• Home
  • blog
  • SPCR’s 2012 Graphics Card/Cooler Test System

SPCR’s 2012 Graphics Card/Cooler Test System

Our updated 2012 Graphics card/cooler test system utilizes a Sandy Bridge CPU and Gigabyte’s versatile Z77X-UD3H motherboard, a modified Define R3 tower case, a Kingwin 1000W 80 Plus Platinum power supply, and cooling products from Thermalright, Scythe, and Antec.

SPCR’s 2012 Graphics Card/Cooler Test System

September 10, 2012 by Lawrence Lee

Our old video card/heatsink test platform consists of an AMD Athlon II X3 processor on a Gigabyte 785G motherboard mounted in an original Antec P180 case. From a purely utilitarian standpoint, it doesn’t have any serious issues but its age combined with a few case nitpicks makes it a prime candidate for an upgrade. The platform was never particularly energy efficient to begin with and the P180 was overly large and heavy for our purposes. Wear and tear had damaged the side panels too, making them difficult to mount and no longer flush against the rest of the chassis.

Our new test system, fundamentally, is similar to its predecessor. It offers us plenty of room to work inside along with ample clearance for longer graphics cards. The environment is very stressful due to a low airflow, low noise configuration which helps us differentiate the acoustic and thermal differences between the various video card and heatsink models. The ability to quickly increase fan speeds is also necessary as many fanless coolers aren’t equipped to survive such a torture test.

Motherboard: Gigabyte Z77X-UD3H

The motherboard is arguably the most important component in any PC build as it defines the limitations of most of the other components and peripherals. An Intel board is an obvious choice for hardcore gamers as their high-end processors deliver superb performance and we appreciate their excellent energy efficiency compared to AMD’s offerings. Unless you have the deep pockets for a LGA2011 Sandy Bridge Extreme system, the feature-rich Z77 chipset for the LGA1155 Sandy Bridge/Ivy Bridge platform is ideal.

Gigabyte’s Z77X-UD3H is an incredibly versatile ATX Z77 board. For high-end gamers, there are three PCI-E 16x slots for using multiple video cards in CrossFireX or SLI. Like most Z77 motherboard, it has excellent overclocking features which can help alleviate any bottlenecking issues with these high-end multi-GPU setups. Integrated graphics is available for use with Lucid’s Virtu GPU virtualization software but it’s also nice to simply have backup available in case a discrete graphics card fails.

The board has an mSATA slot under the CPU socket, presumably to use a small drive to take advantage of Intel’s SSD caching feature. It’s a nice add-on if you’d rather not spare one of the SATA ports for a standard SATA SSD. The board is also equipped with an attractive physical start button and a POST LED indicator which can come in handy when assembling and testing a new system. PWM fan control is offered on all four of the board’s fan headers though only two of them work with DC (3-pin) control.

We appreciate the layout of the board as well, particularly the placement of a PCI-E 1x slot above the first PCI-E 16x slot. This gives some clearance between the top of the card and the CPU/cooler. Some third party GPU heatsinks have large screws on the top which may interfere with larger CPU coolers. It’s also notable that all the cable headers are located at the bottom edge and the SATA ports are laid on their sides, so they won’t get in the way of any large, dual or triple slot video cards/coolers.

CPU: Intel Core i3-2100

The Core i3-2100 is quite popular amongst budget gamers who prefer Intel hardware but can’t afford the ~US$200 for a proper quad core Core i5 or i7. This dual core CPU with Hyper-threading runs at 3.1 GHz and actually keeps up with most games fairly well. For our purposes, we don’t need a particularly high-end chip and the i3-2100’s lower power requirements means we don’t need to run the fan that’s cooling it at loud levels.

RAM: Kingston HyperX Genesis 8GB DDR3-1600

Memory isn’t a critical component for our system but with RAM being so cheap these days, 8GB should be standard in just about any build even though 4GB is often enough, even for a gaming PC as most titles are still 32-bit applications that don’t use more than 2GB in Windows. We’ll be using a 2 x 4GB kit of HyperX Genesis DDR3-1600, high performance DIMMs from Kingston. They’re equipped with low profile heatspreaders which we recommend to avoid interference with third party CPU coolers.

