Zalman CNPS8000: A Worthy Successor?

Zalman has broken with tradition and released a low profile heatsink that is not based on their’s classic “flower” clamped-fins design. It’s the same height as the venerable CNPS7000, but it has heatpipes and it weighs 100 grams less. How does it perform?

July 30, 2006 by Devon
Cooke

For several years, Zalman has relied on its circular “flower” design
to bring some very unusual — and successful — heatsinks to market.
The CNPS7000 was untouchable
when it was first released and is still a respectable heatsink, despite the
fact that it doesn’t have heatpipes. More recently, the
CNPS9500 has found considerable success in the high end market and is notable
as much for its design elegance as its excellent performance.

The CNPS8000 marks a departure from Zalman’s traditional design. It is surprisingly
conventional for a modern heatsink, with four heatpipes to transfer heat from
a copper base to a bed of aluminum fins and a proprietary fan to provide airflow.

The motivating force behind the CNPS8000 seems to have been size;
Zalman lists compatibility with “Slim and Low Profile” computers as
the first feature. However, it’s a little unclear why Zalman needed to produce
such a heatsink. At 62mm, it is no shorter than the CNPS7000 — a design
that has already proven itself for Zalman and is available for about US$20.
The CNPS8000 will need to provide significantly better performance than its
predecessor if it is to provide better value for money.

The retail box is large, and displays the heatsink prominently through a
clear window.

The window is part of this plastic sheath that protects the heatsink during
transit.

As with all of Zalman’s recent heatsinks, the CNPS8000 ships with a Fanmate2
fan controller that allows the fan voltage to be set between 5~11V. The fan
controller is one of Zalman’s biggest claims to fame, and is one of the reasons
why Zalman has acquired a reputation for selling quiet equipment. That aside,
there’s nothing too unusual in the package: A manual, a Zalman sticker, some
thermal interface material, mounting clips, and the heatsink itself.

The included Fanmate2 makes slowing down the stock fan simple.

Several strengths and weaknesses can be gleaned just by reading the specifications.
The CNPS8000 is quite small and light. We are sceptical about the
performance figures as even the best heatsinks struggle to break 0.20 °C/W.
We have never tested a heatsink that reached 0.13 °C/W. Adding to the confusion
is the way that Zalman states the performance figures. According to their spec
sheet, we are supposed to believe that the heatsink performs better in Silent
Mode. We have assumed that this is an error, and have reversed the figures to
restore some plausibility to the figures.

PHYSICAL DETAILS

Aside from the fact that it has a fan embedded in it, the design
of the heatsink is nothing special. It is quite similar to Thermalright’s XP-90
and XP-120 heatsinks. The
fan blows down on a bed of fins that extend upwards from the base. Four
heatpipes help to distribute heat evenly across the fins.

The fins are quite tightly spaced, and, unlike Zalman’s flower
heatsinks, they do not have the advantage of being more widely spaced away from
the center of the heatsink. All of the fins are aluminum (no hybrid AlCu model
here), allowing the weight to be kept to a minimum. At 350 grams, the CNPS8000
is lighter than most other aftermarket heatsinks.

The fan is embedded into the body of the heatsink.

The body of the heatsink is very rigid. The fins do not hang from the heatpipes
like they do on several other heatsinks; they are firmly attached to the base
of the heatsink, maximizing surface area of the fins while keeping height to
a minimum. However, without a clear exhaust path for the air, the most central
fins are unlikely to be very effective.

The fins are fixed at the base.

Four heatpipes are used.

The base is made of machined copper, and is smooth enough to be serviceable.
A certain amount of grain is visible in the surface, but it could not be felt
with a fingernail and is unlikely to affect performance.

Some grain is visible in the base.

FAN DETAILS

The fan is a Zalman creation, and features an odd half-frame that seems
to exist solely to attach the fan to the heatsink. The outline of the Zalman
logo is cut into the side, allowing a small amount of air to escape. It seems
likely that these holes will cause turbulence noise.

The fan is rated for 0.23A, marking it as medium speed. Unfortunately the unique
frame makes the fan impossible to replace, so we will have to count on Zalman
to include a quiet fan, something they have not always done well in the past.

