FX-8350 Overclock

[DLM_012]

I'm going to run P95 on through the night but I think I have it stable at 4.47Ghz. What's the hottest that I should allow my chip to peak at? Under my current cooling setup (Coolermaster V8) and at 1.375v on my CPU I hit around 60-65C. Realistically, I know that my CPU will never be under this kind of load in day to day tasks with me doing video editing and gaming but I'm wondering if these are good temps for stress testing? I know for certain I can achieve better results with some form of water cooling (whether it be closed loop or custom) but seeing as how this CM V8 isn't even a few months old, that's out of the question for me to invest in a new cooling setup already.

Feedback anyone? Despite me having several PC builds under my belt, I am relatively new to overclocking and can use any helpful advice.

[Helltech]

Usually people say over 80c is failing, I don't like to let it go over 70c.

[DLM_012]

Thanks for the feedback. Right now it seems to alternate back and forth between 64 and 65C. I wanted to give it a few minutes to settle in to its max temp so I can sleep easy knowing that it wouldn't fry while I was sleeping lol. I'll update this in the morning with results on how it fared in P95. Once I get this OC'd my machine will pretty much be complete. I'm just about done with my custom sleeving job and I'm waiting for the rear body part of my 600T to come in the mail sometime this week. So excited! :D

EDIT: The computer just froze up on me and there's no way I'm upping the voltage past 1.4 with my cooling setup. It looks like I'll just have to stick with 4.2Ghz OC for the time being:/

[Mentalguy]

amd cpus dont like going over 70c

[ZeroPlasm]

http://www.ncix.com/products/?sku=73199&promoid=1387

http://www.ncix.com/products/?sku=78188

[cheti]

Yeap, as people have stated before - don't let it go over 70C in P95 and you'll be fine.

Even tho at AMD websites, it's stated that 61C is the max temp, but keep in mind, that during the P95 test your CPU will be immensely stressed and in day-to-day operation (even heavy games and encoding). You won't be able to stress the CPU as much.

[Andster29]

Yeap, as people have stated before - don't let it go over 70C in P95 and you'll be fine.

Even tho at AMD websites, it's stated that 61C is the max temp, but keep in mind, that during the P95 test your CPU will be immensely stressed and in day-to-day operation (even heavy games and encoding). You won't be able to stress the CPU as much.

Hint why P95 is called a stress test.

[TechFan@ic]

AMD is weird, claims 61c is the max for the 8350 yet 71 for the 4300, even though both have identical CPU cores, although the latter only has 4 rather than 8 .

Don't let it go over 80, that's when it gets really bad & you could even burn out your socket.

I'd personally keep it under 80 in stress tests and under 70 in games .

[TechFan@ic]

Found this on OCN

The max temp for the FX-8350 is 60c to the individual cores. Beyond that has the chance to permanently damage the processor.

Try using AIDA 64 for temp monitoring. Has a really nice sensor page.

You will have 2 different readings for the CPU temp (The socket or something, this reading doesn't matter). They will either say CPU/CPU Diode/CPU Core. Don't worry about the Diode or the CPU temp, just focus that your CORE temp doesn't go over 60c. I wouldn't even run it continuously at 60c. A safe constant would be 57c max... to the CORE that is. The PCB on your mobo is good to run much hotter, so don't worry about socket temp.

[MunsonedForLife]

Maybe upgrade to an all-in-one liquid cooling system? Just a thought...will probably help with the temps.

[killerx525]

The weird thing is my 1090T also has a max temp limit of 62C which is only 1C difference from the 8350. I've read for my CPU that, my CPU temp is limit is a lot higher then that, so the 8350 maybe the case too.

[Umeshwarnath]

Less than 70 should be just fine for the 8350.....if you see any throtling ...just back off the overclock by a little bit

[DLM_012]

So I pretty much have failed at OC'ing this chip you guys. I think I have bad luck with chips. I'll give a complete run down of my specs so you guys can help me resolve this:

AMD FX-8350 @ stock

Coolermaster V8

Sabertooth 990FX AM3+ board

Corsair 4 X 4GB of 1600Mhz Vengeance quad channel memory

MSI Twin Frozr 7950 OC'd

Seasonic X650 fully modular PSU

Samsung 840 series 500GB SSD

Western Digital Caviar Green 1TB

all housed in my Corsair 600T

Now I haven't taken the time to test the stability but I'm almost sure it's stable at 4.2Ghz no voltage tweaking. Once changing the multiplier or bus speed to go anything beyond that my system will fail in minutes in Prime or AMD's stress testing software. After applying 1.4v to the vcore it takes an hour or two to fail. Shouldn't I be getting better results than this at that voltage?

