIOS to the newest version?

http://www.asrock.com/mb/AMD/Fatal1ty%20990FX%20Killer/?cat=Download&os=BIOS

This is just a program I use to find health issues with my PC. It isn't totally free but the free demo does find a lot of stuff. Demo is the small link in the middle.

http://www.iolo.com/downloads/download-system-mechanic/

Other than that have you done a virus scan? I usually recommend Malwarebytes if you need a free one. 

Link to comment
Share on other sites

Link to post
Share on other sites

Well, it is probably the vrms on the GPU, hopefully Sapphire can replace or shed some light on this. I included a screen shot of my GPU with BF3. The VDDC is pretty much a straight line. Even though it is Nvidia instead of AMD I can't see it being too different. GPUz doesn't show the amps and watts for mine.

attachicon.gifgpuz.png

Just to toss out the last bit of software related stuff. In case it is one of those hardware issues that can be fixed by software. You've installed everything related to the motherboard drivers? 

http://www.asrock.com/mb/AMD/Fatal1ty%20990FX%20Killer/?cat=Download&os=Win8a64

Updated the BIOS to the newest version?

http://www.asrock.com/mb/AMD/Fatal1ty%20990FX%20Killer/?cat=Download&os=BIOS

This is just a program I use to find health issues with my PC. It isn't totally free but the free demo does find a lot of stuff. Demo is the small link in the middle.

http://www.iolo.com/downloads/download-system-mechanic/

Other than that have you done a virus scan? I usually recommend Malwarebytes if you need a free one. 

I sent off a pretty detailed CS Ticket to Sapphire, so we'll see what they say. I used that exact same driver webpage on my laptop and a flashdrive when I first put the computer together to install all the up-to-date drivers (didn't even both using the included CD), but just for kicks and giggles I'll download everything again and try re-installing them for a third time to see if anything changes. BIOS is supposedly up to date but I'm going to run ASRock's BIOS update Utility again anyway. I'll check out that program you sent the link for, thanks! And yeah, one of the very first things I did was run my anti-virus (Avast! Premier) and Malwarebytes to see if that was the issue. :-/

We'll just have to wait to see what Sapphire send back. Thanks again for your help so far!

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

I think that your issue is your power supply, Rosewill power supplies are poor at best and absolutely horrendous at worst, I would get rid of it even if it isn't the cause of your problems.

What are the ratings of the PSU? Read the specs on the sticker on the PSU and type them here.

Basically a good 650W PSU should only have one 12V rail and it should be outputting around 50/54Amps on that rail. If it has 2 or more 12V rails or it has less than 50Amps on that rail then it most likely isn't giving your GPU enough power to run properly.

And that would cause your GPU to throttle in order to keep its power levels within what the PSU can output, which would cause your stuttering in games, and would also explain why when you lower the graphics details to low the stuttering happens a lot less frequently.

CPU: Core i5 2500K @ 4.5GHz | MB: Gigabyte Z68XP-UD3P | RAM: 16GB Kingston HyperX @ 1866MHz | GPU: XFX DD R9 390 | Case: Fractal Design Define S | Storage: 500GB Samsung 850 EVO + WD Caviar Blue 500GB | PSU: Corsair RM650x | Soundcard: Creative Soundblaster X-Fi Titanium
Click here to help feed our lasses Pokemon

Link to comment
Share on other sites

Link to post
Share on other sites

I think that your issue is your power supply, Rosewill power supplies are poor at best and absolutely horrendous at worst, I would get rid of it even if it isn't the cause of your problems.

What are the ratings of the PSU? Read the specs on the sticker on the PSU and type them here.

Basically a good 650W PSU should only have one 12V rail and it should be outputting around 50/54Amps on that rail. If it has 2 or more 12V rails or it has less than 50Amps on that rail then it most likely isn't giving your GPU enough power to run properly.

And that would cause your GPU to throttle in order to keep its power levels within what the PSU can output, which would cause your stuttering in games, and would also explain why when you lower the graphics details to low the stuttering happens a lot less frequently.

It's a single rail PSU with the output: +3.3V @ 24A, +5V @ 24A, +12V @ 54A, -12V @ 0.5A, and +5VSB @ 3A. I'm haven't totally ruled out the PSU, but I just got some new data today. I used to be able to run Far Cry 3 on High just fine, holding at 50-60 FPS consistently with minimal frame rate freezing. I haven't played it for about a week and when I fired it up today I found that it was running between 18-25 FPS on the setting I had previously used, and it was even struggling to run at 30 FPS with everything set to Low. I had GPU-Z open on the other monitor and I found that even with everything turned to Low my GPU was still pinned at 99%. Any thoughts on why that would change so drastically?

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

@tmlhalo @Robin88

I'm having a thought here (said in Capt Jack Sparrow homage)...

The PSU came with 2 cables for GCs. The both have an 8-pin plug at the PSU end but there are only 6 wires. At the other end it splits into 2x 6+2 pin connectors.

post-63986-0-10800500-1420322208_thumb.j

As of right now this is how they are plugged into the GC.

post-63986-0-03244300-1420322207_thumb.j

Right now I obviously only have the one cable run. Here's my question(s). Does it matter which header on the GC each of those 6+2 pin plugs go into? Should I try running the second cable to the card and only having 1 of the 6+2 pin from each cable plugged into each header on the same card? Right now I'm just trying to throw out any idea that might help narrow down the suspect list. Let me know. Thanks

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

@tmlhalo @Robin88

I'm having a thought here (said in Capt Jack Sparrow homage)...

