360 Radiator Shootout Summary

As part of my “In-progress” 360 radiator testing, I’ve started this summary post where I will continue to update as radiators are tested.

SPONSORS

A special thanks to my many sponsors:

Danger Den sponsored the Hardware Labs GTX 360 and Hardware Labs SR1 360 radiators as well as the MC-TDX waterblock.

Aqua Computers sponsored the Aqua Computer’s flow meter.

Swiftech sponsored the MCR320 QP(Quiet Power), MCR320 XP (Extreme Performance) radiators as well as the MCP-35X2 pump.

XSPC sponsored the XSPC RS360, and XSPC RX360 radiators.

Aquatuning sponsored the Alphacool NexXxoS Pro III 360, Alphacool NexXxoS XT45 Full Copper 360, Alphacool NexXxoS XT60 Full Copper 360, and Phobya G-Changer 360 radiators.

EK sponsored the EK Coolstream XT360 and Coolstream XTX360 radiators

Coolgate sponsored the Coolgate Single Row Acrylic 360, the HD 360, and the triple row 360.

Test Bench

Restriction

Generally most radiators are low in restriction relative to your average and even lower restriction CPU block.  Also in general the double thickness radiators usually have the lowest restriction due to the additional tubes where water can flow, but there are some exceptions such as the GTX which uses thinner than typical tubes.

Thermal Performance

Conclusion

Well 16 radiators done and as usual with most watercooling, the differences between radiators is also very small.  When only 6 watts separates the top 11 radiators at 600RPM, testing error alone does become a problem.  On the little bit of error checking I’ve done on a few radiators I am seeing in the neighborhood of 2-3 watt error present in my trendline results.  You should also consider the scale of those 6 watts when a very small thing like switching out your pump from a D5 to DDC could likely yeild you that 6 watts reduction in heat.

.

With that I would say generally the thermal performance differences are relatively small, but there are some generalities:

• Copper Tubes are better than Brass Tubes
• Thicker radiators generally can be tuned for a broader range of RPMs than thinner
• Thicker radiators can perform better than thinner radiators, but the difference is not more than 5-10% typically
• Flat tubed radiators perform better than round tubed radiators
• Slim 30mm radiators can be tuned to perform extremely well in specific areas, but generally that means loosing more in the opposite area (ie a high speed tuned slim radiator won’t perform well at slow speeds)
• Build quality does vary and here you do get what you pay for.
• In performance, you don’t get what you pay for.

Bang for the Buck

I’ve always been one for emphasizing value, and with results this tight, the lower the cost the more bang for the buck. The XSPC RS360 being the lowest price rad holds a very high benefit/cost across the board.  While not quite as cheap in price, the Swiftech MCR320XP gets high marks for being more value oriented and maintaining awesome performance in the mid/high speed ranges.  The Alphacool ST30 is similar in being an amazing value rad, but optimized for slower speed fans.

• Slow Speed Value = Alphacool ST30
• All Around Value = XSPC RS360
• High Speed Value = Swiftech MCR320-XP

Top Performance ($ is no object)

•  Slow Speed Performance = Alphacool ST30, HWlabs SR-1, and the neighboring top 6 are all really really close.
• All Around Performance = Alphacool UT60 & EK XTX are both really good all around performers
• High Speed Performance = Swiftehc MCR320-XP, Alphacool UT60, and the EK XTX are all great higher speed performers

Best Build Quality

• Hardware Labs GTX & SR1
• Aquacomputers AMS
• Several other close runners up.

Most Useful Feature Rich

• Alphacool UT60,XT45,ST30 & Phobya G-Changer for having that screw protector feature.  Also extra ports and bleed screw on the UT60.
• Aquacomputers AMS for the modularity features.

Most Unique

• Aquacomputers AMS for it’s solder free and round tube construction and bare copper core
• Coolgate Copper Radiators for their copper plating finish and the Acrylic for it’s clear end tanks.

In the end there is no one size fits all, but there are plenty of very good options and they all perform very closely to one another.  To say one wipes the floor with another is simply false when typically we’re talking single digit percentage differences and at most 20-30% difference.  Don’t expect that you can upgrade any one core size and see dramatic measurable improvements, you have to consider the finer details and want a change in areas outside of performance.

Cheers!
Martin