• Home
  • blog
  • Zalman 9300AT: Not me too, but me again

Zalman 9300AT: Not me too, but me again

blog image

It was one of the pioneers in quiet computing, but Zalman’s latest product, another shrink of a 3+ year old design, disappoints. The cooling performance is good, but the CNPS9300 AT doesn’t win any prizes for acoustics or value.

Zalman CNPS9300 AT CPU Cooler: The CNPS9700 Slims Down

October 1, 2008 by Lawrence Lee

Zalman CNPS9300 AT
LGA775 & K8 CPU Cooler
Tech Co.
Street Price
$45~50 USD

The Zalman CNPS9700
giant copper CPU cooler offered good performance but suffered
a high price-tag and a nasty sounding fan. These drawbacks overshadowed
the clever design of the heatsink itself, and made it difficult to recommend
for a silent PC. One solution was to replace the integrated fan, a procedure recommended only for the handy. We have not seen a new CPU heatsink design from Zalman
in quite some — the most recent additions to their CPU heatsink
lineup are rehashes of what they’ve done in the past. The CNPS9300
AT is no exception.

The CNPS9700 is no match for the cheap
tower heatpipe coolers that dominate the current market. The CNPS9300 AT is
a drastically slimmed down version of the 9700, so it shouldn’t be able to match
or beat its predecessor’s performance. It’s narrower, lighter, and is equipped
with a smaller fan. While it may offer more clearance around the CPU socket,
it is not significantly shorter, so case compatibility won’t be much better.

The most interesting thing about the CNPS9300 AT is the fact that it exists.
Why did Zalman produce it, and why now? What is its place in the market? Those
are the questions we hope to answer.

Zalman usually keeps their packaging to a minimal, and the CNPS9300 AT
follows that trend. Its box is compact and decorated with a calming blue


The box contains the heatsink in a plastic clamshell container, LGA775
and AMD mounting equipment, thermal grease and instructions. This is the
first time we’ve encountered a Zalman heatsink that does not include one
of their Fan Mate manual fan speed controllers. Its absence is probably
the result of Zalman finally incorporating a fan with a PWM connector.

Zalman CNPS9300 AT: Feature
Highlights (
from the
product web page
Feature & BriefOur Comment
Does not generate noise or vibration in Silent Mode.
Usually not the case with Zalman.
Compatible with Slim Tower / Low Profile Home
Theatre PC Enclosures.
This must be a mistake, probably taken
off the literature from the CNPS8700/7500 series. Slim/low profile cases
are not wide enough to accommodate a 133mm tall CPU cooler.
Compatible with all Intel Socket 775 based Single, Dual, and Quad Core
CPUs, and all AMD Socket AM2+, AM2, 754, 939, 940 based Single, Dual,
and Quad Core CPUs.
For $50, we expect universal compatibility.
100% copper heatsink with aerodynamically
optimized “tunnel” design for maximum cooling efficiency.
The same basic design as the 9500 and
Patented “figure 8” heatpipe design
for cooling performance of up to 4 heatpipes with the use of just 2.
Longer heatpipes can be utilized by curving
them. It’s a common feature.
Low-noise 92mm PWM fan for automatic control
of the fan’s RPM according to the CPU’s temperature.
Finally, Zalman decides to use a PWM
fan. Welcome to the 21st century.
Zalman CNPS9300 AT vs. CNPS9700:
from the product web pages here
and here)
ModelCNPS9300 ATCNPS9700
Dimensions 61.35(L) x 108(W) x 132.5(H) mm90(L) x 124(W) x 142(H) mm
Weight 407g764g
Base MaterialPure Copper & Pure AluminumPure Copper & Aluminum
Dissipation Area2,583cm25,490cm2
ModelCNPS9300 ATCNPS9700
Bearing Type Superflo2 Ball-Bearing
Speed1,500rpm ± 15 % ~ 2,400rpm ±
10 %
1,250rpm ± 30 %
2,800rpm ±
10 %
Noise Level
(at 1m)
20~30dBA ± 10%19.5dBA ± 10%
35dBA ± 10%
Control MethodPWM Control, Auto RestartManual via Fan Mate
Life Expectancy50,000 hoursN/A


The Zalman CNPS9300 has a 92mm ball bearing fan similar to that of the CNPS9500
(the CNPS9700 fan is 110mm wide), but lacks the third heatpipe of the 9500/9700.
It also has a reduced width, and thus lower weight — only 407g.

