The Asus GTX 960 Strix OC Edition has a 17 cm long circuit board braced by a metal backplate. Its sizable heatsink hangs over the PCB by 1.6 cm, and the plastic fan shroud extends past by that another 1.8 cm. The total weigh is 610 g, with the cooling solution accounting for 360 g on its own.  TEST METHODOLOGY Test Platform
Measurement and Analysis Tools
Testing Procedures Our test procedures are designed to determine the power consumption, noise, and heat produced by the card/cooler with the system in various states. In addition to testing under “normal” conditions, we also perform a torture test consisting of FurMark running in conjunction with Prime95 to stress both the graphics card and processor simultaneously. This combination is more demanding on the CPU and GPU than any real gaming session. This final result is not indicative of a real-world situation, but rather a worse-case scenario; If it can cool the card and its components adequately it means there will be some degree of thermal headroom when deployed in a more conventional situation. By adequately cooled, we mean cooled well enough that no misbehavior related to thermal overload is exhibited. Thermal misbehavior in a graphics card can show up in a variety of ways, including:
Any of these misbehaviors are annoying at best and dangerous at worst — dangerous to the health and lifespan of the graphics card, and sometimes to the system OS. Aftermarket coolers are installed on an ASUS GeForce GTX 680 DirectCU II OC, a factory-overclocked single GPU card that draws 225W by our estimates. The stock VRM heatsink is left on if possible. The cooler’s fan(s) is connected to the motherboard (if possible) and its speed is changed to various levels to represent a good cross-section of its airflow and noise performance. Ambient Noise Level For noise measurements, our mic is positioned at a distance of one meter from the center of the case’s left side panel at a 45 degree angle. Our test system’s CPU fan is a low speed Scythe that is set to full speed at all times while the two Fractal 140 mm case fans are connected to case’s integrated fan controller. Three standard speed settings have been established for testing. GPU Test System: Anechoic chamber measurements Case Fan Setting System SPL@1m High 24 dBA Med 15 dBA Low 12~13 dBA When testing video cards and coolers with active cooling, the low setting will be used. For passive cards and heatsinks, all three settings will be tested to determine the effect of system airflow on cooling performance. Estimating DC Power The following power efficiency figures were obtained for the Kingwin LZP-1000 used in our test system: Kingwin LZP-1000 Test Results DC Output (W) 22.2 41.9 65.5 90.7 149.0 199.6 251.2 300.3 400.9 AC Input (W) 35 56 81 105 166 211 265 322 426 Efficiency 63.6% 74.9% 80.8% 86.4% 89.8% 92.8% 92.9% 93.5% 94.1% This data is enough to give us a very good estimate of DC demand in our test system. We extrapolate the DC power output from the measured AC power input based on this data. We won’t go through the math; it’s easy enough to figure out for yourself if you really want to. TEST RESULTS Baseline Power with Integrated Graphics: Power Consumption Measurements: GPU Test System w/Intel HD 2000 IGP System Power Idle Prime95 AC 36W 71W DC (est.) 23W 56W CPU fan at 500 RPM, system fans at 580 RPM. Ambient noise level: 10~11 dBA System noise level: 12~13 dBA Ambient temperature: 21°C System with Discrete Graphics: System Measurements: GPU Test System w/Asus GTX 960 Strix State Idle Prime95 RE6 Demo (Peak) Prime95 + FurMark Prime95 + FurMark GPU Fan Speed off 1550 RPM 2000 RPM 1120 RPM* CPU Temp 17°C 35°C 35°C 46°C 46°C PCH Temp 39°C 41°C 45°C 46°C 49°C GPU Temp 36°C 36°C 62°C 62°C 75°C GPU Core Clock 135 MHz 1418 MHz 1380 MHz 1367 MHz System Power (AC) 46W 84W 191W 218W 219W System Power (DC) 32W 69W 177W 206W 207W System SPL@1m 12~13 dBA 18 dBA 20 dBA 14 dBA *manually adjusted fan speed CPU fan at 500 RPM, system fans at 580 RPM. Ambient noise level: 10~11 dBA System noise level (on int. graphics): 12~13 dBA Ambient temperature: 21°C Like the rest of the GTX 900 Strix line, the 960 model’s fans stay inactive at low temperatures so it’s effectively silent when the system is performing tasks