Is it normal value ? Psu problem?

[dazdazdaz268]

Hi , I did a few test at idle and full load and I see some stranges values (temps and +12 voltage)
idle : https://prnt.sc/h664io
full load:http://prntscr.com/h664u0

Cooler: stock fan
GPU: MSI gtx 970 4gb
motherboard: gigabyte P55A-UD3
PSU: Antec 500 W bronze certified (don't know the name exactly I can give it to you later)

So may i have your thought about this ? thank for your replies !

 

[Dark Lord of Tech]

Use HWINFO to check , it's a lot better.
https://www.hwinfo.com/download.php

 

[Roland Of Gilead]

Your idle/load temps for the I7 870 are a little high. Ideally they should be about mid 30's idle and 70 load. The chip is older, so if you haven't replaced your thermal paste for more than a couple of years, i'd suggest doing so. Something like this would be fine : http://www.arcticsilver.com/as5.htm and will most likely bring your temps back at least 5c or maybe up to 8. You could also do some cleaning in your case, clear out fans etc with some compressed air, and make sure the fan is spinning okay (although it seems fine from the sensors in HWMon)

The temps aren't exactly dangerously high or anything, so don't be too concerned just yet. BTW way are you OC'ing your CPU or GPU? That would add to case temps and contribute to higher idle/load temps.

 

[Dark Lord of Tech]

HWmonitor doesn't real PSU voltages well , it's poor software use HWINFO.

 

[10tacle]

^^Yep. HWmonitor is worthless. Their software is not properly converting hardware monitor readings for whatever reason. I first noticed discrepancies with its readings when I started using OCCT years ago and started comparing readout numbers during tests. The numbers between the two programs under separate testing variated wildly, and I got a third opinion from HWiNFO64 - it matched OCCT's readings.

 

[Roland Of Gilead]

^^Yep. HWmonitor is worthless. Their software is not properly converting hardware monitor readings for whatever reason. I first noticed discrepancies with its readings when I started using OCCT years ago and started comparing readout numbers during tests. The numbers between the two programs under separate testing variated wildly, and I got a third opinion from HWiNFO64 - it matched OCCT's readings.

Argue till your hearts content about whether or not HWmon reports things as you would like. It's neither hear nor there.
My experience with both HWmon and HWinfo are identical (i.e same results). Regardless of the software used to monitor the system, I think your missing the point. Chances are the OP's CPU needs some thermal paste replacement and/or a general clean out of fans/heatsinks. I would expect the temps to drop back to right where they should be. Maybe an additional case fan might help.

 

[10tacle]

You do realize it's more than just temp, but voltage reading he's concerned about, right? As a long time aggressive overclocker, it's not about what I "like" - it's what I need: accurate reporting data. From what I understand, it's an issue between HWmonitor and the chipset used. Some report no problems, others report wild fluctuations. This is both between AMD and Intel chipsets, so the issue is widespread.

Nobody has ever been able to pin down a root cause, but likely it's that CPUID's HWM just doesn't like some vendor motherboards performance sensors. All I can tell you is that my builds, which are either with ASUS or MSI motherboards, have issues with accurate HWM reporting. That's not my opinion. That's a fact. This is why I always recommend getting a second and even better a third opinion for a conclusive consensus before looking into an actual hardware problem and starting the tear down for trouble causing (out of norm temp and voltage readings specifically).

 

[Dark Lord of Tech]

HWINFO is much better than HWmonitor.

 

[Roland Of Gilead]

You do realize it's more than just temp, but voltage reading he's concerned about, right? As a long time aggressive overclocker, it's not about what I "like" - it's what I need: accurate reporting data. From what I understand, it's an issue between HWmonitor and the chipset used. Some report no problems, others report wild fluctuations. This is both between AMD and Intel chipsets, so the issue is widespread.