The fan sits a fair bit higher compared to the CNPS9500/9700. While this
allows for longer heatpipes to be used, the surrounding components may receive
less indirect airflow. It also makes the cooler less likely to interfere
with a large chipset cooler close to the
CPU socket.

The shape of the fins is odd, like the result of
an experiment in ceramics class. From front to back the surface expands,
contracts, then expands again. The lack of width results in the lowest
heat dissipation area for a Zalman heatsink in many years: 2,583cm2. The
CNPS9700 has more than twice that area.

The heatpipes are arranged in a figure eight pattern just like the 9500/9700.
They have less curvature however, resulting in the the fan and fins being
raised higher over the CPU.

The impeller is mounted to a plastic ring at the center of the heatsink
via three screws. The ring in turn is clipped to the fins at three different
points and are held on via tension. The CNPS9700 fan on the other hand
uses a metal stand attached to the base. The fins are packed very tight
at the interior but the gaps between them get progressively wider as the
fins extends outward.

Only the very bottom of the heatpipes make contact with the copper base.
The gaps between them are noticeable.

The base is very flat and polished extremely well. The reflection is crystal clear.


Tight and secure mounting is critical for
good heat transfer between CPU and heatsink. Ease of installation is also important — a simple
mounting scheme means less time spent installing, and reduced likelihood
of user error.

AMD mounting clip.

AMD installation is accommodate using a simple tension clip.
An L-shaped lever must be inserted at one end. There are two positions for this L-lever. According to the manual the outer one is for AMD and the inner
one for Intel. This makes no sense since the current Intel LGA775 CPU socket uses push-pins. On the bright side, the clip can be slipped
between the heatpipes, so that the cooler can be rotated 90 degrees on AMD motherboards,
allowing the fan to point in whatever direction is desired. This is a feature that’s not as common as it should be, due mainly to the rectangular shape of the AMD CPU mounting area.

The mounting for LGA775 is more involved, requiring free access to both sides of the motherboard. It is similar to previous Zalman heatsinks, quite secure and tight. You cannot install it with the motherboard already in the case.

The LGA775 mounting hardware. Installation on an Intel board involves
a more secure, three step process with a metal clip, clip support bracket
and back plate.

The back plate, as you can guess is placed on the back of the motherboard,
lined up with the mounting holes around the CPU socket.

The clip support bracket is then attached to the back plate using four
long bolts. The bracket was tall enough to clear the short capacitors
on our test board.

Finally a clip is placed over the mounting plate and screwed to the
support bracket. There is a notch in the top of the base to keep the heatsink


Some basic physical measurements have been
added to our test routine.

Zalman CNPS9300 AT: SPCR Measurements
400g (heatsink
420g (heatsink, mounting clip and screws)
Fin thickness
~ 0.17mm
Fin spacing
~ 2.16mm (outer edge)
Vertical Clearance (northbridge)
not an issue
(power supply)
~4 mm (this
will depend on the distance from the CPU socket to the edge of the PCB)

As the fin spacing is not uniform through the heatsink, measuring it has
limited comparative value and so we have omitted it from our fin thickness and
spacing comparison table.