that don’t greatly tax the GPU. As a result, when the machine was idle and when the CPU was loaded using Prime95, the test system produced the same 12~13 dBA@1m as it did when running on integrated graphics. Temperatures were low across the board and power consumption was quite modest. Despite running at default settings, the GPU core frequency jumped up to 1418 MHz during the Resident Evil 6 Demo benchmark test, a full 100 MHz higher than specified. The built-in fan control was surprisingly aggressive, keeping the GPU at 62°C with a speed of 1550 RPM. At this level, the machine emitted 18 dBA@1m. This is a fairly demanding real game test, but even at its peak, the total system power consumption was well under 200W AC. This system could be run comfortably on a 350W power supply. Our full load test of Prime95 + FurMark, required 2000 RPM to keep the GPU at the same temperature, resulting in a noise increase of 2 dB. The clock speed dropped to 1380 MHz at this point, but this is still well above stock. 20 dBA@1m is very quiet for even a mid-range graphics card, but it can easily be made quieter by overriding the fan controls. Dropping it down to 1120 RPM caused the GPU to hit 75°C and the CPU and PCH to warm up by an additional 3°C, but at this speed, it measured only 14 dBA@1m, just a hair louder than when the GPU fans were turned off altogether. Typically I try to shoot for a standard GPU temperature of 85°C but unfortunately an odd bug that caused our test system to crash the moment the GPU temperature reached 77°C with the GTX 960. However, this seems to be an isolated problem as it easily handled temperatures of 90°C and above when tested in two different systems. Altogether The Strix GTX 960 is quite promising as it clocks fairly high on its own, the power draw is modest, and the fans spin faster than they need to. There’s plenty of thermal headroom for those in warmer environments, and for overclocking. At lower speeds, the stock fans were quite smooth with a modest amount of rattling from the bearings when observed at close proximity (within one foot). Starting at about 1500 RPM it started to become increasingly whiny; to be expected for fans of this size at this speed. Even so, with the side panel of the case muffling much of the noise produced, it was fairly quiet. I did notice some coil whine during the Resident Evil 6 test (even with V-Sync enabled) but it manifested as a tolerable low pitched buzz rather than an annoying high pitched squeal. Noise & Cooling Comparison Comparison: GPU Test System (FurMark + Prime95) GPU Model Est. Power Draw (DC) GPU Temp SPL @1m Asus GTX 960 Strix 151W 75°C 14 dBA Asus GTX 960 Strix 150W 62°C 20 dBA Asus GTX 980 Strix 198W 71°C 22 dBA Zotac GTX 970 168W 78°C 23 dBA Asus GTX 680 DirectCU II OC 226W 70°C 24 dBA AMD R9 290X 308W 92°C 39 dBA CPU fan at 500 RPM, system fans at 580 RPM. Ambient noise level: 10~11 dBA System noise level (on int. graphics): 12~13 dBA Ambient temperature: 21°C We recently updated our GPU test system, transplanting it in the larger Fractal Design Define R5 to create extra space for water-cooled video cards and coolers. The new case actually allows some cards to perform better thermally/acoustically so I only have four cards to compare the GTX 960 Strix to directly, but the results are very promising. On load, the 960 runs quieter and cooler than most higher tier cards, and slowing the fan manually puts it in its own league. System Power Consumption Comparison The GTX 960 Strix has a lower TDP than its big brothers so naturally it has an energy efficiency advantage on heavy load, beating out the 970 and 980 by about 30W in the Resident Evil 6 test. It also unexpectedly bested them when idle and during video playback as well. While I wouldn’t recommend a mid-range card for a non-gaming PC, it looks like it can pull double duty without drawing too much power from the wall. GPU Power Consumption Comparison The power consumption of an add-on video card can be estimated by comparing the total system power draw with and without the card installed. Our results were derived thus: 1. Power consumption of the graphics card at idle: when Prime95 is run on a system, the video card is not stressed at all and stays idle. This is true whether