Heatsink test – Joe
SUMMARY: An excellent choice for low-noise, very good P4 cooling; looks like a Prescott Pacifier.
The good guys at Thermalright were nice enough to send an engineering sample of their 120mm P4 heatpipe heatsink for a test spin. This unit will be available only for P4 cooling. In contrast to some heatpipes I have tested, the heatpipes are soldered to the copper base – some used thermal paste instead, with performance suffering as a result.
Mounting is straight forward – there are two clips which engage the P4 retention bracket. The preferred mounting procedure is to engage the two clips at the back end of the heatsink and then rock it forward toward the RAM slots to engage the retention bracket.
This is a large heatsink – to give some idea of its footprint, the following pics show it mounted on a motherboard. Looking from the top down
shows the overhang over the motherboard; I outlined in red the heatsink’s base. While it may look as if it interferes with the first DIMM slot, the pipes angle up to clear it. Looking edge on
shows the overhang over the edge of the motherboard. A view from the top edge of the motherboard
shows the fit among the motherboard’s components. Note that there will be beneficial cooling effects due to this overhang on components surrounding the socket – most notably, there may be secondary cooling of the Northbridge if passive cooling is used on it.
The Thermalright 120mm P4 Heatpipe was tested on an Acorp 4S845A motherboard with a modified P4 1500 to read CPU case temps. For testing, I used two fans – a US Toyo #USTF1203812HW 38 mm unit and a Delta 25mm #WFB1212M; unfortunately the TOYO died (I was radically overvolting it), so I can’t report noise levels on it. I varied rpms on both to give a performance range.
| Motherboard | CPU Case Temp | Ambient Temp | Delta | C/W | MBM Temp |
| Delta 998 rpm, <50 dBA² | 37.6 C | 25.7 C | 11.9 | 0.21 | 34 C¹ |
| Delta 1502 rpm, <50 dBA² | 36.2 C | 25.4 C | 10.8 | 0.19 | 32 C¹ |
| Delta 1966 rpm, 54 dBA² | 35.2 C | 25.1 C | 10.1 | 0.17 | 30 C¹ |
| Delta 2503 rpm, 59 dBA² | 34.6 C | 25.1 C | 9.5 | 0.16 | 29 C¹ |
| TOYO 420 rpm | 42.5 C | 26.5 C | 16.3 | 0.28 | 36 C¹ |
| TOYO 1333 rpm | 35.6 C | 25.0 C | 10.6 | 0.18 | 31 C¹ |
| TOYO 2003 rpm | 34.2 C | 24.4 C | 9.8 | 0.17 | 29 C¹ |
| TOYO 2502 rpm | 33.9 C | 24.4 C | 9.5 | 0.16 | 29 C¹ |
| TOYO 2945 rpm | 33.6 C | 24.5 C | 9.1 | 0.16 | 28 C¹ |
I thought it would be interesting to compare the 120’s performance to a similar heatpipe – Thermalright’s SP94. I tested this with an 80mm Delta fan – below is a graph comparing performance levels:
I should note that in my tests, 50 dBA measured 8″ from the fan intake corresponds to about 30 dBA measured 3 feet from the fan; this is a very quiet noise level. The 120mm fan at 1500 rpm is quiet, and below that it’s about inaudible.
The graph shows that the 120mm fan flattens out the noise-C/W profile rather nicely. Speeding up the 120mm fan to rpms above 2500 rpm will increase noise to uncomfortable levels, defeating the purpose of using the 120mm fan. For reduced noise at very good performance levels, the 120mm is a clear winner.
Die Simulator results place Thermalright’s 120mm P4 Heatpipe in the topmost rank of P4 heatsinks tested to date (Heatsink Ranking).
Thermalright’s 120mm P4 Heatpipe is well made and an excellent choice for low noise P4 cooling. Mating Thermalright’s Heatpipe with a variable speed fan might be the preferred approach for the power user looking for peak performance followed by a quiet night’s sleep. Prescott users might want to consider this heatsink for effective cooling.
Thanks again to Thermalright for sending this our way.
The Thermalright Heatpipe will be available from The Heatsink Factory; if this is of real interest to you, you might want to reserve one early.
²Note that manufacturers measure fan noise usually 3 feet from the fan.
Be the first to comment