E
E
Evgeny Elizarov2013-01-20 19:30:21
Iron
Evgeny Elizarov, 2013-01-20 19:30:21

What's wrong with the processor?

There is a home server with the following configuration:
GigaByte GA-G33-DS3R
Intel Core 2 Quad Q8300
Patriot DDR-II DIMM 1Gb PC-6400 (x2)
Kingston HyperX DDR-II DIMM 2Gb PC-6400 (x2)
6 hard drives and a Chieftec case with a Chieftec PSU APS-650C 650W

The machine has been working for several years, but recently I have to actively convert video: Air Video Server for streaming to a tablet + I encode a lot of video using ffmpeg, and then a problem emerged - after some time of work (sometimes 10 minutes is enough, and a maximum of about 3 hours) the machine just reboots. The first thought is the temperature, but as the graphs show, the temperature does not rise above 58 degrees under load, despite the fact that the Scythe Ninja Mini is cooled (without a fan) and there is only one cooler for blowing in the case. Today I took and changed the processor multiplier from 7 to 6, as a result, at a lower frequency, it has been encoding video for about 6 hours (the temperature is only 2 degrees lower). On the motherboard, all reboots are turned off when a certain temperature is reached, everything that can only influence this is turned off. As a result - I can’t fully understand - what is the problem, I sin on a percent or bp / mother. Unfortunately, I don’t have a replacement processor (and in stores it’s no longer possible to find something equivalent for the 775th socket in stores, and the PSU is not at hand either), so it’s extremely difficult to test all this somehow . It’s just that if it’s the processor, then it turns out that you have to change the entire platform (I generally put up with this), but there are fears that this may eventually turn out to be a power supply unit and changing the system will be completely unnecessary.

Can anyone suggest something smart?

Answer the question

In order to leave comments, you need to log in

9 answer(s)
S
srka, 2013-01-21
@srka

nutrition problem. Check the capacitors on the mother. + Try with another PSU.
+ once you get there, change the thermal paste on the processor.

N
Nikolay45, 2013-01-21
@Nikolay45

Everything is correct. Quote - "When the processor clock frequency is reduced by 20%, the performance decreases by 13%, and the power consumption by 49%!" Look here - www.compress.ru/article.aspx?id=16962&iid=786 I suppose you abused passive cooling for a 95 watt processor, and the PSU fan swallows superheated air from the case. The capacities of the power supply unit or / and the CPU power inverter have practically dried up (despite the fact that they are on the MP with a solid polymer, but they are not eternal either). Leaking alkaline capacitors in the PSU are often determined visually, but the gel does not flow, you have to solder and measure it. I think that the replacement / repair of the PSU will solve your problem for six months or a year, then the MP will catch up - there are no miracles. Good luck.

V
ValdikSS, 2013-01-20
@ValdikSS

I think it's BP's fault. I had something similar.

M
m08pvv, 2013-01-20
@m08pvv

Sometimes there may still be a problem with thermal paste - has it been changed for a long time?

N
nochkin, 2013-01-20
@nochkin

There is no way to check without replacement. Unless you can try to pull out the memory modules in pairs (there are several of them), to check if it is not buggy. Although, according to the description, it looks more like a percent or a motherboard.

E
Equillibrium, 2013-01-20
@Equillibrium

1. try to increase the voltage on the core a little and put on active cooling (the temperature sensor could start to lie).
2. try lowering the voltage along with the frequency (GHz does not affect heat dissipation as much as the core voltage)
After each manipulation, run prime95 in Blend mode for 4 hours

S
space_monkey, 2013-01-20
@space_monkey

Try turning off half the hardware.
And drive, if it stops cutting down, then most likely the power supply.

V
VeMax, 2013-01-20
@VeMax

And yet I advise you to pay attention to the bridges, because. a similar behavior was observed when the temperature on one bridge was exceeded.

Q
qxfusion, 2013-01-20
@qxfusion

What video card do you have? In practice, I saw a problem that the video card driver sometimes rebooted the system due to an overheated GPU. Therefore, you first need to determine the source of the problem - software or hard.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question