You probably see this a lot, but this is my first time posting anything here. So forgive me if this is in the wrong spot, and be gentle.

Anyway, I just recently bought and built a new computer. As someone who has up until now been almost exclusively on the console side of the wall, most of this stuff is new to me. Here are the specs for the components I bought:

CPU: AMD FX-8120 Zambezi 3.1GHz

Motherboard: GIGABYTE GA-970A-D3

RAM: G.SKILL Ripjaws X Series 8GB DDR3

GPU: GIGABYTE Radeon HD 7850 2GB

PSU: Diablotek UL Series 550W

So, now that that's out of the way, here's my problem. I was playing Tomb Raider with the settings maxed and it was looking fantastic. Really smooth frame rate, textures were all nice, etc. Then, about an hour in, the whole thing fell apart. I lost the video signal to the monitor and I heard a clipping noise in my headphones as if the last fraction of a second of the sound was looping over and over. I had to do a hard reboot to get everything going again. But then the weird part was, everything just fired right back up. I was able to jump right back in the game, and played for another hour or so, and the same thing happened. There was only cool air being blown out of all the box's exhaust ports (including the one on the video card itself) but I know that probably doesn't mean a whole lot.

I went looking around online and found out that the card is overclocked from the factory and that sometimes that causes problems. So I used the software that came with it and turned it back down. Still had the problem. Updated drivers, checked that the card and motherboard both had the most recent BIOS files (they do) and the problem persists. Then I had the notion that maybe it's not getting enough power, so I went looking at that. The video card calls for a minimum of a 500W psu. I have a 575W. But then the CPU saws that it has a Thermal Design Power of 125W. Reading up on that leads me to believe that this means it will use a max of 125W while it's being taxed heavily to run the fans enough to cool it. Basic math tells me that if the CPU and GPU are both using all the power they will use, that's 625W and more than the 575W my PSU is willing to offer.

So, my question for you fine folk is this: Does this sound like a different, more powerful PSU is the answer? Or is there something wrong with the video card? I don't really want to spend more money on a different PSU if I don't need it, and while I know a little bit about this stuff, I'm not confident enough to diagnose a problem without input.

The TDP on a 7850 is more like 130W or less. The 500W is a general recommendation for a complete system. Putting your system through a power supply calculator, I wouldn't expect you'd have a problem with a good 500W+ power supply.

I'd first take a look at temperatures under load, especially for CPU and GPU. HWMonitor is one option. And make sure all your fans are in fact spinning while the system is on. Open the side panel and/or take a look at the RPM readings in HWMonitor or Speedfan. (Some will show zero, if you're not using every single fan connection on the board.) If the fans are running normally and CPU temp is sky high, it could be your heatsink isn't seated properly.

Also try other games to see if you get similar results. It could be a game/driver issue specific to that game.

A PSU is split into rails, each rail can supply a certain amount of power before it starts to overload. If you have a PSU with multiple rails you have to be mindful of which connectors you are using as you can overload a single rail, which could be causing the issues you have here. Its hard to see if a PSU has rails without physically seeing the information sticker on the side of it.

Your first bet is to check system temperature, then make sure every component in the system is connecting properly (this can cause wierd issues.)

PSU's are most efficent when they are working at around half load or a bit higher, this keeps their heat down (as less power voltage needs to be turned into heat) and can save some in power over a longer period. If you have the cash it wont hurt jumping up to a 750 however it may not fix these issues if it comes down to something like heat.

I didn't mention it in the first post, but it actually did this the first time while playing Civilization 5, and I thought it was that at first. But then Tomb Raider did it again. If it was temperature, wouldn't it take some time for it to come back up?

you're smarter than most, figuring out what the tdp of the cpu was all by yourself.

as wraithtek said though, the 500 watt value that they are talking about regarding the graphics card is a recommendation for how powerful the entire psu should be to power a typical system with a 7850 in it, and it is also very exaggerated to take into account all the crappy psu's that are out there.

We should probably figure out if your psu is.

i couldnt find an ul series 550 watter on diablotek's website. ( never heard of that brand before by the way )i did find another 550 watter on the site though. and looking at it's specs it delivers 432 watts on the rails that power the graphics card and the cpu.

which should be fine for your components, as long as you dont up the cpu or gpu voltages while overclocking.

components cool extremely quickly as soon as the demanding task stops, so no it wouldn't need time to recover after one of these incidents.

download hwmonitor and run a 15 min tomb raider session then jump out and tell us what the max recorded values were, for the cpu and gpu.

Please clarify.

Also have you plugged the 6 pin pci-e connector into the side of he graphics card.

But to be honest it should be running the PSU seems powerful enough never heard of diablotek thou.

If i were to guess it would be either be overheating or a power draw issues.

Power Supply. Sorry about the confusion. The listing was entered wrong on newegg before, it seems. It's a 575W power supply, and the rest of the specs are in the link there.

So anyway, I did a little bit of playing around with stuff last night. First, it was mentioned up there that I make sure everything is plugged in right, so I went ahead and did that. Pulled the vid card all the way out, unplugged everything, then plugged it all back in. I also used a different 6-pin connector, just in case there was something wrong with the one it was using before. Made sure everything was plugged all the way in and snug. Also made sure the wires were tucked away from the card's fans and such.

The video card came with its own monitoring software, which is what I used to turn the overclock down originally. I went looking into the fan speed after you guys mentioned overheating, and it turns out it was set to "auto" mode and stayed around 50% speed until the card hit 60°C. So I set it to manual and told it to run at 100% all the time. I think that may have done the trick. It ran for almost an hour without any issues, so I got brave and turned the overclocking back up to what the factory had it set at. I then played for 2 more hours with that software running to monitor temperature, and I didn't see it go above 50°C the whole time (I have it running on a second monitor to watch while I play).

For the sake of being thorough, I did the 15 minute test with HWMonitor that was suggested just now. As for the max recorded values, the CPU's max was 46°C (114°F), and the GPU's max was 45°C (113°F). I'm honestly not sure if that's good or bad, since this is the first time I've ever monitored temperatures of computer components. The most I've done before this is put my hand over my Xbox 360's exhaust port and said "Damn, that's hot."

But like I said, it was running pretty well last night. Based on what you guys said, I feel like it was overheating and maybe turning the fan all the way up helped? I appreciate all the help you guys offered, and I appreciate more that I didn't get any kind of "lol newb" vibes. If it makes sense to you guys that turning the fan up would fix the problem, let me know. Thanks again :)

Did you play something during these 15 minutes?Cause 45 c is cool as hell.

When it comes to graphics cards, they can run around 90 c without any issues.
Most cards have a threshold at about 105c, but you should keep em under 90.
So it really shouldn't have mattered if your graphics card ran at 60c before you increased the fanspeed.
Intel cpu's prefer it under 80c.

Regarding putting your hand on something,We humans get scalded if something around 50c or hotter, we might think that the computer is about to melt, but it is far from the operational thresholds of any pc component, cept for psus and harddrives.

I was playing Tomb Raider, and actually turned everything up to ultra just to be sure. The card has two big-ass fans on it so I'd hope it would stay cool. I know it was getting hotter than that before because the fans were making as much noise when it messed up before as they are now. I guess running them that fast from the start makes it easier to keep the temp in control. I played it more after that test for about an hour and the temp only went to 49°C. Tomorrow night, since I don't have to wake up early for work, I'm gonna try to do an extended run just to make sure it's actually working right.

So, the video card malfunctioned again while the temperature monitor had it never going above 48°C. I've decided to call it quits on that card and sent it back to Newegg. Gonna order a new one. Thanks again for all the help, guys.