Jamesaevans’s Gaming HTPC Build Log

As far as monitoring temps go there are a whole host of apps out there but if you hop to the CPUID site there are a good selection of monitoring apps including CPUID Hardware Monitor.
CPUID - System & hardware benchmark, monitoring, reporting
Great little app and does exactly what it says on the tin and gives real time plus min & max temp readouts.

Prime95 would be a good choice for stress testing.

Hi RedMonkey,

Thanks for the advice, just spent the last hour checking out the utilities that you recommened. :thumbsup:

The CPUID Hardware Monitor looks very good as it dispays voltages, temperatures, fans speed and logs min and max values. Along side it Prime95 looks to be a good way to test the CPU and Memory.

So that just leaves the graphics card, I'm sure I could just play some games or try out Nvidia - Design Garage. But if I want to stress test the graphics card do you have any thoughts on the various utilities from Futuremark or any other alternatives?

One final question, what sort of temperatures should I be looking for?

Thanks
James
 
3D Mark 11 is a good one for stressing your GPU. I've avoided Furmark because I hear it's a card-killer.

You can download MSI Afterburner (and Kombuster) which allows you to assign overclocking profiles to your card and stress test it. Also you can use Core Temp which is useful and has a configurable alarm that warns you when your temps hit a pre-defined setting - handy if you're paranoid about burning something out.
 
Hi RedMonkey,

So that just leaves the graphics card, I'm sure I could just play some games or try out Nvidia - Design Garage. But if I want to stress test the graphics card do you have any thoughts on the various utilities from Futuremark or any other alternatives?

Futuremark...3dmark06 or 3dmark11 should suit your purposes for testing out your graphics card. Canyon Flight test in 3dmark06 has been a recommended function and still seems to get used for the majority of review benches.

One final question, what sort of temperatures should I be looking for?

Thanks
James

Great question James, temps are always quite subjective and this is where the obsession begins for me for getting best performance at the lowest temps....and I know I'm not alone in this forum with that particular obsession:D. That said, I think you've built wisely and done as much as you can for good cooling on air.....

It all depends on how much of an OC you apply to either the the CPU or GPU as this will push the heat up.
I had a good look around for the thermal range of you GPU but couldn't find anything from the manufacturer that confirmed. You may want to check out the links to a couple of reviews that give the thermal readouts for idle and load. Obviously, you're temps will be not be the same due to differing hardware and case but I think you could still use this as a rough guide quite comfortably.
Gainward GTX 570 Phantom Power and Thermals | bit-tech.net
Gainward GeForce GTX 570 Phantom - Overclockers Australia

With regards to your CPU....it already runs cool ;). But if you check out the Intel site for your chip the thermal ceiling is:
TCASE 72.6°C
Anything above this and you'll be cooking :thumbsdow

Quite honestly I think for both your CPU & GPU you could be looking at temps of 30c - 60c depending on load etc.
I very rarely go over approx 40c when idling.....I consider idling doing normal tasks like watching TV & recording etc and 40c is on a very hot day!
 
I ran Prime95 yesterday. I would suggest rather than "benchmark" actually using the programme to search for primes seems to keep the cpu cores at 100% for longer. During the benchmarking not all cores are active all the time.

As for max temps, you will read a number of conflicting account of maximum temps. And you will read things about consistent long term temps vs peak load temps.

But basically in these benchmarking exercises you want to be below the Tj-max, which for SandyBridge is 98°C
Some of these temp progs (like CoreTemp) can report distance from Tjmax rather than actual temp. The advantage of this is that you might note that RealTemp and CoreTemp and other tools can report different CPU temps, but if you ask the programme to report the Tjmax, they also vary, generally by the same amount.

So if you like, the headroom will be the same on either programme. The other point to note is unlike older chips, tjmax is read from the cpu directly and so temps should be accurate across a broad range.

For reference. My MoBo idles aroung 32°C and CPU between 27-36°C depending on power saving mode, but nearly always no higher than 16x multiplier.
During Prime95 the multiplier went to 37x and temps to around 60°C using stock cooler and Asus set to quiet (so cooler speed only increased 10%, where I know it can go from around 1000 rpm to 2000 rpm).
 
13 - Testing (Idle, Blu-ray and 3D Rendering)
Thank you everyone for the advice regarding temperature monitoring and stress testing. Based on the advice I have downloaded CPUID Hardware Monitor and performed some testing this evening.