The fan can be removed, but it’s hard to imagine how it could be replaced.

INSTALLATION

Two installation brackets are included, one for Intel’s Socket 775, and one
for the various K8 sockets on the market. The brackets are split into two halves
that slide and lock together, allowing the brackets to be added to or removed
from the heatsink without requiring tools.

Because there are so few parts, putting the bracket together is quite intuitive.
The photos below show how the two halves mate and how they attach to the heatsink.
The halves do not fit tightly, but they are a little difficult to mate properly
and it is easy to push them together only to find that one of the arms has not
connected properly. With due care and attention, it does not take long to install.

The two halves of the mounting bracket mate together…

…like so.

In actual use, they slide together over the base of the heatsink, allowing
the mounting system to be switched easily.

The K8 bracket uses the retention module that has come standard on K8 motherboards
since the first Athlon 64 was released. The system is quite familiar and easy
to understand, but can be hard to use because of the high tension involved.

Socket 775 does not have a retention module, so Zalman has opted to use a backplate
and screws to secure the heatsink. The screws are built into the bracket itself,
and aligned perfectly with the backplate that was already on our motherboard.
Installing the heatsink was a simple matter of tightening the screws. While
not as easy to use as the plastic clips used on Intel’s stock mounting mechanism,
the screw-and-backplate method is probably more secure, as there is less plastic
under tension. Note that the four screws for 775 are accessible with a screwdrive from the top through the gap between the fan blades and the frame.

The bracket for Socket 775 screwed to the backplate.

TESTING

As time goes on, fewer and fewer heatsinks continue to support Socket 478,
which is rapidly fading into the history of computing. This is the case with
the CNPS8000, which does not fit on our standard 478 socket heatsink
testing rig.

Instead, our Socket 775 test bench was used. This is the same
system used to test the Freezer 7 Pro. Details of the system are outlined
below. So far, only five heatsinks have been tested on this system: The
Arctic Cooling Freezer 7 Pro, the
Spire Verticool II, the Scythe Ninja, and the Thermalright XP-120, and,
most recently, the Arctic Cooling
Alpine 7. No other heatsinks we have tested are directly comparable with
the results of this review. The Pentium 520 used in this test is cooler than
most of Intel’s soon-to-be-forgotten Prescott and Presler chips, but it is still
15-20W hotter than the P4-2.8 Northwood used in our socket 478 HS testing platform.
On the other hand, it is also hotter than almost about every AMD processor on
the market, not to mention Intel’s new Core 2 Duo chips, many of which draw
much less power than the Intel 520.

On the test bench…

Test Platform

• Intel
  520 processor (P4-2.8 Prescott, 1 MB cache, 800 MHz FSB in
  775 casing). The Thermal Design Power is 84.0W. The Maximum
  Power, as calculated by CPUHeat
  & CPUMSR, is 100.76W.
• AOpen
  i915Ga-PLF motherboard – Intel 915G Chipset; built-in VGA. Reviewed
  at SPCR.
• OCZ DDRAM PC-4000, 512 MB
• Fujitsu
  MHT2080BH 80GB notebook drive in Smart
  Drive (hard drive noise containment box).
• Antec Neo HE 430
  power supply
• Arctic Silver
  Ceramique Thermal Compound
• Two-level open platform with foam damping feet. Motherboard on
  top; most other components below. Eases heatsink changes and setup.

Measurement & Analysis Tools

• CPUBurn
  processor stress software
• SpeedFan
  version 4.27 software to show CPU temperature
• A custom-built variable DC power supply that allows us to dial in
  exactly what voltage is powering the fan
• B&K model 1613 sound level meter
• Seasonic
  Power Angel AC Power meter
• High resolution digital
  audio recording system

Noise measurements were made with the fan powered from the lab DC power supply
with everything else turned off to ensure minimal ambient noise. Airflow
measurements for this heatsink were not made due to the difficulty of measuring
the stock fan accurately.

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 ambient conditions during testing were 16 dBA and 26°C. This is significantly
warmer than the usual lab conditions, and reflect a recent spate of hot weather.
Please keep this in mind when comparing results.