Comparison: Fin Thickness & Spacing
Fin Thickness
Fin Spacing
Scythe Ninja
0.31 mm
3.95 mm
Thermalright HR-01 Plus
0.45 mm
3.15 mm
Noctua NH-U12P
0.44 mm
2.63 mm
Noctua NH-C12P
0.47 mm
2.54 mm
Xigmatek HDT-S1283
0.33 mm
1.96 mm
Zerotherm Zen FZ120
0.37 mm
1.80 mm
Thermalright Ultra-120
0.45 mm
1.42 mm

Testing was done according to our
unique heatsink testing methodology
, and the reference fan was profiled
using our standard fan testing
. A quick summary of the components, tools, and procedures
follows below.

Key Components in Heatsink Test Platform:

  • Intel
    Pentium D950
    Presler core. TDP of 130W; under our test load, it measures
    78W including efficiency losses in the VRMs.
  • ASUS
    motherboard. A basic microATX board with integrated graphics
    and plenty of room around the CPU socket.
  • Samsung MP0402H
    40GB 2.5" notebook drive
  • 1
    GB stick of Corsair XMS2
    DDR2 memory.
  • FSP Zen 300W
    fanless power supply.
  • Arctic Silver
    : Special fast-curing thermal interface material, designed
    specifically for test labs.
  • Nexus 120 fan (part of our standard testing
    methodology; used when possible with heatsinks that fit 120x25mm fans)

Test Tools

  • Seasonic
    Power Angel
    for measuring AC power at the wall to ensure that the
    heat output remains consistent.
  • Custom-built, four-channel variable DC power supply,
    used to regulate the fan speed during the test.
  • PC-based spectrum analyzer: SpectraPlus with ACO Pacific mic and M-Audio digital audio interfaces.
  • Bruel & Kjaer (B&K) model 2203 sound level meter (SLM)
  • Anechoic chamber with ambient level of 11 dBA or lower
  • Various other tools for testing fans, as documented
    in our standard fan testing

Software Tools

  • SpeedFan
    , used to monitor the on-chip thermal sensor. This sensor is not
    calibrated, so results are not universally applicable.
  • CPUBurn
    , used to stress the CPU heavily, generating more heat than most
    real applications. Two instances are used to ensure that both cores are stressed.
  • Throttlewatch
    , used to monitor the throttling feature of the CPU to determine
    when overheating occurs.

Noise measurements were made with the fan powered from the lab’s variable DC
power supply while the rest of the system was off to ensure that system noise
did not skew the measurements.

Load testing was accomplished using CPUBurn to stress the processor, and the
graph function in SpeedFan was used to make sure that the load temperature was
stable for at least ten minutes. The stock fan was tested at various voltages
to represent a good cross-section of its airflow and noise performance.


Stock Fan Testing

The stock fan was tested briefly for its noise characteristics. The results
were not surprising.

37 dBA@1m
2380 RPM
30 dBA@1m
1790 RPM
25 dBA@1m
1050 RPM
21 dBA@1m
790 RPM
18 dBA@1m
790 RPM
*Test platform placed on soft foam

The CNPS9300 AT fan is very loud, especially at high speeds — an unfortunate
trademark of Zalman fans. The CNPS9300 is subjectively poorer than its predecessors
due to the vibrations transmitted by the fan — far more noticeable
on this particular model than any of Zalman’s previous coolers. When we placed the test board
on foam, SPL dropped by 3 dBA at 5V. In addition, though
the Zalman has finally gone with a PWM fan, reducing the fan voltage did not
result in drastic reductions in fan speed as we typically see with PWM fans.

Cooling Results

Zalman CNPS9300
Fan Voltage
SPL @1m
°C Rise
37 dBA
30 dBA
25 dBA
21 dBA
Load Temp: CPUBurn for ~10 mins.
°C Rise: Temperature rise above ambient (23°C) at load.
°C/W: based on the amount of heat dissipated by the CPU (measured
78W); lower is better.

Fan @ 12V: Performance was excellent at 14°C above ambient temperature.
This is only two degrees higher than the best load temperature we’ve ever recorded.
Of course at 12V the fan was just plain loud. Very growly and aggressive. The
vibration noise was somewhat masked by turbulence.

