- Nov 11, 2013
- 25
- 0
- 10,540
Hey guys,
Last year marked my entrance into PC gaming, and so far I'm loving it. Right now (as I'm typing this, in fact) I am currently OC-ing my CPU. I think I know what I'm doing, but I have some questions. I'm sorry if some of my questions are answered by other threads or resources, but the issue I'm having is some guides mention some things like they are essential to consider while other guides never mention them at all, as well as just completely conflicting information, even from reliable sources.
First off, my specs:
AMD Athlon x4 750k
Sapphire Radeon HD 7870
10G of DDR3 ram in 4 sticks, 3 different brands
ASRock FM2A75 Pro-4 M motherboard
High quality 500W PSU
Air heat sinking (Looking into water/liquid cooling soon; that's my next upgrade
) As I side note, I normally run my computer with the side panel off. I know it increases dust accumulation, but I am excellent at keeping my PC components clean and dust bunny free and I always have duster on hand. For me, it dramatically reduces my CPU temp - around 10c in most cases!
Also, I am a computer scientist so I am comfortable with any lingo you want to throw around.
Next, here is my OC-ing process:
Firstly, a program called ASRock eXtreme Tuner Utility came with my mobo (http://www.youtube.com/watch?v=_7naB8kIZXE). It's a windows program, like others of its kind, that allows for CPU/RAM tuning without needing a restart or interaction with BIOS. I've read up on the program and others that do a similar thing and I am very aware of certain concerns OC veterans have with regards to in-windows OC-ing. I personally am okay with BIOS and can use it if I feel like it's necessary, but I'm not looking to push more than 4.0Ghz or 4.5 at the very max.
That means I don't expect any instabilities to come out of anything except Vcore issues, so I don't need the level of control one gets from BIOS tweaking. Some specific issues I've read about AXTU are temperature readings not being accurate; I have other programs in place that have better credibility with these types of readings. I also like not having to constantly reboot, because repetitious rebooting can put stress on HDDs, especially the older ones I'm using. As for being able to turn off power saving features and other OC nuisances, all of those settings are also available within windows.
I am using CPUZ to monitor whether changes made in AXTU are actually reaching the hardware, as well as to monitor Vcore readings. AXTU also monitors Vcore, but since it has issues with temperature readings, I thought I should also keep an eye on this with another program as well. However unlike temperature readings which have conflicted, so far Vcore readings from AXTU and CPUZ have essentially been the same.
Next, I am using HWMonitor. I swear CPUID is going to own me within a month
. This is obviously being used for temperature monitoring, as well as a third reference for Vcore readings.
To stress test, I am using Prime95.
I am also using task manager to keep a larger view on what's happening.
As for the actual process:
I started off at a multiplier of 36 (baseline for this CPU) with a Vcore of 1.35v (baseline for this CPU). I go into AXTU and bump up the multiplier by one and wait a minute or so as a "basic stability test." After making sure it's stable at low demand, I then pull up Prime95 and start a torture test. While testing, I watch the CPU temperature to make sure it does not exceed 70c and watch the Vcore to make sure it doesn't jump from 1.35 to 1.6. I let Prime95 go for 10 minutes or so, and then I restart the process.
Here are my questions:
1. Am I correct in believing I should stay away from temperatures exceeding 70c and Vcores exceeding 1.5v?
2. Am I correct in assuming I can knock around 7c off of any temperature readings I get from the CPU as a package? I have read online that although the CPU as a whole might be at something like 60c, the individual cores are actually running 5 to 10c cooler, and those are the specific components that can't overheat. Is this safe/validated?
3. Is this the correct order of troubleshooting for an unstable configuration? Revert last change -> Increase Vcore and retry, up to 1.5v -> if still unstable, check the RAM -> other issues.
4. Am I correct in assuming any temperature readings I get while using Prime95 are worst case and most situations, even gaming, will not cause the CPU to heat to maximum for more than a minute or some other negligible time frame?
5. Am I safe in trusting heat guarding features, or features that kill the computer if the CPU gets too hot? I've had this happen to me before and so far my CPU hasn't shown any signs of damage from that instance so I assume these features are effective.
6. If I get to a stable setting, but get into a game and the framerate stutters horribly, is it safe to assume it's likely an insufficient Vcore setting, and is it usually a lack of voltage that is the issue? Can too much voltage also cause this issue?
7. Will reducing the Vcore help reduce CPU heating?
8. In what situations would I want to consider messing with NB, SB or DRAM voltages? I know most people don't need to mess with these settings to OC, but some people have for reasons that I'm having a hard time finding online and these settings are available within AXTU. Is 1.1v for NB and SB standard, and is 1.5v for RAM standard like I am assuming? Also, when would I want to change the timings of my RAM? Right now it is at 9-9-9-24-33 @ 600mhz
9. What is the point of CPU voltage offsetting and in what situations would I want to use it? I believe it is essentially a power saving feature and I have disabled it because it was causing risky Vcore levels, but I'm still not entirely sure about that conclusion.
10. When would I want to mess with PCI frequencies? My CPU does not have integrated graphics, so adjusting APU frequencies are pointless, but PCI frequencies are available for changing in AXTU. Right now I'm keeping it at 100mhz
Thanks!