Power Supply: Kingwin Lazer Platinum 1000W

1K watt power supplies are still overkill for any single GPU system but the Kingwin LZP-1000 gives us plenty of headroom for future high-power cards and most importantly, it’s incredibly quiet so it will have a minimal impact on the system’s overall noise level. On our PSU test platform, the fan stayed at its minimum speed with a load of up to 400W DC, producing only 17 dBA@1m in the process. The modular cables are also much appreciated.

Case: Fractal Design Define R3 (modified)

The Define R3 is one of the most popular quiet cases on the market thanks in part to its minimalist exterior, user-friendly interior, included fan controller, and the presence of acoustic dampening foam. It doesn’t meet all our criteria however, particularly with regard to its non-removable hard drive cage which limits the total video card length. We ended up ripping out the drive cage entirely, rivets and all as we didn’t require any 3.5 inch drives anyway and it makes for better system airflow. We also replaced the stock fans as acoustically, they weren’t good enough for a reference system.

Storage: Corsair Force GT 120GB

An SSD is a good investment whether you’re playing games or just testing hardware. Applications load up lickety-split and they are completely silent. The Corsair Force GT is a popular model, one of the many SandForce SF-2281 based solid state drives on the market and one step down from the Corsair Force GS. This particular drive is a manufacturer refurbished model that can often be found at a considerable discount.

CPU Cooler: Thermalright HR-02 Macho (prototype) & Scythe Slip Stream (500 RPM)

The Thermalright HR-02 Macho is one of the best heatsinks available and we were fortunate enough to receive a prototype sample that never made it to market. Its performance isn’t nearly as good as the final product but it’s good enough to cool our Core i3 chip with only a 500 RPM fan, a Scythe Slip Stream. Even at full speed, the SY1225SL12SL is one of the quietest and smoothest sounding case fans available, so no control is necessary.

System Fans: 2 x Antec TrueQuiet 120

For our build we wanted to hit a very low noise level while also having the option of increasing system airflow for fanless graphics card/coolers. Antec’s TrueQuiet 120‘s aren’t the best performers but they have truly excellent sound quality and padded corners that inhibit vibration. They have a decent range as well, almost inaudible in our case at ~600 RPM and fairly loud at the top speed of ~1130 RPM. As the Define R3’s intake fan mount uses a clip system designed square frames, we had to add some adhesive foam to the sides to create a snug fit. The two fans were connected to the motherboard and controlled via SpeedFan.

2012 GPU Test System Assembled

Our finished platform is a fine example of a quiet PC with good airflow and plenty of room for long graphics cards, even dual GPU versions. Airflow is mostly unimpeded, running front to back.

The Define R3 doesn’t have tons of room behind the motherboard tray but it is sufficient as long as the thicker cables aren’t tied together.

Power Consumption Comparison

System Measurements: Old GPU Test System
(AMD Radeon HD 4200 IGP)
Measurement
Idle
CPU Load
System Power (AC)
52W
140W
System Power (DC)
unknown
111W
System noise level: 13 dBA@1m
System Measurements: New GPU Test System
(Intel HD 2000 IGP)
Measurement
Idle
CPU Load
System Power (AC)
36W
74W
System Power (DC)
unknown
61W
System noise level: 12~13 dBA@1m
(stock fans at ~580 RPM)

The new GPU test system is substantially more energy efficient with a 16W savings when idle and a massive 50W reduction on load. Furthermore, the Core i3-2100 is actually considerably faster than the previous Athlon II X3 435 despite having one less physical core so it’s less likely to bottleneck a discrete graphics card in any 3D tests we conduct.

TEST METHODOLOGY

Our test procedure is an in-system test, designed to:

1. Determine whether the cooler is adequate for use in a low-noise system.
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:

  • Sudden system shutdown, reboot without warning, or loss of display signal
  • Jaggies and other visual artifacts on the screen.
  • Motion slowing and/or screen freezing.

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.

2. Estimate the card’s power consumption. This is a good indicator of how efficient
the card is, and it affects how hot the GPU runs. The lower the better.

3. Determine how well the card decodes high definition video.

Test Platform

Measurement and Analysis Tools

3D Performance Benchmarks (for low-end/budget graphics processors only)

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)
65.5
90.7
149.0
199.6
251.2
300.3
400.9
AC Input (W)
81
105
166
211
265
322
426
Efficiency
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.

