The Setsugen is a unique GPU cooler that uses only a single extra slot and has a slim 120mm fan next to the GPU core blowing outward through the heatsink body below. We put the Setsugen through an arduous torture test. Can this odd single-fan design tame one of the hottest cards on the market, the Radeon HD 4890?
February 28, 2010 by Lawrence Lee
Product | Scythe Setsugen GPU Cooler |
Manufacturer | Scythe |
Street Price | ~US$40 |
The best CPU coolers have tower designs extending high over the socket with
the fan blowing the hot air emanating from the heatsink toward the rear or ceiling
exhaust. For video cards, this isn’t an option as it would put a lot of strain
on the PCB and block most or all of the expansion slots underneath. For high-end
cards, the stock cooling solution typically involves enclosing the heatsink
and attempting to exhaust out the back using blower-style fans which are not
that efficient.
The box next to a Radeon HD 4870. |
Most aftermarket graphics card coolers position the fan(s) on the outside of
the heatsink blowing toward the GPU core and circuit board similar to downblowing
CPU heatsinks. The Setsugen is the first GPU cooler we’ve come across that actually
does the opposite. It has a single 120mm fan underneath the main fin mass blowing
outward through the heatsink (downward in a typical ATX case). This seems like
a flawed approach as heat rises so in theory the exhausted air will simply drift
back up toward the GPU. In addition, placing the fan between the PCB and heatsink
gives it very little breathing room and no source of cool intake air.
Its biggest asset is its slim profile that won’t take up more than one extra
expansion slot, which seemingly makes it an attractive option for dual video
card systems. On most Crossfire/SLI motherboards, there are only two slots between
the full length PCI-E connectors. Using the Setsugen on the upper card would
leave one slot of breathing room underneath, while larger heatsinks, including
those of the dual fan variety only have half a slot of clearance, or none at
all. However without a second fan, the Setsugen cannot deliver direct airflow
over the voltage regulation area on most high power cards. Our test card for
GPU coolers is the notoriously hot Radeon
HD 4890, which we know first hand has very hot VRMs and draws almost
150W. It is perhaps more than the Setsugen can handle.
Package contents. |
The package contents include mounting hardware, a series of smaller heatsinks
for cooling memory, VRMs and other chips, and the cooler itself. A manual fan
speed is connected to the fan, designed to be mounted in an expansion slot so
it can be adjusted from the outside.
Scythe Setsugen: Key Features (from the product web page) | |
Feature & Brief | Our Comment |
M.A.P.S. (Multiple Airflow Pass-Through Structure) By introducing the newly developed M.A.P.S. (Multiple Airflow Pass-Through Structure), the heatsink has been optimized to fully utilize the ability of the included Scythe original “Slip Stream Slim” high airflow 120 mm case fan. | Basically means slits breaking up the heatsink body into several sections which lowers resistance to airflow. The same feature can be found on their Mugen-2 CPU cooler. |
Slip Stream 120 mm Slim with Fan Controller The new Slip Stream 120 mm Slim is a consistent further development with all the improvements of the original Slip Stream 120 mm fans. Equipped with the fan controller, you can control the fan speed to meet your cooling requirements. | Slip Streams are renowned for their smooth, quiet acoustics. The fan controller included attaches to a PCI slot bracket so it can be controlled manually from the outside. |
Flexibility Scythe Setsugen VGA Cooler improves the cooling performance for graphic cards for faster graphic processing and more performance for your PC! | Setsugen is compatible with the majority of Nvidia and ATI graphics cards going back up to 4 generations. |
Scythe Setsugen: Specifications (from the product web page) | |
Model Name: | Setsugen VGA Cooler |
Model #: | SCVSG-1000 |
Manufacturer: | Scythe Co., Ltd. Japan |
Overall Dimensions | 180 x 33.5 x 139 mm / 7.09 x 1.32 x 5.47 in |
Weight: | 375 g (13.22 oz) |
Supplied Fan | Scythe Slip Stream 120 mm SLIM (with fan controller) |
Fan Dimensions | 120 x 120 x 12 mm / 4.72 x 4.72 x 0.47 in |
Fan Speed | 800 (± 30%) ~ 2.000 rpm (± 10%) |
Noise Level | 19.53 ~ 33.67 dBA |
Air Flow | 19.40 ~ 45.47 CFM |
Compatibility | |
nVIDIA | Geforce Fx5900 /6600 /6800 /7300 /7600 /7800 / 7900 /8500 /8600 /8800GTS (G92) /8800 GT (G92) 8800 (G80) /9600 GT /9600 GSO /9800 GT /9800 GTX / 9800 GTX+ / GTS 250 / GTX 260 / GTX 275 / GTX 280 / GTX 285 |
ATI | 9*** /X*** /X1300 /X1600 /X1650 /X1800 /X1900 / X1950 /HD2600 /HD2900XT /HD3650 /HD3850 /HD3870 / HD4770 /HD4830 /HD4850 /HD4870 /HD4890 /HD 5850 / HD 5870 *Radeon HD 5770, HD 5750 are incompatible |
PHYSICAL DETAILS
The Setsugen is composed of four copper heatpipes and a set of 67 friction-fit
aluminum fins and bears a mild resemblance to the Accelero
S1, though its a much shorter cooler and its fins are more densely packed.