Fan @ 9V: The CPU temperature increased 3°C. The fan sounded less harsh,
and more tonal, whiny.

Fan @ 7V: Thermal rise was an additional one degree. The fan exhibited chuffing
and the vibrations became more noticeable.

Fan @ 5V: Cooling suffered another 5°C — it would seem the cooler’s
sweet spot is between 9V and 7V (roughly 1000~1800 RPM). The fan chuffed and clicked up close. The
vibration though dominated the noise signature — it was very conspicuous,
as if the fan were not secured properly.


Though the noise measurements we took of the CNPS9300 AT were taken in our
new anechoic chamber, items generating over 20 dBA@1m measure roughly the
same in both our old and new testing facilities. Since all our measurements are over 20 dBA@1m, the results can be compared
to those taken previously.

Zalman CNPS9300 AT: Comparables
Thermaltake V1
SPL @1m
°C Rise
SPL @1m
°C Rise
SPL @1m
°C Rise
30 dBA
30 dBA
30 dBA
24 dBA
23 dBA
25 dBA
21 dBA
20 dBA
21 dBA

The performance of the CNPS9300 AT is eerily similar to two other loud heatsinks
we’ve tested previously. There was very little measurable difference between
the CNPS9300, CNPS9700 and another loud CPU cooler, the Thermaltake
, at equivalent noise levels. Surprisingly, despite having only two
heatpipes and a greatly reduced width and surface area, the CNPS9300 performs
almost exactly the same as the CNPS9700. As the mounting system is the same,
and fan is actually smaller, this suggests that the 9700 may have been
bigger than it needed to be. (Editor’s Note: On the other hand, with a significantly hotter CPU, you might see an advantage in the 9700. However, not many current CPUs run hotter than our test platform’s Pentium D950, which has a TDP of 130W.)


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 second intervals of room ambient, followed by
10 seconds of 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.


The CNPS9300 AT is basically a compact version of the CNPS9700 with few improvements,
yet the performance is comparable. It would seem that Zalman’s design is fundamentally
a good one. If any other heatsink lost half
its surface area, its cooling ability would probably be seriously compromised.

The mounting system for Intel motherboards is secure, and while the same cannot
be said for AMD boards, the heatsink’s weight is quite low, actually under the 450g maximum recommendation of both AMD and Intel, so it’s not much of an issue. Unlike most AMD heatsinks, Zalman’s mounting system allows the fan to blow in any direction.

Where Zalman continues to fail is in fan acoustics. It is much too loud. The vibration problem could have been a one-off due to bearing damage in transit, but that’s just a conjecture. It seems to have less copper than their
other CPU coolers, and it’s cheaper, but $45-$50
is still too much for this level of performance and noise.
There are many cheaper alternatives that outperform the CNPS9300 AT,
in both cooling and noise. So where does the CNPS9300 AT fit in? It doesn’t really
— there’s no situation where it would be the best choice.

Zalman seems set in their ways, continuing to use bad fans and expensive copper
designs while the rest of the cooling world moves on. They were pioneers in low-noise
computing, but today their innovations are few and far in between. We can only shrug and wait to see if Zalman slides further into mediocrity or rejuvenates itself in the coming months.

Zalman CNPS9300 AT

* Good, not great performance
* Low weight
* Secure LGA775 installation
* Rotatable AMD installation


* Very loud fan
* Bad vibration problem
* Expensive

Our thanks to Zalman
for the CNPS9300 AT heatsink sample.

* * *

Articles of Related Interest
NH-U12P Tower Cooler

Zen FZ120 CPU Cooler

NH-C12P: A Top-Down Cooler Rises Up

HR-01 Plus: 2nd Gen Killer Tower Cooler

Zipang 14cm fan "blow-down" CPU cooler

Thermaltake V1: "Peacock
Tail" Cooler

* * *

Discuss this review in the SPCR forums.

Leave a Comment

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