Ambient Noise Level

Our test system’s CPU fan is a low speed Scythe that is set to full speed at all times. The two Antec TrueQuiet 120 case fans are connected to the motherboard and are controlled using SpeedFan. Three standard speed settings have been established for testing.

GPU Test System:
Anechoic chamber measurements
Setting
System Fan Speed
System SPL@1m
High (loud)
1130 RPM
26 dBA
Med (quiet)
820 RPM
18 dBA
Low (silent)
580 RPM
12~13 dBA
Note: mic is positioned at a distance of one meter from the center of the case’s left side panel at a 45 degree angle.

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.

Video Test Suite


1080p | 24fps | ~22 mbps

H.264/MKV: A custom 1080p H.264 encoded clip inside an Matroska container.


1080p | 24fps | ~2.3 mbps

Flash 1080p: The Dark Knight Rises Official Trailer #3, a YouTube HD trailer in 1080p.

Testing Procedures

Our first test involves monitoring the system power consumption as well as CPU and GPU temperatures during
different states, idle, under load with Prime95 to stress the processor, and Prime95 plus FurMark to stress both the CPU and GPU simultaneously. This last state is an extremely stressful, worst case scenario test which generates
more heat and higher power consumption than can be produced by a modern video
game. If the card can survive this torture in our low airflow system, it should be
able to function normally in the vast majority of PCs. Noise levels are measured and recorded as well; if we deem the card’s fan control to be overly aggressive, we can adjust them at our discretion using various software tools.

Our second test procedure is to run the system through a video test suite featuring
a variety of high definition clips played with PowerDVD and Mozilla Firefox (for Flash video). During playback, a CPU usage graph is created
by the Windows Task Manger for analysis to determine the average CPU usage.
High CPU usage is indicative of poor video decoding ability. If the video (and/or
audio) skips or freezes, we conclude the GPU (in conjunction with the processor)
is inadequate to decompress the clip properly.

Lastly, for low-end and budget graphics cards, we also run a few gaming benchmarks to get a general idea of the GPU’s 3D performance. We don’t feel this is necessary for high-end models as there are many websites that do this in painstaking detail.

GPU Cooler Testing

Heatsink testing requires only the Prime95 plus FurMark stress test to be used. The fan(s) (if applicable) are connected to a custom external fan controller and tested at various speeds to represent a good cross-section of its airflow and noise performance.

Our GPU cooler test card is an HIS Radeon HD 5870 iCooler V Turbo, a factory-overclocked single GPU card that draws about 236W by our estimates. The stock VRM heatsink is left on for convenience.

Tested Cards

Card Model
Core Clock
Memory Clock
Memory Type
650 MHz
900 MHz
GDDR3 512MB
550 MHz
1000 MHz
GDDR5 512MB
650 MHz
1000 MHz
GDDR5 512MB
800 MHz
1125 MHz
GDDR5 1GB
ASUS Radeon HD 6850
DirectCU

(EAH6850 DC/2DIS/1GD5)
790 MHz
1000 MHz
GDDR5 1GB
900 MHz
1050 MHz
GDDR5 1GB
875 MHz
1225 MHz
GDDR5 1GB
ASUS GeForce GT 430
(ENGT430/DI/1GD3(LP))
700 MHz
800 MHz
DDR3 1GB
ASUS GeForce GTS 450
DirectCU

(ENGTS450 DIRECTCU/DI)
783 MHz
902 MHz
GDDR5 1GB
823 MHz
1002 MHz
GDDR5 2GB
1020 MHz*
1502 MHz
GDDR5 2GB
*GPU Boost up to 1111 MHz (1084 MHz according to specifications)

Breaking in our new test system is a wide variety of both AMD and Nvidia graphics cards from the past few years, ranging from entry level passively cooled cards to high performance models with large, non-reference coolers.

TEST RESULTS

Power Consumption

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 PCIe 16X 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 add-on 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.

The energy efficiency of modern day graphics cards is quite good. With the exception of the few high performance models in our roundup, most of the cards idle using less than 15W. On load, the HIS HD 5870 had the highest draw of 236W, one of the reasons its our reference card for GPU heatsink testing. Also note that both the HD 6870 and HD 5870 were tested using the GELID Icy Vision, an excellent third party cooler with its fans running at 5V (~1260 RPM), possibly skewing their results slightly as better cooling increases overall energy efficiency.

