• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Need help with FX-4170

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

CloudStrife

Registered
Joined
Apr 24, 2012
Im fairly new when it comes to overclocking but im not new when it comes to building systems, kinda. i had been outta the game for awhile and built a new system in july of last year, im pretty sure. yesterday i got this core from Newegg, upgrading from a Athlon II x4 640.

in short my issue is with P95, i started running this yesterday after getting this new core put in and cores were failing, its not just one and its not always the same.

sometimes after 1 min, and this is in a usual case, core 1 will fail, almost immediately followed by core 2 failing and then cores 3 and 4 would go on just fine.

in other cases, after fiddling around with P95 "settings", more or less, i got different results, then back to a "fresh install" of P95 i got yet again different results, and since then ive gotten different results.

different cores will fail at different times on each test
the most ive had fail on me was 3 cores and that was after approx. 8mins.

the error i was getting every time was
FATAL ERROR: Rounding was 0.5, expected less than 0.4

on every core i would get this error, regardless of in which order they failed. after doing some research alot of suggestions i was seeing was a Vcore bump. i kinda sorta did this, but only bumped it .025. everything up to this point was stock, there were not changes to FSB, HTlink, or multiplier
after bumping the voltage .025 i got this and then IMMEDIATELY stopped
then got rounding error but was stated differently, was "rounding was 0.49xxxxxxxx, expected less than 0.4" on core 1
and also got SUMOUT error on core 2

so i reset my settings back to normal, and then proceeded to go in the opposite direction, this time tho i decreased the multiplier from the stock x21, to x20 for a 4ghz clock

im currently approx 30+ mins in with no problems
current CPU Vcore is 1.39V
core temp @ 48*C
individual temps @ 58*C

sorry if this is a jumbled mess, and i know im "fresh meat" here so im just looking for some help.
i suspect already it is most likely a voltage issue?

also if any other info is needed please let me know

thanks

im running P95 v26.6, 64 bit version
OS is Windows 7 Ultimate 64 bit

computer specs are

AMD FX-4170
CORSAIR CMPSU-750TX 750W ATX12V v2.2 SLI - Power Supply
GIGABYTE GA-870A-UD3 AM3 AMD 870, rev 3.1 with bios version FEE to allow for AM3+ cores - MOBO
G.SKILL Ripjaws X Series 4GB (2 x 2GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) - ram set 1
G.SKILL Ripjaws X Series 8GB (2 x 4GB) 240-Pin DDR3 SDRAM DDR3 1333 (PC3 10666) - ram set 2
ZALMAN CNPS5X SZ 92mm Hydraulic CPU Cooler - CPU heatsink/fan
 
If it was working fine with the Athlon (to exclude any other hardware errors), and the FX-4170 is at stock, I'd say RMA it if it's not even stable at stock settings.
 
If it was working fine with the Athlon (to exclude any other hardware errors), and the FX-4170 is at stock, I'd say RMA it if it's not even stable at stock settings.

yes the athlon II was fine, i did oc the Athlon II and aside from getting a little hotter *which is expected* it was still stable after a quick P95 run, dont know long term.

but as for the FX-4170 @ stock its not stable
@ a x20 multiplier it seems just fine

/edit
also not the answer i was looking for but it may be my only course
 
If it's not stable the way it came out of the factory floor, it's defective, and there's nothing more to it. Send it back and get a replacement or a refund.
 
If it's not stable the way it came out of the factory floor, it's defective, and there's nothing more to it. Send it back and get a replacement or a refund.

sigh, yeah, but ill give it one more shot at stock, cant hurt
 
When you say individual temps in your original post, you mean your individual core temps are at 58C? If so, that's too high! 55C is the maximum safe temperature for the FX chips. That could have something to do with it being defective as well.
 
When you say individual temps in your original post, you mean your individual core temps are at 58C? If so, that's too high! 55C is the maximum safe temperature for the FX chips. That could have something to do with it being defective as well.

thats the temp its at when it was running @ x20, and was running for nearly 1-1.5 hours

/edit
heres it running at stock multiplier

/edit2
if a core "fails"
ill post a attachment of that as well

/edit3
core 1 "failed"

first one is around start
second one is almost immediately after core 1 failed
 

Attachments

  • cpu issue 2.png
    cpu issue 2.png
    188.3 KB · Views: 985
  • cpu issue 3.png
    cpu issue 3.png
    187.3 KB · Views: 953
Last edited:
Either those temps are too high, or you have a bad temp sensor. Maybe it's unstable because of temps. Also, what is the stock voltage for the 4170? Because I thought it was 1.25V, not 1.4V. That would cause some pretty toasty temps alone.
 
Either those temps are too high, or you have a bad temp sensor. Maybe it's unstable because of temps. Also, what is the stock voltage for the 4170? Because I thought it was 1.25V, not 1.4V. That would cause some pretty toasty temps alone.

and thats one thing thats confusing me
alot of what ive seen on people ocing the 4170 is that they are bumping the voltage UP TO 1.38+

yet my Bios says the "normal" voltage for my CPU is 1.4125
 
That is very strange. I made my 4100 into a 4170 and I'm only using 1.25V. :D My temps don't go above 30C on the Hyper 212+ even after 8 hours of p95. I have no idea how your Zalman cooler performs, but I would hope it's better than the stock cooler. Was that the same one you had on your Athlon?
 