It measures 7.1″ long and weighs in at 360 grams according to our digital
scale.
The fan is a slim Slip Stream 120mm model with a rated speed |
The clearance between the fan and the base is very small. As |
Like Scythe’s Mugen-2 |
On average, the fins are 0.30mm thick and 1.55mm apart. |
Like most Scythe heatsinks, its base is flat and the nickel-plating |
INSTALLATION
The most critical aspect of installation is for the heatsink to be securely
mounted. A firm mating results in good contact between the heatsink’s base and
the GPU core and more efficient heat conduction. Ease of installation is also
important; a simple mounting scheme means less hardware to fiddle with, and
a reduced likelihood of screwing up.
|
As a result of this quirk, we actually had to install |
The longer sides of the screws go through the PCB and are secured |
|
|
TEST METHODOLOGY
Our test procedure is an in-system test, designed to 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.
Test Platform
- Intel
Pentium D 930 Presler core processor. Official TDP of 95W. - Gigabyte G-Power 2 Pro
heatsink, modified with a Scythe
Slip Stream 500RPM 120mm fan. - AOpen
i945Ga-PHS motherboard – Intel i945Ga Chipset; built-in VGA. - Corsair
XMS2 memory 1GB, DDR2-800 - HIS Radeon HD 4890 Turbo Edition
graphics card. Approximate power draw of 149W. - Seagate Momentus 5400.3
160GB, 5400RPM, notebook hard drive, suspended - Seasonic S12-600
ATX12V v2.0 compliant power supply, modified with a Scythe
Slip Stream 800RPM 120mm fan @ 5V. - Antec P180B case,
modified in detail below. - Nexus 120mm
fan connected to a variable fan speed controller. - Microsoft
Windows XP Professional SP3 operating system - ATI
Catalyst 9.11 graphics driver
Measurement and Analysis Tools
- CPUBurn
processor stress software. - FurMark
stability test to stress the GPU. - GPU-Z to
monitor GPU temperatures.. - A custom-built variable fan speed controller to power the system
fan - PC-based spectrum analyzer:
SpectraPlus with ACO Pacific mic and M-Audio digital
audio interfaces. - Anechoic chamber
with ambient level of 11 dBA or lower
|
GPU coolers are tested on a HIS Radeon HD 4890 Turbo Edition, one of the most
power hungry single GPU cards on the market, drawing almost 150W by our estimates.
The stock cooling plate is left on to cool the memory chips and the voltage regulators
as the 4870/4890 series requires good VRM cooling and the VRM heatsinks included
with most aftermarket VGA coolers are usually too small to be effective.
Our main test consists of FurMark stability test running in conjunction with
CPUBurn to stress both the graphics card and processor simultaneously. This
combination produces more CPU/GPU stress than a typical gaming session. As our
test system has very limited airflow, our results are not indicative of a real-world
situation, but rather a worse-case scenario. If the heatsink in question can
cool the card and its components adequately in this environment it means there
will be some degree of thermal headroom when deployed in a more conventional
situation. GPU temperatures are recorded using GPU-Z. On our HD 4890 test card,
there are three main on-die sensors, as well as three on the primary voltage
regulators. We average the results of each set of sensors.
The cooler is tested at various speeds to represent a good cross-section of
its airflow and noise performance. Noise is measured and recorded with our test
system on with the heatsink installed. 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.