[TechFan@ic]

So I pretty much have failed at OC'ing this chip you guys. I think I have bad luck with chips. I'll give a complete run down of my specs so you guys can help me resolve this: AMD FX-8350 @ stock Coolermaster V8 Sabertooth 990FX AM3+ board Corsair 4 X 4GB of 1600Mhz Vengeance quad channel memory MSI Twin Frozr 7950 OC'd Seasonic X650 fully modular PSU Samsung 840 series 500GB SSD Western Digital Caviar Green 1TB all housed in my Corsair 600T Now I haven't taken the time to test the stability but I'm almost sure it's stable at 4.2Ghz no voltage tweaking. Once changing the multiplier or bus speed to go anything beyond that my system will fail in minutes in Prime or AMD's stress testing software. After applying 1.4v to the vcore it takes an hour or two to fail. Shouldn't I be getting better results than this at that voltage?

Try shutting down one module out of the 4 at a time and stress test to find out if a specific module is holding you back .

If you find a specific module that's holding you back , return the chip and get a new one.

Also check this guide : http://www.overclock.net/t/1140459/b...results-coming

[TheCanadianGamer16]

that temp limit is wierd, its the same for my 6100, yet it peaks that ON WATERCOOLING at stock in prime95, im only using a single 120mm but if push pull with cougar vortex's and a 120mm rad can't keep it cool under load, i think there is a problem, tho i have been able to keep it stable at 4Ghz, tho with prime it would kill it, but a good stability and temp monitor is OCCT, it will run continuously until either a core fails, or temps get higher than your set amount

[DLM_012]

So I pretty much have failed at OC'ing this chip you guys. I think I have bad luck with chips. I'll give a complete run down of my specs so you guys can help me resolve this: AMD FX-8350 @ stock Coolermaster V8 Sabertooth 990FX AM3+ board Corsair 4 X 4GB of 1600Mhz Vengeance quad channel memory MSI Twin Frozr 7950 OC'd Seasonic X650 fully modular PSU Samsung 840 series 500GB SSD Western Digital Caviar Green 1TB all housed in my Corsair 600T Now I haven't taken the time to test the stability but I'm almost sure it's stable at 4.2Ghz no voltage tweaking. Once changing the multiplier or bus speed to go anything beyond that my system will fail in minutes in Prime or AMD's stress testing software. After applying 1.4v to the vcore it takes an hour or two to fail. Shouldn't I be getting better results than this at that voltage?

Try shutting down one module out of the 4 at a time and stress test to find out if a specific module is holding you back . If you find a specific module that's holding you back ' date=' return the chip and get a new one. Also check this guide : http://www.overclock.net/t/1140459/b...results-coming I've tried the things mentioned in that guide but considering that applying voltage to my PC has absolutely no effect on my potential overclocking ability it won't work. I'll try your first suggestion though. If I remember correctly, the 7th module and either the 5th or 6th always give me errors first. Sometimes it's my eighth module that gives me errors.

that temp limit is wierd' date=' its the same for my 6100, yet it peaks that ON WATERCOOLING at stock in prime95, im only using a single 120mm but if push pull with cougar vortex's and a 120mm rad can't keep it cool under load, i think there is a problem, tho i have been able to keep it stable at 4Ghz, tho with prime it would kill it, but a good stability and temp monitor is OCCT, it will run continuously until either a core fails, or temps get higher than your set amount[/quote']OCCT fails too. Even though I'm not running into any current stability problems, I'm beginning to think that I should just run my chip at stock and see if it fails prime.

[DLM_012]

It's not even stable at 4.2Ghz with stock voltage. I'm beginning to regret buying this processor :/ This thing should have been binned instead of leaving the factory. It's not my RAM that's holding me back either. I checked using my 1866Mhz Patriot RAM that I was running in my computer when I had my 965BE and I'm getting the same results with errors in P95. I'll proceed to do a 24hr Prime test at stock everything to determine if I have a bad chip.