The cards tested had very similar CPU usage during video playback, 1~2% for our 1080p H.264/MKV test clip, and 8~9% for our YouTube HD sample. Power consumption however varied greatly, dependent mostly on the core/memory clocks used by the cards’ respective video decoder chips. The higher-end HD 5000/6000 series cards are particularly prone to this issue, being substantially more power hungry than their low-end counterparts and competitors on the Nvidia side.

The GTX 680 is especially impressive in this regard, consuming only 6~8W more than its idle usage while the HD 5870 and HD 6870 produced an increase four to five times higher.

Noise & Cooling

System Measurements: GPU Test System
Model
Idle
Load
GPU Temp
SPL @1m
GPU Temp
SPL @1m
Sapphire HD 7750 Ultimate*
38°C
12~13 dBA
77°C
12~13 dBA
AMD HD 5450*
44°C
12~13 dBA
70°C
12~13 dBA
HIS HD 5870 Turbo
(Icy Vision at 5V)
35°C
17~18 dBA
87°C
17~18 dBA
AMD HD 6870
(Icy Vision at 5V)
38°C
17~18 dBA
78°C
17~18 dBA
HIS HD 5550 Silence*
(system fans on med)
52°C
18 dBA
106°C
18 dBA
ASUS GT 430
35°C
23 dBA
78°C
23 dBA
ASUS HD 6850 DirectCU
39°C
17 dBA
77°C
25 dBA
ASUS GTS 450 DirectCU
31°C
19~20 dBA
66°C
26 dBA
AMD HD 6570
44°C
15 dBA
90°C
27~28 dBA
ASUS GTX 680
DirectCU II OC
32°C
15 dBA
73°C
27~28 dBA
Gainward GTX 560
Ti Phantom
37°C
18 dBA
86°C
37 dBA
Ambient temperature: 22°C
*passively cooled

Video card noise has also improved over the years with our idle test system producing less than 20 dBA@1m with almost all of the cards tested. The one exception, the Asus GT 430, is one of those rare models that lacks fan control altogether. Another card of note is the HIS HD 5550 Silence which overheated in our test system’s default fan setting; it survived after increasing the speed to medium but with a temperature above 100°C. Its heatsink isn’t very well designed, lacking surface area due to its very broad fins.

A few of the cards also had overaggressive fan behavior in our opinion. The three ASUS DirectCU cards had plenty of thermal headroom; with some user induced custom fan control all three can be considerably quieter while maintaining an adequate level of cooling.

MP3 Sound Recording

This recording was 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. It’s intended to give you an idea of how our test system sounds
in actual use — one meter is a reasonable typical distance between a computer
or computer component and your ear. The recording contains stretches of ambient
noise that you can use to judge the relative loudness of the subject. Be aware
that very quiet subjects may not be audible — if we couldn’t hear it from
one meter, chances are we couldn’t record it either!

The recording starts with 5~10 seconds of room ambiance, followed by 5~10 seconds
of the GPU test system with its case fans at various speeds. 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.

FINAL THOUGHTS

Our 2012 GPU test system has similar qualities as the old platform, being a very quiet machine with excellent acoustics that allows us to easily discern differences in noise produced by modern day graphics cards. It represents a good thermal challenge too, an airflow starved scenario that puts video card heatsinks to a more extreme test than a typical quiet PC. By default, the system’s case fans are set to almost inaudible levels to expose poorly designed coolers and/or energy inefficient cards. Even such a card would have a chance as our fan control setup allows us to give a little help if necessary.

We believe that this more modern, energy efficient setup is a more accurate representation of quiet systems found in the wild and will help us more accurately evaluate video cards and coolers for acoustics, thermals, and power consumption.

* * *

Articles of Related Interest
ASUS GeForce GTX 680 DirectCU II OC

Sapphire HD 7750 Ultimate Edition

ASUS DirectCU & AMD Radeon HD 6850 Graphics Cards

AMD Radeon HD 6570 & 6670 Budget GPUs

Arctic Cooling Accelero Xtreme Plus GPU Cooler

AMD Radeon HD 6870 Graphics Card

* * *

Discuss this article in the
SPCR forums.

Leave a Comment

Your email address will not be published. Required fields are marked *