Idle Temperatures
I started by leaving the system idle for about 30mins to get a benchmark to compare against the other tests. Here is a screen shot showing the CPUID Hardware Monitor results:
5688201510_5410c948c3_z.jpg

Motherboard Temps: 32'C / 25'C / 28'C
Fan Speed: 1,000 RPM
CPU Core Temps: 30'C
GPU Temp: 36'C
GPU Fan Speed: 1,200 RPM
GPU Fan Power: 30%

One thing I did notice at this point is that FANIN0, which is the intake fan on the CPU cooller and connected to the CPU FAN header on the motherboard was only showing 26 RPM. Looking at the two fans on the CPU Cooler both appeared to be running at the same speed, and certainly if the fan had been running at just 26 RPM I'm sure this would have been very obvious.

I swapped the two fans over and both the readings in CPUID Hardware Monitor and visual speed of the fans remained constant, so I think this may be an issue with the way CPUID Hardware Monitor is picking up the fan speed from the Gigabyte motherboard, but I'll need to investigate if other owners of this motherboard are having the same problem. :confused:


Blu-ray Temperatures
Next I moved onto a Blu-ray playback test using Kung Fu Panda, which gave the following CPUID Hardware Monitor results:

Motherboard Temps: 39'C / 25'C / 30'C
Fan Speed: 1,200 RPM
CPU Core Temps: 40'C
GPU Temp: 50'C
GPU Fan Speed: 1,200 RPM
GPU Fan Power: 40%

Here are a couple of screen shots showing the Blu-ray image quality (click through to high-res):







I've not have chance to reseach the HD audio bit streaming issue any further, but the sound quality is certainly already better than from my PS3, the clarity of speach in particular is significantly improved.


3D Rendering Temperatures
For the final test this evening I decided to try out some 3D rendering using the Nvidia Design Garage. I let the rendering engine run for 1,000 passses which produced the following results in CPUID Hardware Monitor:

Motherboard Temps: 47'C / 25'C / 43'C
Fan Speed: 1,200 RPM
CPU Core Temps: 51'C
GPU Temp: 72'C
GPU Fan Speed: 2,400 RPM
GPU Fan Power: 62%

Here is the rendered image of a speed yellow Porsche GT2 after 1,000 passes (click through to high-res):


More testing to come with 3dmark11, Prime95 and Need for Speed Shift 2 in the next few days, but any thoughts on the low fan speed for the CPU fan would be very much appreciated.

James
 
Last edited:
Hi mate,

The GTX570 you have, how do you output the audio from it? Does it even have audio via HDMI?

Thanks.
 
One thing I did notice at this point is that FANIN0, which is the intake fan on the CPU cooller and connected to the CPU FAN header on the motherboard was only showing 26 RPM. Looking at the two fans on the CPU Cooler both appeared to be running at the same speed, and certainly if the fan had been running at just 26 RPM I'm sure this would have been very obvious.

I swapped the two fans over and both the readings in CPUID Hardware Monitor and visual speed of the fans remained constant, so I think this may be an issue with the way CPUID Hardware Monitor is picking up the fan speed from the Gigabyte motherboard, but I'll need to investigate if other owners of this motherboard are having the same problem. :confused:

James

Hi James,

I'm wondering if this may be a bug in CPUID HM....download Speedfan and see if you get the same result.
SpeedFan - Access temperature sensor in your computer

I know that CPUID HM has recently gone up a revision so it's possible that's where the fault could be.
If this doesn't prove to be the case then you could try to flash the BIOS on the MB but only do this if others are reporting the same issue in the Gigabyte forums. I'm not familiar with you MB settings but do you have any kind of fan control in the BIOS for min & max speeds?

Overall I think your temps are looking pretty normal for the individual tasks that you set.
 
Hi mate,

The GTX570 you have, how do you output the audio from it? Does it even have audio via HDMI?

Thanks.

Hi Useless_Help,

The GTX 570 graphics card that I have (Gainward GTX 570 Phantom) has a full sized HDMI port which outputs the audio along with the picture. The full sized HDMI port was another reason that I went for this particular make/model of GTX 570 as I was not keen on the graphics cards which come with a DisplayPort or Mini-HDMI port and which require the use of an adaptor, which could come potentially come lose.

Hope that's helpful.

James
 
How come your res is not set to 1080p? That desktop looks like it's in 720p.

Hi wolvers,

The display is set to 1080p, but when I took the screen shot I was trying out a setting in Windows 7 to make the icons and text larger so that it is easier to read / navigate when sitting on the sofa.