TEST RESULTS
Noise
The included fan controller had a range of 830~2020rpm which is the equivalent
of 4.4V~12.0V. With our custom fan controller, we found that the fan’s starting
voltage was 4.3V, so the included controller effectively gives you full control.
Stock Fan Measurements (in system) | ||
Voltage | Speed | SPL@1m |
12V | 2020 RPM | 30 dBA |
9V | 1680 RPM | 26 dBA |
7V | 1400 RPM | 21 dBA |
6V | 1210 RPM | 18 dBA |
5V | 1000 RPM | 15 dBA |
4.3V | 800 RPM | 13~14 dBA |
Ambient noise level: 11 dBA (12 dBA with the test system on) |
The noise generated by the Setsugen’s fan has a whiny, turbulent character
which becomes smoother and less annoying at lower speeds. That being said, it
is still much better than most aftermarket coolers as it doesn’t hum and there
is almost no chatter emanating from the fan’s bearings. The turbulence is very
pronounced due to the fan’s constrictive environment which causes it to be fairly
loud. In our test system, it was audible throughout its range at one meter,
becoming tolerable at 7V, but not really quiet until 6V.
|
|
Performance
Test Results: Scythe Setsugen | |||
Fan Voltage | SPL@1m | Avg. Core Temp | Avg. VRM Temp |
12V | 30 dBA | 84°C | 115°C |
9V | 26 dBA | 91°C | 128°C |
7V | 21 dBA | 95°C | 132°C |
6V | 18 dBA | 103°C | 146°C |
5V | 15 dBA | stopped | stopped |
Ambient temperature: 23°C Ambient noise level: 11 dBA (12 dBA with the test system on) |
The Setsugen was only able to keep the GPU temperature below 90°C at higher
fan speeds. At 6V, the only quiet level we fully tested, the GPU temperature
increased past 100°C and the VRMs reached 146°C. At 5V, the VRM temperature
quickly surpassed 150°C, within 5~10°C of card failure, so we stopped
testing.
These results were obtained using the stock cooling plate; if our HD 4890 did
not have one, and we were forced to use the small heatsinks provided by Scythe,
we likely would’ve stopped at 9V. The VRMs ran 15~20°C hotter on the dual
fan Musashi using Scythe’s heatsinks rather than the stock cooling plate, and
we imagine it would be even higher on the single fan Setsugen.
Now you may be wondering if you can flip the fan around and blow it onto the
GPU. With some effort the wire clips holding the fan can be removed, but simply
reversing the fan’s orientation isn’t possible. Like most fans it doesn’t have
struts on the back, so the blades and hub on the backside rub up against the
heatsink. In a typical tower case, gravity will pull the it downward and prevent
it from spinning altogether. If you strap the fan onto the outside of the heatsink,
you might as well get the dual fan Scythe Musashi instead as the thickness would
be comparable.
Comparison
Scythe Setsugen vs. ZEROtherm CoolMaxx 2000 & Zalman VF1000 LED | ||||
Cooler | Fan Voltage | SPL@1m | Avg. Core Temp | Avg. VRM Temp |
Setsugen | 7V | 21 dBA | 95°C | 132°C |
CoolMaxx 2000 | 9V | 20 dBA | 90°C | 138°C |
VF1000 LED | 9V | 20 dBA | 100°C | 139°C |
8V | 19 dBA | 102°C | 145°C | |
CoolMaxx 2000 | 7V | 18 dBA | 95°C | 149°C |
Setsugen | 6V | 18 dBA | 103°C | 146°C |
Ambient temperature: 23°C Ambient noise level: 11 dBA (12 dBA with the test system on) |
The ZEROtherm CoolMaxx
2000 has a similar design to the Setsugen in that its fan is underneath
the heatsink. However, it’s smaller and lighter, and the fan utilized is a thicker
80mm model that blows toward the GPU. It beats the Setsugen in GPU core temperature
by a good margin and with lower noise levels to boot. Subjectively the Setsugen’s
fan sounds much better though and it generates slightly better VRM temperatures.
In this particular case, the HD 4890’s VRMs are more likely to cause instability
so the Setsugen may actually be a better choice.