The PSU came with 2 cables for GCs. The both have an 8-pin plug at the PSU end but there are only 6 wires. At the other end it splits into 2x 6+2 pin connectors.

attachicon.gif10918391_10153056268384319_1404378891_o.jpg

As of right now this is how they are plugged into the GC.

attachicon.gif10886306_10153056282674319_1565701746_o.jpg

Right now I obviously only have the one cable run. Here's my question(s). Does it matter which header on the GC each of those 6+2 pin plugs go into? Should I try running the second cable to the card and only having 1 of the 6+2 pin from each cable plugged into each header on the same card? Right now I'm just trying to throw out any idea that might help narrow down the suspect list. Let me know. Thanks

To be honest the detailed inner workings of a power supply are a little beyond my experience. I pulled one of the spare PCIe cables I have from my psu out the bag and shows this.

post-165866-0-68069100-1420326222_thumb.

12 pin into the psu to 2 (6 + 2) pin connectors. So it certainly couldn't hurt trying 2 different cables instead of a single cable. I know a 6 pin PCIe connector is 150w and an 8 pin is 250w. You can run a larger amount of power by having a thicker gauge wire but I don't know what Rosewell has done specifically. On the side note did System Mechanic find anything? Also while you change the PCIe 6 pin cable maybe try plugging the computer directly into the wall if you happen to be running through an old surge protector? Any word from Sapphire?

Link to comment
Share on other sites

Link to post
Share on other sites

I never liked 5400rpm drives. maybe the seek time is too long?

 

what other apss are you running at the same time you are gaming.?

Link to comment
Share on other sites

Link to post
Share on other sites

To be honest the detailed inner workings of a power supply are a little beyond my experience. I pulled one of the spare PCIe cables I have from my psu out the bag and shows this.

attachicon.gifIMG_20150103_175856759.jpg

12 pin into the psu to 2 (6 + 2) pin connectors. So it certainly couldn't hurt trying 2 different cables instead of a single cable. I know a 6 pin PCIe connector is 150w and an 8 pin is 250w. You can run a larger amount of power by having a thicker gauge wire but I don't know what Rosewell has done specifically. On the side note did System Mechanic find anything? Also while you change the PCIe 6 pin cable maybe try plugging the computer directly into the wall if you happen to be running through an old surge protector? Any word from Sapphire?

System Mech only found some junk files and registry errors that it corrected, but nothing major from what I could tell. I'll try running a second cable but with some data I just collected I don't know that it is going to help. (See below) I haven't heard from Sapphire yet, and I don't really expect to until sometime this next week.

 

I never liked 5400rpm drives. maybe the seek time is too long?

 

what other apss are you running at the same time you are gaming.?

I am starting to wonder if it is the slower HDD causing a bottle neck as part of the problem, because I just used FurMark to run their 720 and 1080 benchmarks as well as the GPU stress test and the power draw is fine and steady where it should be. So I'm starting to wonder if maybe the GC is reducing the power draw when it's sitting there waiting for data from the HDD, and then once it gets the data it needs to kick back in, causing the stutter/freezing. But that still leave the question of why Far Cry 3 is suddenly running so crappy on low settings now when a week ago it was running great on high settings. I'm at a loss at this point. Maybe I just need to save up and buy a large SSD to have my games on to see if that fixes the issue. :-/

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

System Mech only found some junk files and registry errors that it corrected, but nothing major from what I could tell. I'll try running a second cable but with some data I just collected I don't know that it is going to help. (See below) I haven't heard from Sapphire yet, and I don't really expect to until sometime this next week.

 

I am starting to wonder if it is the slower HDD causing a bottle neck as part of the problem, because I just used FurMark to run their 720 and 1080 benchmarks as well as the GPU stress test and the power draw is fine and steady where it should be. So I'm starting to wonder if maybe the GC is reducing the power draw when it's sitting there waiting for data from the HDD, and then once it gets the data it needs to kick back in, causing the stutter/freezing. But that still leave the question of why Far Cry 3 is suddenly running so crappy on low settings now when a week ago it was running great on high settings. I'm at a loss at this point. Maybe I just need to save up and buy a large SSD to have my games on to see if that fixes the issue. :-/

In your primary post you said your hard disk drive is Western Digital 1 TB Black? Do you happen to have a link to the model number? I'm fairly sure current WD 1 TB blacks are 7200 RPM. 

Link to comment
Share on other sites

Link to post
Share on other sites

the one I found was 5400.

 

did you do a cmos reset by any chance? wonder what the ram is actually running at. maybe something a little off and a reset might help. ( voltages/speed, etc. )

Link to comment
Share on other sites

Link to post
Share on other sites

In your primary post you said your hard disk drive is Western Digital 1 TB Black? Do you happen to have a link to the model number? I'm fairly sure current WD 1 TB blacks are 7200 RPM. 

It's a 5400rpm HDD. It's the Black^2 (Black Squared)

http://www.newegg.com/Product/Product.aspx?Item=N82E16822236642

END OF LINE

-- Project Deep Freeze Build Log --

Quote me so that I always know when you reply, feel free to snip if the quote is long. May your FPS be high and your temperatures low.

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×