the video card is integrated or an add-on PCI-E x16 device. Hence, when the power consumption of the base system under Prime95 is subtracted from the power consumption of the same test with the graphics card installed, we obtain the increase in idle power of the discrete card over the integrated graphics chip. 2. Power consumption of the graphics card under load: the power draw of the system is measured with the add-on video card, with Prime95 and FurMark running simultaneously. Then the power of the baseline system (with integrated graphics) running just Prime95 is subtracted. The difference is the load power of the add-on card. Any load on the CPU from FurMark should not skew the results, since the CPU was running at full load in both systems. Both results are scaled by the efficiency of the power supply (tested here) to obtain a final estimate of the DC power consumption. Note: the actual power of the add-on card cannot be derived using this method because the integrated graphics may draw some power even when not in use. However, the relative difference between the cards should be accurate. By my estimates, the GTX 960 Strix uses just 13W, the lowest of any mid/high-range card we’ve ever tested. The entire 2nd generation Maxwell cards do well in this metric but the 960 in particular stands out. On full load, the GTX 960 Strix uses about 150W, a considerable savings compared to most older cards. Software The aforementioned thermal headroom can be taken advantage of by the included GPU Tweak utility. Asus’ video card tuning app allows you to increase the GPU voltage by 50 mV and raise the power target by 15%, enabling to hit higher clock speeds automatically. The app has a nice charting/monitoring function and the ability to dynamically control the fan speeds as well. MP3 Sound Recordings These recordings were made with a high resolution, lab quality, digital recording system inside SPCR’s own 11 dBA ambient anechoic chamber, then converted to LAME 128kbps encoded MP3s. We’ve listened long and hard to ensure there is no audible degradation from the original WAV files to these MP3s. They represent a quick snapshot of what we heard during the review. Each recording starts with ambient noise, then 5~10 second segments of product at various states. For the most realistic results, set the volume so that the starting ambient level is just barely audible, then don’t change the volume setting again while comparing all the sound files.
FINAL THOUGHTS The consensus among other review sites pegs the GTX 960’s performance at a level similar to the GTX 760 and AMD’s R9 285/280X, making it a perfect mid-range option for those interested in a good gaming experience at 1080p resolution. If prices had stayed status quo, it would have been a coup for the GTX 960, but AMD has recently lowered their prices so their counterparts can now be found at the US$200 price-point as well. If you look at it from a purely performance:price standpoint, the new Maxwell card doesn’t have an edge. Energy efficiency and size is really where the GTX 960’s strengths truly lie, with the Strix OC Edition acting as a fine example of all that entails. The card’s power demands are impressively low, so with a decent sized cooler, you’re almost guaranteed quiet and cool operation. Under a heavy synthetic load, our sample card managed to stay relatively cool, even with reduced fan speeds that brought the noise level down to near silent levels. Along with quieter operation, gamers can expect a lower power bill, possibly less reliance on air conditioning, and the ability to push the clock speeds higher without much cost in noise or cooling, even if it’s already overclocked out of the box as is the Strix Edition. Those in the market for a new system can now look at lower wattage power supplies and smaller cases thanks to the shorter PCB. The GTX 960 isn’t quite the game changer some expected it to be, but it does bring all the advantages of Nvidia’s second generation Maxwell technology to a much larger crowd. The hardware has been significantly cut-down compared to its big brothers so it doesn’t outperform its direct competitors, but all the side benefits give it a clear advantage, making it easily the best option in its price range. |
