Calling all nerds (gaming) GET IN HERE!

[squidmaster]

My first hand experience is that they're trash. You're first hand experience is using a way over powered PSU to do a very low power job and that's working alright. That's not much of an experience.

[decay]

I wouldn't trust Raidmax PSU's, especially not cheap models with split 12V rails. I'd look into getting either a Corsair or Seasonic with single rail setup's and 40A+. Don't be cheap on something like your PSU, you want something that can adequately power your machine under heavy loads without any strain on it.

replacing his PSU at both of your advice would have been a complete waste of money-

these statements about raidmax are based on what?

my (first-hand) experience with them is they do fine, and i'm asking a lot more of my hardware than you are.

[DatUtahGuy]

Calling all nerds (gaming) GET IN HERE!

cool I will uninstall it and see how the fan is working

[jclar]

IMG

@jclar, are you saying I should "tear apart" the case of the video card and clean it?

CPU temps look great, almost definitely a problem exclusive to the video card and unrelated to the case temperature. Assuming you're on the Intel cooler you would be experiencing temps quite a bit higher than that if it weren't the case.

Yes, disassemble the cooling assembly. Remove dust.

That's what I would do. I'd have some alcohol wipes, some thermal pads, and thermal paste handly. I've seen brand new high-end video cards that people have taken apart only to discover the GPU was not even half-covered with thermal paste. There's really no risk of damaging anything as long as you're careful, and of course you're not sacrificing the warranty or anything.

Definitely. And although it will probably be covered adequately with thermal paste, almost any card out there is hindered by the fact that A) The thermal paste is of extremely low quality, as quality thermal paste is very expensive in comparison and B) The thermal paste is applied too thick with enough material for several applications. Thermal paste is a weak point in the cooling assembly and an extremely thin amount should be used.

With temperatures like yours I think it is extremely likely that either your GPU fan is toast or just completely clogged up. Reapplying thermal paste may not be necessary, but is required if the GPU is separated from the HS/F assembly.

Often times when a fan fails it will be very resistant to spinning. When flicked with your finger the fan should complete several revolutions if good. If the fan only slightly moves it is faulty. IIRC XFX has a double lifetime warranty on their cards now, I would think your 6870 would qualify for that. If the fan is bad simply RMA the card.

[E30 Wagen]

@jclar, are you saying I should "tear apart" the case of the video card and clean it?

That's what I would do. I'd have some alcohol wipes, some thermal pads, and thermal paste handly. I've seen brand new high-end video cards that people have taken apart only to discover the GPU was not even half-covered with thermal paste. There's really no risk of damaging anything as long as you're careful, and of course you're not sacrificing the warranty or anything.

[DatUtahGuy]

@jclar, are you saying I should "tear apart" the case of the video card and clean it?

[MJCRO]

I wouldn't trust Raidmax PSU's, especially not cheap models with split 12V rails. I'd look into getting either a Corsair or Seasonic with single rail setup's and 40A+. Don't be cheap on something like your PSU, you want something that can adequately power your machine under heavy loads without any strain on it.

[jclar]

Raidmax PSUs blow, regardless of whether it is working now you should swap it out for a decent unit. 530W is plenty for a single 6870, not that I'm confident that a Raidmax PSU rated at 530W can actually deliver that consistently. To give an idea, I'm using a quality Antec PSU @ 520W on a similar rig without issue, even with an OC'ed i5.

CPUID HWMonitor is an excellent temperature monitoring program from the makers of CPU-Z. Using this in conjunction with FurMark will give you good reliable results quickly. Similarly, if CPU overheating is suspected you could use Prime95 in conjunction with HWMonitor.

65*C for a 6870 is much too hot while web browsing. I'm also using an XFX 6870 with a non-reference(supposedly worse than reference) cooler. While typing this I am at 46*C. During full load I will hang right around 67*C or so. You may want to make sure that your HS/F assembly on the GFX card is clean, and you can likely improve your temperatures slightly by reapplying thermal paste using something of quality like AS5.

Temperature in the case is likely not the cause of the GPU temperatures you posted. If you post the temperature of your CPU idle and under load via Prime95/HWMonitor this will give me a little more insight as to whether or not that is the case. I will assume you are using the Intel cooler that came with the processor.

[decay]

i've been mining bitcoin with two ATI-based cards (7770 & 7950) for about half a year, which means they're at WOT all day.

PSUs get noisy when they're loaded near capacity all the time, and seeing as this rig's in the middle of my living room, i wanted it quiet.

[squidmaster]

Why? Maybe because you'd need two quad-core xeon server CPUs attached to a server mobo running 8 sticks of ram, 2 hi-perf SATA drives, dual hd 7970 GCs connected via cross-fire, 4 i pods charging, 6 120mm fans attached to a fan controller, a sound-plaster pci-e sound card, an AGEIA physX PPU, AND a corsair hydro 100 water-cooling unit in your computer to even warrant the power output of an 850w PSU.

http://www.extreme.outervision.com/PSUEngine

[decay]

my PSU is also a raidmax (850W) and it handles the load of two video cards fine. i don't think that's the root cause here.

[TimbyMaTombo]

I don't have any personal experience with them, but from what I've read, Raidmax isn't the best brand of PSUs. There is a chance that even though it's rated for 550w, it might not actually be producing it...

[decay]

i think a good case with fans is what you want- i've been mining bitcoin with two ATI-based cards (7770 & 7950) for about half a year, which means they're at WOT all day.

this case: http://www.raidmax.com/chassis/platinum.html keeps everything nice and cool; those two fans in the lower half of the front face point right at the video cards.

[squidmaster]

http://www.amazon.com/Zalman-Z-MACHINE-GV1000-Z-machine-Cooler/dp/B0012X1KSU/ref=sr_1_16?ie=UTF8&qid=1377969988&sr=8-16&keywords=zalman+vga

[DatUtahGuy]

What brand is your power supply? How many amps can the 12volt rail handle. Are there 2 12v rails? Im betting your psu is shite and can't hang under heavy loads.

Also, 144 degrees F is nothing. Its only 62 degrees C. The 68xx series has the fans almost idle at those temps on purpose to cut down on noise.

Run a logger and get to gaming. What are you max temps under real stress?

PSU is a Raidmax, 2 rails.
Here is my testing so far, took the side panel off, got a window fan and aimed it at the opened side of my tower, the card is idling at 61c right now, I played BF3 on mostly high settings earlier(NOTE THAT WHEN I SAY I PLAYED BF3, I SPENT MORE TIME ON THE LOADOUT MENU THAN PLAYING, STILL TRYING TO BIND KEYS PROPERLY AND GET USED TO CONTROLS. I WOULD GET KILLED IN 2 MIN AND SPEND 7 RECONFIGURING KEYS), I was able to play for about an hour before the game started sputtering, quit immediately and checked the temps, it was up to 94c. Definitely a cooling issue, I think the power supply is hanging in there for now.

I tried to go into BIOS for windows 8 (http://www.youtube.com/watch?v=QGiG1oljjZI), but instead of restarting into BIOS, screen stays dark and the tower beeps 5 times, nothing else happens. I was trying to change the temps at which the fans come on for my GPU.