James
 
Hi James,

I'm wondering if this may be a bug in CPUID HM....download Speedfan and see if you get the same result.
SpeedFan - Access temperature sensor in your computer

I know that CPUID HM has recently gone up a revision so it's possible that's where the fault could be.
If this doesn't prove to be the case then you could try to flash the BIOS on the MB but only do this if others are reporting the same issue in the Gigabyte forums. I'm not familiar with you MB settings but do you have any kind of fan control in the BIOS for min & max speeds?

Overall I think your temps are looking pretty normal for the individual tasks that you set.

Hi RedMonkey,

Thanks for the suggetsions, I'll try SpeedFan and/or Gigabytes own monitoring software when I get home tonight and will post my results.

I did some research last night and I think the CPU FAN header may have defaulted to PWM control rather than Voltage control based on a couple of posts that I found on other forums. Having consulted the motherboard manual there is a setting in the BIOS for the CPU FAN header control mode [AUTO/PWM/VOLTAGE], so I'll give this a go tonight as well and see if this resolves the issue.

Thanks
James
 
Hi wolvers,

The display is set to 1080p, but when I took the screen shot I was trying out a setting in Windows 7 to make the icons and text larger so that it is easier to read / navigate when sitting on the sofa.

James

Righty ho. :thumbsup:
 
Hi wolvers,

The display is set to 1080p, but when I took the screen shot I was trying out a setting in Windows 7 to make the icons and text larger so that it is easier to read / navigate when sitting on the sofa.

James

set the desktop back to 1080p, then go to control panel- display, and up the dpi :)
 
Hi wolvers,

The display is set to 1080p, but when I took the screen shot I was trying out a setting in Windows 7 to make the icons and text larger so that it is easier to read / navigate when sitting on the sofa.

James

Set the dpi to 125 in control panel or higher and leave the res at 1080p. (right click desktop and then select screen res, now select make writing larger or smaller). You can also customise the text size in web browsers.

You should never have to drop the res to make writing easier to read from the sofa. :)

Edit: just noticed grangeys post. :D
 
Razor, Grangy, I am assuming he might have done that, as that is "a setting in Windows 7". I cannot view the screenshots, it might be different.

As for the fan speed. On my MoBo and stock cooler I get a wrong reading of CPU cooler speed in CPUID-HM. Asus app reports fine. I will check SpeedFan too.

Speaking of these programmes. When I run then windows pops up a box, which you have to click "allow" or something before it runs, as it says the program wishes to make changes to your computer. Is there anyway to stop this, XP was never this annoying.
 
icstm - Go into user accounts and then select change user account settings to off, save and reboot. :)
 
icstm - Go into user accounts and then select change user account settings to off, save and reboot. :)
yes I have seen those settings. (there is a link to that page when the pop up comes). When I read that it makes it sound as though there will be an occasional situation when the pop-up might be useful.

Do most of you switch it off. I must say, having used an IBM/Intel compatible PC since the late 80s I have only once got malware/virus, but that might just be because I have a wooden desk so touch wood each time I use my PC (or maybe I shouldn't be that superstitious)
 
14 - Resolving the CPU fan speed issue

Gigabyte EasyTune6
In an attempt to try and get to the botton of the CPU fan speed issue I installed the EasyTune6 utility from Gigabyte that includes fan speed monitoring. The application installed but refused to launch, which looks like a common error based on the form posts that I found on the subject. Very disappointing that a manufacturer puts out a utility that doesn't run on PCs based on their own motherboard. :mad:

SpeedFan
I then downloaded SpeedFan, which instaled and launched without issue, and showed that the CPU fan was spinning at a much more normal 1,200 RPM rather than the 18 RPM shown in HWMonitor. Click the screen shot for high resolution version:

Therefore, it looks like a bug with the HWMonitor utility, which is a shame as otherwise I like the rest of the utility and the information it provides. I have sent an email to the developer containing a link to the above screen shot so hopefully this will be resovled shortly.


BIOS Settings
To be sure I went into the PC Health section in the BIOS and this confirmed that both fans were operating at around 1,200 RPM.

I had read that in some cases the CPU fan might have been auto configured as a PWM controlled fan when it is actually normal voltage controlled fan, and that this could lead to fan speed monitoring issues. Therefore, I went into the BIOS settings and manually set the fan to "Voltage" in the hope that this might solve the fan speed issue.

However, when I came back into Windows and lauched HWMonitor and SpeedFan to see if this had made any difference, I discovered that the CPU fan had disappeared from both utilities. Therefore, I subsequently reversed this change to the BIOS and will wait to see if the issue with HWMonitor is resolved in a future release.

