Swiftech MCR X20 DRIVE REV3 - Manual (Page 6)

Copyright Swiftech® 2011 – All rights reserved – Last revision date: 12-13-11 – One or more Patents Pending - Rouchon Industries, Inc., dba Swiftech® –151 West Victoria
St., Long Beach, CA 90805 – Toll free (US) 888-857-9438 – Tel. (310) 763-0336, Fax (310) 763-7095 - E Mail: [email protected] – URL: http://www.swiftech.com -
Information subject to change without notice – Page 6
As mentioned earlier however, the consequence of parallelizing cooling devices is that the flow rate inside of said devices is also divided, therefore
slower. So we now need to introduce another concept to further qualify the rationale behind parallelization: the heat flux generated by the different
electronic devices, i.e. the rate of heat energy that they transfer through a given surface.
CPU’s
• Modern CPUs generate a lot of heat (up to and sometimes higher than 200 W), which is transferred through a very small die surface (the die is
the actual silicon, and it is usually protected by a metallic plate called a heat spreader or IHS). Among other things, what it means in practical terms
is that higher flow rates will have relatively more impact on the CPU operating temperature than on any other devices. For this reason, and in most
configurations, the Apogee™ HD CPU waterblock will preferably always be connected in series with the main line, so it can benefit from the
highest possible flow rate.
ALL other devices except radiators
• GPUs, whether they have an IHS or not, also generate a lot of heat (sometimes even more than CPU’s). However the physical size of the dies is
substantially larger than that of any desktop CPUs. The resulting lower heat flux makes GPUs much less sensitive to flow rate. In fact, when both
are liquid cooled, we can readily observe that the GPU operating temperature is always much lower than that of the CPU. For this reason, it is 1/
always preferable to parallelize multiple graphics cards with each-other, and 2/ when one or more GPU blocks are used in conjunction with one or
more other devices like chipset and/or memory, it
is always beneficial to parallelize the GPU(s) with said devices using the Apogee™ HD multi-port
option.
• Chipsets, Memory, Hard Drives and pretty much everything else one would want to liquid cool in a PC can also be placed in the same use-
category as GPUs, either because they have a low or moderate heat flux, or because the total amount of heat emitted by the devices can be
handled without sophisticated cooling techniques. What it boils down to, is that they are even less flow-sensitive, and we submit that parallelization
of these blocks should in fact become a standard.
Radiators
The higher the flow rate inside of a radiator, the quicker it will dissipate heat. For this reason, radiators will always remain on the primary line, just
like the CPU block, in order to benefit from the highest possible flow rate.
In conclusion, we can see that the multi-
port Apogee™ HD when coupled with the MCR Drive Rev3 radiators makes a compelling case for
optimizing complex loops: it maximizes the flow rate where it matters most (on the CPU, and radiator) while offering a splitter-free parallelization of
up to three other components (GPUs, chipset, etc.).
Alternate configuration:
The Apogee™ HD allows an alternate configuration: by using the main outlet as an entry port instead of the inlet, you can then parallelize the CPU
with up to two more components: a second CPU, a GPU, a Chipset, etc. While it remains true as explained earlier that CPUs benefit from higher
flow rate than other components, the few degrees in performance gains might not be consequential to some users. In these situations then, using