Last year marked my entrance into PC gaming, and so far I'm loving it. Right now (as I'm typing this, in fact) I am currently OC-ing my CPU. I think I know what I'm doing, but I have some questions. I'm sorry if some of my questions are answered by other threads or resources, but the issue I'm having is some guides mention some things like they are essential to consider while other guides never mention them at all, as well as just completely conflicting information, even from reliable sources.
First off, my specs:
AMD Athlon x4 750k
Sapphire Radeon HD 7870
10G of DDR3 ram in 4 sticks, 3 different brands
ASRock FM2A75 Pro-4 M motherboard
High quality 500W PSU
Air heat sinking (Looking into water/liquid cooling soon; that's my next upgrade
) As I side note, I normally run my computer with the side panel off. I know it increases dust accumulation, but I am excellent at keeping my PC components clean and dust bunny free and I always have duster on hand. For me, it dramatically reduces my CPU temp - around 10c in most cases!Also, I am a computer scientist so I am comfortable with any lingo you want to throw around.
Next, here is my OC-ing process:
Firstly, a program called ASRock eXtreme Tuner Utility came with my mobo (http://www.youtube.com/watch?v=_7naB8kIZXE). It's a windows program, like others of its kind, that allows for CPU/RAM tuning without needing a restart or interaction with BIOS. I've read up on the program and others that do a similar thing and I am very aware of certain concerns OC veterans have with regards to in-windows OC-ing. I personally am okay with BIOS and can use it if I feel like it's necessary, but I'm not looking to push more than 4.0Ghz or 4.5 at the very max.
That means I don't expect any instabilities to come out of anything except Vcore issues, so I don't need the level of control one gets from BIOS tweaking. Some specific issues I've read about AXTU are temperature readings not being accurate; I have other programs in place that have better credibility with these types of readings. I also like not having to constantly reboot, because repetitious rebooting can put stress on HDDs, especially the older ones I'm using. As for being able to turn off power saving features and other OC nuisances, all of those settings are also available within windows.
I am using CPUZ to monitor whether changes made in AXTU are actually reaching the hardware, as well as to monitor Vcore readings. AXTU also monitors Vcore, but since it has issues with temperature readings, I thought I should also keep an eye on this with another program as well. However unlike temperature readings which have conflicted, so far Vcore readings from AXTU and CPUZ have essentially been the same.
Next, I am using HWMonitor. I swear CPUID is going to own me within a month
. This is obviously being used for temperature monitoring, as well as a third reference for Vcore readings.To stress test, I am using Prime95.
I am also using task manager to keep a larger view on what's happening.
As for the actual process:
I started off at a multiplier of 36 (baseline for this CPU) with a Vcore of 1.35v (baseline for this CPU). I go into AXTU and bump up the multiplier by one and wait a minute or so as a "basic stability test." After making sure it's stable at low demand, I then pull up Prime95 and start a torture test. While testing, I watch the CPU temperature to make sure it does not exceed 70c and watch the Vcore to make sure it doesn't jump from 1.35 to 1.6. I let Prime95 go for 10 minutes or so, and then I restart the process.
Here are my questions:
1. Am I correct in believing I should stay away from temperatures exceeding 70c and Vcores exceeding 1.5v?
2. Am I correct in assuming I can knock around 7c off of any temperature readings I get from the CPU as a package? I have read online that although the CPU as a whole might be at something like 60c, the individual cores are actually running 5 to 10c cooler, and those are the specific components that can't overheat. Is this safe/validated?
3. Is this the correct order of troubleshooting for an unstable configuration? Revert last change -> Increase Vcore and retry, up to 1.5v -> if still unstable, check the RAM -> other issues.
4. Am I correct in assuming any temperature readings I get while using Prime95 are worst case and most situations, even gaming, will not cause the CPU to heat to maximum for more than a minute or some other negligible time frame?
5. Am I safe in trusting heat guarding features, or features that kill the computer if the CPU gets too hot? I've had this happen to me before and so far my CPU hasn't shown any signs of damage from that instance so I assume these features are effective.
6. If I get to a stable setting, but get into a game and the framerate stutters horribly, is it safe to assume it's likely an insufficient Vcore setting, and is it usually a lack of voltage that is the issue? Can too much voltage also cause this issue?
7. Will reducing the Vcore help reduce CPU heating?
8. In what situations would I want to consider messing with NB, SB or DRAM voltages? I know most people don't need to mess with these settings to OC, but some people have for reasons that I'm having a hard time finding online and these settings are available within AXTU. Is 1.1v for NB and SB standard, and is 1.5v for RAM standard like I am assuming? Also, when would I want to change the timings of my RAM? Right now it is at 9-9-9-24-33 @ 600mhz
9. What is the point of CPU voltage offsetting and in what situations would I want to use it? I believe it is essentially a power saving feature and I have disabled it because it was causing risky Vcore levels, but I'm still not entirely sure about that conclusion.
10. When would I want to mess with PCI frequencies? My CPU does not have integrated graphics, so adjusting APU frequencies are pointless, but PCI frequencies are available for changing in AXTU. Right now I'm keeping it at 100mhz
Thanks!
Facebook
Twitter
Instagram