Modern graphics cards from both AMD and Nvidia employ protection mechanisms to ramp up fan speeds, and eventually throttle back clock rates and voltages if they get too hot. This technology doesn't always work to keep your system stable (particularly when you're overclocking). Rather, it's meant to keep the hardware from getting damaged. So it's not unheard of for an over-tuned card to crash, requiring a reset.
There has been much debate about how hot is too hot for a GPU. However, higher temperatures, if they're tolerated by the equipment, are actually desirable as they result in better heat dissipation overall (as the difference with ambient temperature, and thus amount of heat that can be transferred, is higher). At least from a technical perspective, AMD's frustration over reactions to the Hawaii GPU's thermal ceiling is understandable. There are no long-term studies that I'm aware of speaking to the viability of given temperature set points. From my own experiences with device stability, I have to rely on manufacturer specifications.
On the other hand, it is a well-known fact that silicon transistors broadly perform better at lower temperatures. That is the main reason you see competitive overclockers using liquid nitrogen to get the chips they're testing as cold as possible. In general, lower temperatures help facilitate more overclocking headroom.

Some of the most power-hungry cards in the world are the Radeon HD 7990 (375 W TDP) and GeForce GTX 690 (300 W TDP). Both are dual-GPU cards. Single-GPU boards tend to be quite a bit lower, though the Radeon R9 290-series cards creep up closer to 300 W. In either case, that's a lot of heat to dissipate.
Volumes have been written about graphics card cooling, so we wont delve into that. Rather, we're interested in what actually happens when you begin applying load to a modern GPU.
- You launch a processing-intensive application like a 3D game or your favorite bitcoin miner
- The card's clock rates increase to their nominal/boost values; the board starts warming up due to greater current absorption
- Fan speed progressively rises, up to a point defined by firmware; usually it'll taper off when acoustics approach 50 dB(A)
- If the programmed fan speed isn't enough to keep the GPU's temperature below a certain level, clock rates scale back until the temperature falls below the set threshold
- Your card should operate stably within a relatively narrow frequency and temperature range until the application driving the load is shut down
As you can imagine, the exact thermal throttling point depends on many factors, including the specific load, the enclosure's airflow, the ambient air temperature, and even ambient air pressure. That's why cards throttle at different times, or not at all. This thermal throttling point can be used to define a reference level of performance. And if we set a card's fan speed (and thus noise level) manually, we can create a noise-dependent measurement level. What use is that? Let's find out...
- Performance That Matters: Going Beyond A Graphics Card's Lap Time
- Graphics Card Myth Busting: How We Tested
- To Enable Or Disable V-Sync: That Is The Question
- Do I Need To Worry About Input Lag?
- The Myths Surrounding Graphics Card Memory
- More Graphics Memory Measurements
- Thermal Management In A Modern Graphics Card
- Testing Performance At A Constant 40 dB(A)
- Can Overclocking Hurt Performance At 40 dB(A)?
Create a new thread in the UK Article comments forum about this subject
-
0 HideJak_Sparra , 10 February 2014 13:38Can't wait for part 2. I have a Sapphire R9 290 Tri-X and am enjoying the smoothest gameplay I've ever experienced at 1920x1200 with ultra settings in games like BF4. BUT, I'm interested in getting a 2560x180p monitor for gaming. The only thing holding me back is that I'm worried how much of a drop in FPS I will see. Hope the next article covers stuff like that. Also, as more and more gamers get more system RAM, I'd love to see an article that covers what happens when you use RAM as a RAMdisc and stick the pagefile on it. Would it act nearly as fast as VRAM?
-
0 Hiderolli59 , 10 February 2014 13:51Great article, waiting on part 2.
-
0 HideJonathan Cave , 10 February 2014 15:11Great Article.
-
0 HideJak_Sparra , 10 February 2014 15:43Can't wait for part 2. I have a Sapphire R9 290 Tri-X and am enjoying the smoothest gameplay I've ever experienced at 1920x1200 with ultra settings in games like BF4. BUT, I'm interested in getting a 2560x180p monitor for gaming. The only thing holding me back is that I'm worried how much of a drop in FPS I will see. Hope the next article covers stuff like that. Also, as more and more gamers get more system RAM, I'd love to see an article that covers what happens when you use RAM as a RAMdisc and stick the pagefile on it. Would it act nearly as fast as VRAM?
-
0 HideJonathan Cave , 10 February 2014 15:47Great Article.
-
0 Hidekyzarvs , 10 February 2014 15:54@Jak - this is covered on page 6?"What happens when graphics memory is completely consumed? The short answer is that graphics data starts getting swapped to system memory over the PCI Express bus. Practically, this means performance slows dramatically, particularly when textures are being loaded. You don't want this to happen. It'll make any game unplayable due to massive stuttering."
-
0 HideSunius , 10 February 2014 16:40Hey, about memory usage and Windows AERO: did you try benchmarking peak memory usage between various windows versions when in fullscreen mode? Going to fullscreen mode should effectively make Windows use 0 graphics memory as far as it's concerned (hence why it takes a while to switch to and out of fullscreen - it is moving data out and into video memory).
-
0 HideWossnames , 11 February 2014 14:0510 dB isn't "twice as loud". It is ten times the sound pressure. 3 dB is about the double sound pressure.However, as we humans do not perceive sound linearly, around 6 dB (actually about four times the sound pressure) is generally perceived as "twice as loud".
-
0 HideHEXiT , 12 February 2014 22:19nice... pretty much confirms what i was thinking about overclocking, the results little in the way of real performance gains.for your next foray into overclocking could you do real world cpu performance. as gamers may well be in for a shock... with very limited returns for a massive overclock, power draw and reduced cpu life.productivity on the other hand can bring real returns... so it would be nice to have this confirmed.
-
0 Hidecdrkf , 13 February 2014 11:16One myth I think you missed and really should highlight with respect to graphics memory is that as you say the amount doesn't effect performance, but the TYPE of memory really does.There are a lot of lower end cards popping up equipped with large amounts of DDR3 memory (e.g. an R9 250 with 2gb DDR3), and these are categorically a worse buy than a similarly priced card equipped with less but faster GDDR 5 memory...
-
0 Hidejabel_sk , 23 February 2014 00:42Skyrim is a not a good game to benchmark VRAM because it's a horrible port. However, the modding community has in what some would argue, redesigned the engine's memory allocation system altogether.