Scythe Setsugen vs. Musashi | ||||
Cooler | Fan Voltage | SPL@1m | Avg. Core Temp | Avg. VRM Temp |
Setsugen | 7V | 21 dBA | 95°C | 132°C |
Musashi | 10V | 19 dBA | 90°C | 117°C |
Setsugen | 6V | 18 dBA | 103°C | 146°C |
Musashi | 9V | 17 dBA | 93°C | 125°C |
Ambient temperature: 23°C Ambient noise level: 11 dBA (12 dBA with the test system on) |
The Setsugen is clearly no match for the Musashi which also manages to be quieter
despite having two fans. At 9V it bests Setsugen at 6V by 10°C in core temperature
while producing 1 dBA less noise. The VRM cooling is even better with the Setsugen
losing by more than 20°C. If you can spare the extra room, the Musashi is
a far more efficient heatsink.
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.
These recordings are intended to give you an idea of how the product 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 10 seconds of room ambiance, followed by 10 seconds
of the VGA test system without a video card installed, and then the actual product’s
noise at various levels. 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.
- VGA Test
System with Scythe Setsugen at one meter
— 5V (15 dBA@1m)
— 6V (18 dBA@1m)
— 7V (21 dBA@1m)
— 9V (26 dBA@1m)
— 12V (30 dBA@1m)
Comparatives:
- VGA
Test System with Scythe Musashi at one meter
— 7V (14 dBA@1m)
— 8V (15 dBA@1m)
— 9V (17 dBA@1m)
— 10V (19 dBA@1m)
— 12V (23 dBA@1m)
- VGA
Test System with ZEROtherm CoolMaxx 2000 at one meter
— 5V (15 dBA@1m)
— 7V (18 dBA@1m)
— 9V (20 dBA@1m)
— 12V (22~23 dBA@1m)
- VGA
Test System with Zalman VF1000 LED at one meter
— 8V (17 dBA@1m)
— 9V (19 dBA@1m)
— 10V (20 dBA@1m)
— 11V (22 dBA@1m)
— 12V (22 dBA@1m)
FINAL THOUGHTS
The Setsugen managed to cool our HD
4890 test card quietly but just barely with temperatures rising beyond
our comfort level. As a result we cannot recommend it for cards with very high
power draws, say 125W and above, unless the fan is kept at higher speeds. This
unfortunately undermines one of its biggest strengths: its thickness. As the
Setsugen only takes up one extra slot, one would immediately believe it would
be ideal for a quiet CrossFire or SLI system. If you want to keep the noise
level down, you’ll need to stick to midrange models, the type of cards that
probably aren’t worth dualing.
The overall design of the Scythe Setsugen is deficient in several ways. Scythe
fans are known for their good acoustics and low noise level but the tight confines
of the fan increases turbulence effects, making the cooler much louder than
it would have been otherwise. Given its proximity to the GPU core, if the fan
were flipped to blow toward it, we believe it would give better results. Unfortunately
this isn’t an easy change to implement as the back of the fan would press up
against the heatsink, preventing it from spinning. While the fan is large, it
is simply not wide enough to deliver direct airflow over the VRMs on most graphics
cards, and on high-end models in particular, VRM cooling is paramount. We also
take issue with the change Scythe implemented in the mounting system, specifically
the way the screws on the mounting plate slide around. Unless you are extremely
lucky, it is almost guaranteed that the contact between the heatsink base and
GPU core will be uneven.
Despite all its problems, it is still probably the best choice we’ve come
across as far as single-fan, dual slot coolers go. If you need a quiet cooler
that occupies no more than one extra slot for a card with a moderate power draw,
the Setsugen fits the bill rather nicely. If you have more room to spare, a
dual fan model is ideal. Its cousin, the Scythe
Musashi is the same weight, about half a slot thicker and a much better
performer. Best of all, it can be found for as little as $40, which is actually
$5 cheaper than the Setsugen.
Scythe Setsugen | |
PROS * Good high airflow performance | CONS * Poor low airflow performance |
Our thanks to Scythe
for the Setsugen sample used in this review.
* * *
Articles of Related Interest
ZEROtherm CoolMaxx 2000 GPU Cooler
Scythe Musashi Dual Fan GPU Cooler
[PostScript Added]
HIS Radeon HD 4890 Turbo Edition
Thermaltake Duorb VGA Cooler:
Are Two Orbs Better Than One?
Xigmatek Battle-Axe: First Direct-Touch
Heatpipe VGA Cooler
Arctic Cooling Accelero S1 VGA
Cooler
* * *