James
 
Last edited:
15 - Setting the screen magnification

Earlier in this thread wolvers and a couple of others commented that it looked like I had my screen resolution set too low. However, the resolution was set to full HD (1920x1080) but I had enabled screen magnification at 150% which made the text and icons larger and thus easier to read when sat on the sofa in my living room.

To try and illustrate how the magnification setting works I have taken screen shots at 100%, 125% and 150% magnification. On each occasion I resized the browser window to the minimum width required to show the AVForums home page without requiring me to scroll the window horizontally. I hope that makes sense. :)

At normal 100% magnification:
5695630142_337d7eeba7_z.jpg


At 125% magnification:
5695630426_d8d9447791_z.jpg


At 150% magnification:
5695056399_9a21645080_z.jpg



If you look closly on the 125% magnification screen shot I have also expanded the Custom DPI settings window. My inital expectaion when I saw this option was that it would work independently from the magnification settings. However, after some investigation it appears that changing the Custom DPI setting is simply another way of setting the magnification setting.

As everyone will have different screen sizes, sit different distances from the screen and have different eye sight, the magnification setting used will be down to personal preference. For me a magnification setting of 125% is a good compromise, but if I was going to be reading lots of web pages on my HTPC I would probably increase that to the 150% setting.

James
 
Last edited:
As I thought you were altering Windows OS settings.
have you found a way to have this set differently for different monitors plugged in at the same time?
I am trying to set my 19" monitor to 100% and my TV to either 125 or 150%
 
As I thought you were altering Windows OS settings.
have you found a way to have this set differently for different monitors plugged in at the same time?
I am trying to set my 19" monitor to 100% and my TV to either 125 or 150%

Hi icstm,

I only have my HTPC connected to my plasma screen and don't have any other monitors to try connecting at the same time.

I'm guessing that you can only have one magnification setting in Windows irrespective of the number of monitors that you have connected to your PC. If you set your 19" monitor to a higher resolution than your TV (assuming it supports this) then I would think that you might get a similar result. But as I said not something I've ever tried to do.

James
 
16 - Testing (Prime95 and 3DMark11)

Following on from the initial set of testing which focused on temperatures during Idle, Blu-ray playback and 3D Rendering, I undertook a second series of tests using Prime95 and 3DMark11 to test the cooling performance and stability of my build.

For comparison here are the Idle temperatures recorded in the initial testing:

Motherboard Temps: 32'C / 25'C / 28'C
Fan Speed: 1,000 RPM
CPU Core Temps: 30'C
GPU Temp: 36'C
GPU Fan Speed: 1,200 RPM
GPU Fan Power: 30%


Prime95
The Prime95 utility provides the ability to stress test your CPU and Memory by undertaking complex mathematical calculations. To really test the stability of your PC it is recommended that you leave Prime95 running for 24 hours. However, as I was equaly interested in the thermal cooling performance I ran the utility for just over 1 hour before taking this screen shot (click for high res version):

Using HWMonitor the key temperature readings were:
Motherboard Temps: 40'C / 25'C / 53'C
Fan Speed: 1,275 RPM
CPU Core Temps: 62'C
GPU Temp: 47'C
GPU Fan Speed: 1,320 RPM
GPU Fan Power: 34%


3DMark11
I then moved onto the 3DMark11 utility which is designed to test the graphics card. The free version which I downloaded only comes with a limited set of tests and profiles that you can use for testing, but I though it would be adequate for my purposes. After runnng the through tests that were available to me I took the following screen shot (click for high res version):

Using HWMonitor the key temperature readings were:
Motherboard Temps: 48'C / 25'C / 44'C
Fan Speed: 1,250 RPM
CPU Core Temps: 62'C
GPU Temp: 74'C
GPU Fan Speed: 2,490 RPM
GPU Fan Power: 64%


Conclusion
Based upon the results from both sets if testing it appears that my HTPC is stable and the cooling for the motherboard, CPU and graphics card is working well.

Now its time to install a game and see how my HTPC performs when it is used for gaming.

James
 
Last edited:
15 - Testing (Prime95 and 3DMark11)

Conclusion
Based upon the results from both sets if testing it appears that my HTPC is stable and the cooling for the motherboard, CPU and graphics card is working well.

Now its time to install a game and see how my HTPC performs when it is used for gaming.

James

Hi James
Great to see all the hard work you have put in has paid off. Your temps are looking really good and I'm sure after a few hours of sustained gaming you'll still be ok.
 

The latest video from AVForums

Is 4K Blu-ray Worth It?
Subscribe to our YouTube channel
Back
Top Bottom