Power Consumption, Temperature, Noise
We had already measured the power consumption of the GeForce GT 240 on our older testbed, so we performed new measurements using a new testbed with the following configuration:
- Intel Core 2 Quad Q6600 CPU (3GHz, 1333 MHz FSB x 9, LGA775)
- DFI LANParty UT ICFX3200-T2R/G mainboard (ATI CrossFire Xpress 3200 chipset)
- PC2-1066 SDRAM (2x2 GB, 1066MHz)
- Enermax Liberty ELT620AWT PSU (620W)
- Microsoft Windows 7 Ultimate 7 64-bit
- CyberLink PowerDVD 9 Ultra/"Serenity" BD (1080p VC-1, 20 Mbit)
- Crysis Warhead
- OCCT Perestroika 3.1.0
The new testbed for measuring electric characteristics of graphics cards is based on the measurement tool developed by Oleg Artamonov and described in his report PC Power Consumption: How Many Watts Do We Need? The tool facilitates and automates the measurement process.
We changed the tests, too. We discarded old synthetic benchmarks from Futuremark and replaced them with real-life tasks such as playing high-definition video and a modern 3D shooter with DirectX 10 support. We also added OCCT Perestroika: GPU that can load any graphics card fully so that we could measure the power consumption under maximum load. Practice suggests that OCCT: GPU copes better with this job than the notorious OpenGL-based FurMark the graphics card developers have learned to protect their products against. We used the following test settings:
- CyberLink PowerDVD 9: FullScreen, hardware acceleration enabled
- Crysis Warhead: 1600x1200, FSAA 4x, DirectX 10/Enthusiast, "frost" map
- OCCT Perestroika GPU: 1600x1200, FullScreen, Shader Complexity 8
Except for the maximum load simulation with OCCT, we measured power consumption in each mode for 60 seconds. After OCCT: GPU had triggered the PSU’s protection a couple of times on a system with a Radeon HD 5970, we limited its run time to 10 seconds.
We got the following data using the new method:
The card behaves predictably. The only interesting thing is the behavior when playing high-definition video in PowerDVD 9: the load on the 12V line is rather high for the first 5-8 seconds, reaching over 2 amperes at peaks, but then lowers considerably. The power management system seems to realize that this task does not need lots of resources and enables power-saving features, reducing the GPU and memory clock rates.
The 3.3V line is almost unused while the maximum current on the 12V line is not higher than 3.7 amperes, which means 44.4 watts. Yes, the GeForce GT 240 is indeed an economical solution, surpassing the Radeon HD 4770. ON the other hand, the Radeon HD 5750 is almost as economical while delivering a much higher performance in 3D applications.
* - proprietary cooling system
** - passive cooling system
The cooling system installed on the Gainward GeForce GT240 1024MB GDDR5 GS is indeed more efficient but the difference is a mere 1°C under load and 2°C in idle mode. Is it quieter then?
We measured the noise with a noise-level meter Velleman DVM1326. The reference point for our noise measurement tests is 43dBA which is the level of ambient noise in our test lab as measured at a distance of 1 meter from the testbed with a passively cooled graphics card inside. When we installed the tested graphics cards, we got the following results:
So, the 80mm fan of the Gigabyte card’s cooler is not loud in any mode (its noise cannot be heard against the noise from our Enermax Galaxy DXX EGX1000EWL power supply) while the Palit GT 240 Sonic accelerates its fan in 3D mode, the noise-level meter reporting this quite clearly. The noise is audible although not really loud. Its spectrum is not irritating.
Our samples of the cards performed differently at overclocking.
The Palit GT 240 Sonic did better in terms of memory frequency, reaching 1050 (4200) MHz, but the Gainward GeForce GT240 1024MB GDDR5 GS proves its Golden Sample nature by outperforming its mate in terms of GPU frequency: 665/1619MHz is quite a serious overclocking gain and we tested the card at such frequencies. So, if you are into overclocking, you may want to prefer the Gainward version, which is also quieter in 3D mode. Now, let’s proceed to gaming tests.