That is very strange. I made my 4100 into a 4170 and I'm only using 1.25V. :D My temps don't go above 30C on the Hyper 212+ even after 8 hours of p95. I have no idea how your Zalman cooler performs, but I would hope it's better than the stock cooler. Was that the same one you had on your Athlon?

yes, and my athlon was always very cool. but it was also weaker, more or less, and didnt require as much "draw" i guess u could say. i cant remember the exacts but i do know that the core temps in Aida64 were always lower the CPU temp, and the CPU temp even after being oc'd to about 3.2-3.3 never went above 50*C, so i wanna say the core temps were somewhere around 43-48*C with a full load
idle i wanna say they were somewhere around 20-25*C

when this core is idle, it says the core temps, this is with the reduced voltage and clock speed *i wanna say like .9-1.0V and like 1.5ghz clock* the core temps are between 7*C and 16*C

but once it starts passing the boundries of the 33% usage, the temp goes up rather quickly

the other 3 cores right now are still running just fine on P95, so is it possible its just a cooling issue?
or is there possible a way for me to lower the voltage?
your 4100 @ 4.2ghz, is basically the same core from what ive read, so i shouldnt need this voltage anyway should i?

sorry prolly another jumbled mess but bear with me im starting to get tired :p
 
ywhen this core is idle, it says the core temps, this is with the reduced voltage and clock speed *i wanna say like .9-1.0V and like 1.5ghz clock* the core temps are between 7*C and 16*C

That's about what mine does at idle. It goes to 1.4GHz and 0.95V, and has impossibly low temps, but that's the same for all the FX cpus that I've seen. It's impossible for the core temps to be that low unless you're sitting on your computer in a freezer. :p I think they're supposed to be accurate at load, which is when it matters.

the other 3 cores right now are still running just fine on P95, so is it possible its just a cooling issue?
or is there possible a way for me to lower the voltage?
your 4100 @ 4.2ghz, is basically the same core from what ive read, so i shouldnt need this voltage anyway should i?

It might very well be a cooling issue now that we get down to it. Like I said, 55C is the maximum safe temperature, any higher than that and it will likely affect CPU performance and cause instability. Also, there is no difference between the FX-4100 and FX-4170 at all other than a higher multiplier out-of-box on the 4170, so that's why I said I've made a 4170 out of mine by overclocking it. Out of curiousity, drop your Vcore down to 1.25 and your multiplier down to 18 (FX-4100 stock settings) and see what your temps are like at load, and if they're good, then try bringing the multiplier back up without touching the voltage.
 
That's about what mine does at idle. It goes to 1.4GHz and 0.95V, and has impossibly low temps, but that's the same for all the FX cpus that I've seen. It's impossible for the core temps to be that low unless you're sitting on your computer in a freezer. :p I think they're supposed to be accurate at load, which is when it matters.



It might very well be a cooling issue now that we get down to it. Like I said, 55C is the maximum safe temperature, any higher than that and it will likely affect CPU performance and cause instability. Also, there is no difference between the FX-4100 and FX-4170 at all other than a higher multiplier out-of-box on the 4170, so that's why I said I've made a 4170 out of mine by overclocking it. Out of curiousity, drop your Vcore down to 1.25 and your multiplier down to 18 (FX-4100 stock settings) and see what your temps are like at load, and if they're good, then try bringing the multiplier back up without touching the voltage.

ok ill give that a shot
 
heres about a 10 min run
*bios says 1.2625v, -1.5v from "normal" 1.4125* 1.25v with 18 multiplier

/edit
and a 20 min run
going to go ahead and push the multiplier back up to 21

1st is 10 mins
2nd is 20 mins
 

Attachments

  • cpu issue 4.png
    cpu issue 4.png
    192 KB · Views: 933
  • cpu issue 5.png
    cpu issue 5.png
    189 KB · Views: 925
Last edited:
Those temps look much better, like I said, just go up a little with your multiplier and test for stability and temps and see if you can make it back to 4.2GHz on 1.25V, except have a stable and much cooler running CPU. :D
 
alright, heres what i have after about 15 mins, and then after core 1 finishes test 10 for the 720k

after that, my system froze on me and was completely unresponsive, i did not get a blue screen tho
 

Attachments

  • cpu fixed....png
    cpu fixed....png
    194.7 KB · Views: 939
  • sigh.png
    sigh.png
    186 KB · Views: 910
alright, heres what i have after about 15 mins, and then after core 1 finishes test 10 for the 720k

after that, my system froze on me and was completely unresponsive, i did not get a blue screen tho

Looks like core #1 failed. Why's your FSB at 201MHz, might I ask? Why not try changing that back to 200 and let's see what happens. Everything else looks good, temperature and voltage wise.
 
Looks like core #1 failed. Why's your FSB at 201MHz, might I ask? Why not try changing that back to 200 and let's see what happens. Everything else looks good, temperature and voltage wise.

bios says its at 200
but once i get in windows it puts it @ 201
was doing that with my athlon as well
 
Back