Nobody has ever been able to pin down a root cause, but likely it's that CPUID's HWM just doesn't like some vendor motherboards performance sensors. All I can tell you that my builds, which are either with ASUS or MSI motherboards, have issues with accurate HWM reporting. That's not my opinion. That's a fact. This is why I always recommend getting a second and even better a third opinion for a conclusive consensus.

I will accept your experience of it (HWMon). That's how you view it. And as much as you point to people having issues, there are equally others, like me, who have no issues, so it's not really fact then is it!? I have both monitoring apps. I use them frequently along with others, as I'm always trying to push my system. BOTH report the same results on my system. There maybe different nomenclature, but the results, broadly speaking will be roughly the same, and that applies to voltage readings too, I'm happy to post a screenshot illustrating that. Now, it's very much a subjective thing. So I respect your opinion, but to say HWMon is worthless, is a bit of a reach, don't you think? It works perfectly well, and I've an ASUS mobo, it works just fine, as does the reporting between both HWMon/info. Same results.

 

[Dark Lord of Tech]

HWINFO is updated more often to adapt to new hardware releases, therefore it tends to give better results , the developers always have a BETA loaded behind new releases.

 

[Roland Of Gilead]

HWINFO is much better than HWmonitor.

HWINFO is updated more often to adapt to new hardware releases, therefore it tends to give better results , the developers always have a BETA loaded behind new releases.

Well, that doesn't make much sense now does it!

here : https://www.cpuid.com/softwares/hwmonitor.html you can see the latest edition of HWmon for Coffee Lake, SL- x etc - what makes you believe it's not updated regularly enough to use the new hardware releases?

Or are you just arguing for arguments sake?

Anyway, I think the OP has enough to go on right now with some simple steps to hopefully fix his issue.

 

[10tacle]

LOL I can't believe we are having a friendly beer table argument over HWmonitor's performance monitoring. For many, it's a fail. For others, it's not. For me, it's worthless. So yes, I will concede that I should have included the disclaimer "from my experience" as to being worthless. I figure it's about a 50/50 chance of getting an accurate reading across different build combinations, chipset generations, etc.

Bottom line: make sure the software that is monitoring the hardware is giving accurate readings. As I stated, trusting one software utility, any utility, is not enough before spending more time digging into physical hardware problems (and never run two monitoring programs at the same time). But yes, hopefully he has enough info to further research what his issue may (or may not) be.

 

[Roland Of Gilead]

LOL I can't believe we are having a friendly beer table argument over HWmonitor's performance monitoring. For many, it's a fail. For others, it's not. For me, it's worthless. So yes, I will concede that I should have included the disclaimer "from my experience" as to being worthless. I figure it's about a 50/50 chance of getting an accurate reading across different build combinations, chipset generations, etc.

Bottom line: make sure the software that is monitoring the hardware is giving accurate readings. As I stated, trusting one software utility, any utility, is not enough before spending more time digging into physical hardware problems (and never run two monitoring programs at the same time). But yes, hopefully he has enough info to further research what his issue may (or may not) be.

i'd wholeheartedly agree with that have a good one

 

[dazdazdaz268]

Hi , First of all thanks for all your replies ! It's a "budget" gaming computer , so it has been cleaned and the thermal paste has been changed just before the test (maybe I didn't use enough of thermal paste).The main issue I was concerned about was the voltage that seem to be particulary high : I used few monitoring programs : OCCT , speedfan and hwmonitor , the result was always the same : 16V maximal and 15 V average.
Test with HWiNFO after 20 minutes of full load : temps : https://prnt.sc/h6ak36
voltage: https://prnt.sc/h6akhr

So do you think this is the correct values ?

 

[Roland Of Gilead]

can you try a different PSU to see if the voltages are similar? It may be the overvoltage protection not working as it should. Maybe SR-71 can advise as he is the PSU Guru.


The temps on full load at 74 or thereabouts don't seem that high, I wouldn't want them much more than that. You could get a new cooler to replace the stock one, and reduce those temps a whole lot.