Jump to content


  • Content Count

  • Joined

  • Last visited

Posts posted by rustynator

  1. Those temps seem fine...and the idle temp seems very good. In response to DragonTattooz's statement, yes it is easily possible to get 28 idle with air, as long as your ambient temps are around 23C (normal sort of temps) and you have a decent air cooler, good air flow and nice neat cabling.


    My old 3700 used to do 26-28C idle on air cooling, but my new 3700 is a different kettle of fish...does 31-33 idle with stock volts!

  2. Thanks Michael :) Here is a little Prime I have been working on at my stock volts and 3.11Ghz (notice the better temps this time) :




    For some reason the temps on this CPU arent very good. At idle I get around 30C at stock volts, hence my load temps are around 40-42C. This gets far worse when I increase the Vcore, and hence the chip is not scaling well with voltage at the moment (was prime stable at 3.145Ghz at 1.44V).


    The strange this is that my old 3700 had great temps at 1.54V...idle was 28C and load was 40C. I cannot understand why the new chip is running so hot. I have reseated my TT BT loads of times with varying amounts of AS5 and still no luck. If I was brave I would remove the IHS, but I cant imagine knackering this CPU, so I have wussed out (for now!).

  3. I think you might be in luck mate. I have a KACAE 0602GPMW and it does 3.11Ghz at stock volts (1.392V) on air cooling, and this is totally prime stable. Here is a pic I took of my OC last week (my temps have dropped since then so I have managed to go up to 1.392V on the same Vcore):




    Your CPU is a week older than mine, so it should do pretty well on stock volts (maybe 3.0GHz depending on the cooling)

  4. Cjoe when you talk about voltage variations, what sort of fluctuations are we talking? My Vcore, when measure with a multimetre, is 1.51-1.52 at idle and 1.54 on full load. Is this the sort of thing you are talking about? Also, I take it you mean the rail voltages as well. I thought that variance in voltages was down to the voltage regulator of the motherboard as well as the quality/power of your PSU?


    There has been a flurry of voting during the last few hours, with the categories of 1.575-1.60 and 1.60-1.625 taking the lead. Thanks for all the votes/help, keep them coming.

  5. I have heard that figure of 10-15% before (in custom Pc magazine if I remember rightly). So that would mean for my 3700 sandy (stock Vcore of 1.40V) the safe voltage might be between 1.54V-1.61V. This seems to kind of agree with the poll so far, however the voters are at the top end of this margin with the most popular categories being 1.6-1.65V.

  6. I think a decent, "safe" lifespan for a CPU would be 3 years, however that does not necessarily mean that the CPU has to be you main CPU for 3 years. Although I may change my CPU in 2 years or less, I would still like to be able to sell the CPU or keep it for a bit longer.


    As for usage, I am talking about day to day use, such as surfing the web, along with things like gaming.


    cjoe7 I didnt realise how important a steady source of electricity was. By electricity, do you mean voltage, and hence that it is important for your Vcore or Vdimm not to vary erratically?


    The poll is showing some interesting results at the moment, with 1.625-1.650 topping the charts at this very second in time. This is higher than I was expecting. Keep the votes coming thick and fast.....

  7. I have been trying to find out what the max safe CPU voltage is and have had a lot if mixed responses...on overclock.net the general opinion seems to be between 1.575-1.65V. What do the people of DFI street think?


    Basically, what is the maximum safe Vcore for an AMD CPU (such as a AMD 64 3200-4000, and X2 or an opteron) if temps were not an issue and were kept under the 45-50C safety barrier by a good air/water cooling setup (not an extreme setup like phase change)???

  8. I tried doing the reading using a molex as my grounding source. This gave on average a reading of 0.02V higher than the reading taken using the proper method. This is due to what supershanks mentioned....here is a bit of theory


    Using the formula V=IxR, if the current (I) is kept constant by the power supply in the computer, and the resistance ® increases due to there being more wires and circuitry between the Vcore measuring point and the grounding point, the IxR will be a bigger value therefore V will be larger.

  9. Neither the bios or windows (i.e. CPUz) is accurate for measuring Vcore. When I tested the Vcore with a multimeter I found that the actual reading was about 0.05V higher than CPUz and 0.03V higher than the Bios, so at 1.66V you might actually be running at 1.70V!!!

  10. Right...I have taken the readings by the proper way (not the molex way) and they show that you CANNOT trust windows in the slightest. I set the Vcore at 1.525 in the bios. In windows it shows up as 1.52V, in the bios it shows as 1.54V, but the voltmeter reads 1.55V when idle and 1.57V when on full load. THat is a difference if up to .05V!


    Now for the Vdimm. In the bios I set the Vdimm at 2.70V. In windows this value reads 2.7V, in bios it reads 2.74V but on the voltmeter it reads 2.77V at both idle and load. A difference of 0.07V!


    This just proves how unreliable both the bios and windows are at reading voltages.

  • Create New...