TEST RESULTS

Fan @ 12V: The 40 dBA@1m noise measurement says everything that is
necessary about the CNPS8000 at full tilt. The heatsink is not even close
to usable in a quiet system at this level. The noise was characterized by
a loud whine and a significant clatter of airflow and mechanical noise. To make matters worse, the performance also left much to be desired. With
the exception of the lowly $15 Arctic Cooling Alpine, every other heatsink
we have tested on this test bench was cooler at a lower noise level.

Fan @ 9V: The fan was still too noisy and too whiny to be usable at
9V, and the performance numbers continued to disappoint.

Fan @ 7V: It wasn’t until the fan was at 7V that we would even consider
using the CNPS8000 in a quiet system, when it just barely slipped under the
30 dBA@1m threshold that we consider quiet. Even then, the whine was still
quite prominent, although lower in pitch than before. A significant amount
of turbulence noise was also audible.

Thankfully, performance did not drop much, although we would not want to
cool our 100W processor at this level. A cooler AMD chip would have a better
chance at this level. The small performance drop is somewhat of an empty
victory, since at this noise level most of the other heatsinks we’ve tested
perform as well as the CNPS8000 at 12V. The CNPS8000 cooled about as effectively as the much cheaper Arctic
Cooling Alpine at this noise level.

Fan @ 5V: The fan was finally quiet, though far from inaudible
or silent, and its noise character still left much to be desired. It still produced a deep growling hum.

The maximum temperature of 60°C was not enough to cause our processor
to throttle, although it was on the high side. Embedded in a case
with a higher ambient temperature, it’s unlikely that the heatsink would cool
effectively at this fan speed. Once again, performance was not worth mentioning; at the 23~24 dBA@1m level
of noise, the CNPS8000 was thoroughly trounced by most of the other heatsinks we’ve tested.

COMPARISON

It is impossible to make the CNPS8000 look good by comparing it to other heatsinks.
The table below shows every other aftermarket heatsink we’ve tested on this
test bed. To keep data to a manageable amount, the heatsinks were compared at
approximately 23 dBA@1m, the lowest noise level we achieved for the CNPS8000,
and the highest noise level produced by our reference Nexus 120mm fan. The stock
Intel heatsink was excluded from the comparison, as we do not have a reliable
result for it at 23 dBA@1m. The results speak for themselves.

There is little to be analyzed here. The CNPS8000 more or less ties the Arctic
Cooling Alpine. A different comparison quickly breaks the tie: The Alpine can
be found online for as little as US$6.35, while the current market price for the CNPS8000
is US$40~50.

NOISE RECORDINGS IN MP3 FORMAT

Zalman CNPS8000: 5V-7V-9V-12V, 5s Ambient
between levels: One
Meter, One Foot

Comparatives:

Alpine 64: 5V-7V-9V-12V, 5s Ambient
between levels: One Meter,
One Foot

Scythe Mine: 5V-7V-9V-12V, 5s Ambient
between levels: One
Meter, One Foot

Nexus 92mm fan: 5V-7V-9V-12V, 5s Ambient
between levels: One Meter,
One Foot

FINAL CONCLUSIONS

We have to shake our heads and wonder what Zalman
was thinking. They have years of experience building heatsinks, and they have created some excellent products in the past. But, the CNPS8000 is not one of
them. Not only is the performance below what we’ve come to expect, the fan is also
noisier than any of Zalman’s past efforts — and it can’t be replaced.

It does have good points, but they are the minor points usually
used to differentiate similar products for marketing purposes. It is reasonably small (though larger
than the CNPS7000), lightweight, and installation is simple enough. If this
was an inexpensive heatsink, these might be enough to make it worthwhile, but, as
it is currently selling for US$40~50 range, it is clearly not intended for
the low end. The CNPS7000 is available
for US$20, and can probably one-up its newest sibling in terms of both noise
and performance.

Much thanks to Zalman
for the CNPS8000 sample.

* * *

Articles of Related Interest

* * *

Comment
on this article in our Forums.