Hardware Labs Black Ice GTX 480 Radiator Review

Introduction

Welcome to my pressure drop testing of the HWlabs Black Ice GTX 480. That's right it's a 480 (4each X 120mm fans), and this is the only radiator on the market to hold not three but four full 120mm fans. I would like to give special thanks to Jeremy from Danger Den for sending me this radiator sample for testing. These 480GTX radiators have been very popular amongst the extreme forum members and it was nice to get my hands on two of these beauties to step up another notch in my own system.

With this review I will initially focus on pressure drop testing as well as thermal performance. First off, let's take a look at the size comparison of the 480GTX next to the PA 120.3. They are both extremely thick radiators, and the 480GTX is simply one more fan long. I'm a strong believer in recommending as much radiator as you can possibly fit. When it comes to water cooling, the biggest advantage comes with the ability to dissipate heat over a much larger surface area than you can with conventional tower air cooler, so you might as well go BIG and the 480GTX does just that!...It's HUGE!!

Radiator Characteristics

Ultra Thin 19.0mm x 1.2mm Tubes

Front to Rear Two Pass Flow

Compact Core High Density Fin Structure

HWlabs also has a higher density fin structure optimized for medium to high speed fans (although you'll see "Slow Speed" fans do well also). It has a couple of unique features. Fist the "Folds Per Inch" is on the order of 20 folds as opposed to the 11 of the PA120.3 and Fesser fin densities. Also it has a compact core in which is uses another layer of copper sheet between the folds to allow double rows of fin folds in between each set of tubes. In addition the GTX fins themselves are not louvered, instead it is relying on the density of the structure for turbulence and heat transfer (Less dust getting trapped in louvers). All of this is purposely part of the design to maximize the total surface area and the contact area between the tubes and fin structure. This added heat transfer area means it should be able to exchange more heat with less air. Below is a fin comparison of pictures at the same scale for the 480GTX, PA120.3, and a prototype of the new Feser X-Changer.

Hydraulics and Pressure Drop Testing

The most scientific way to determine a blocks hydraulic resistance is to test pressure drop. Pressure drop is a measurement of pressure loss across a radiator that varies with flow rate. This is basically a measurement of energy loss, and directly influences how much flow rate you will have.

Equipment:

Dwyer Digital Manometer 477 Mark V - Accuracy .5% of Full Scale. Range 0-20.00 PSI range, Resolution .01 PSI
King Instruments 7520 Series 0-5GPM, 250mm scale - Accuracy 2% of Full Scale. Range 0-5GPM, Resolution .1 GPM (can be interpolated to .02GPM)
Water Source - Household water pressure - 50PSI at >5GPM - Because flow rate readings are instantaneous, household tap water and water pressure are a good and powerful source for pressure drop testing.

It was in interesting test, and one that I completed three times just to make sure over two radiators and different directions. Comparing to other radiators it would fit into the "High Restriction" category, but in comparison to your average CPU block, I wouldn't consider this a huge issue. At 2 gallons per minute it compares reasonably similar to a D-Tek Fuzion with washer and no nozzle which isn't considered a highly restrictive CPU block setup. I would just recommend using a stronger pump like the Laing DDC or D5 pumps and ensure you maintain at least 1 GPM. The "interesting" part about it besides being more restrictive is how linear the results are, most pressure drop curves have a much more pronounced "curve" to it. My theory is the flat and thin tubes inside the radiator are expanding a very small amount as pressure climbs which lowers the restriction. More pressure, more tube expansion, less restriction creating a relatively flat curve. So as expected it's a bit on the restrictive side, now we just need to see if this tradeoff was worth it for better thermal efficiency.

Thermal Testing Specifications

Thermal Testing

As expected the large size has this radiator producing leading results particularly in the slow, medium, high, and ultra high fan scenarios.

You can use the above graph to estimate your water temperature. To estimate a heat load you can use the following tools:

Extreme Outer Visions PSU calculator - Just note heat load minus the 38 watts included in the start.
Swiftech's Conroe/Kentsfield CPU calculator - This is actual an estimator for the kit, but it also gives what I think is a more accurate heat load for the processor. Just plug in the field and it will output a heat load in watts.

About heat load - Remember tha heat load is the actual heat going into the water and it's not equal to power consumption. There is a portion of the power consumption that will get lost into the air through the motherboard, cards, etc and it's also going to depend on what you consider load. Also for GPU only block, you wont be dumping all that memory heat into your water, so you'll have again a fair amount lower heat entering the coolant than those that use full coverage GPU blocks. With that I would warn you that most of these PSU calculators probably are giving you a fair amount higher heat load than you really have. How much error do the PSU calculators have? Good question, but I wouldn't be suprised if they are 20-30% higher than actual. I may run some experiments on some of the hardware I have on hand, but it's unlikely we'll ever have a good means to estimate heat load accurately. That's ok though, you can use the power consumption numbers to design from and use it as a cushion.
Now you can take your heat load in watts and follow it up vertically to the fan scenario. When you cross that line, turn left and follow it horizontally until it impacts the vertical axis on the left. This temperature delta will be your number that you need. To estimate your water temperature simply add this number to the ambient.

So what does this mean? You are cooling with the water that leaves your radiator and enters your CPU block. So any improvement in this number will directly translate into an improvement in your CPU temperature. For example if your water temperature delta is 10C and your loaded core temperature was 50C. If you improved your water delta by 5C either by adding another radiator or increase your fan performance, your new core temperature will be 45C assuming your ambient temperature remained the same.

I would recommend if starting out that you shoot for a 5C delta for extreme performance, even though it's common practice for radiator manufacturers to list 10C delta curves...I believe 5C is a better target for extreme performance. A few years ago a 10C delta was still a very good gain over air cooling, now I think it's better to shoot for 5C to really make the most out of water cooling's advantage. After all, the real benefit is having the ability to dissipate heat from an unconfined space. Air coolers will always be limited to the space around and above the CPU socket. The size of the radiator or series of radiators is only limited by the user's personal preference...go big if you want big performance and include several radiators in the same loop, nothing wrong with more heat dissipation especially if you have a heat load that needs it. My suggestion is to target a 5C delta for extreme performance or a 10C delta for average performance.

Water Temperature Estimator File

How does it compare to the PA120.3

At 1120RPM (Between the Ultra Low and Low Speed) both radiators perform roughly the same. Above that the 480GTX dissipates more heat, below that the PA120.3 dissipates more heat.

Obviously the 480GTX has a size advantage here(4 fans vs 3 fans), and it is producing better thermal results for most of the fan scenarios. The PA120. 3 does however flex it's muscles in the ultra slow speed category where it actually outperforms the bigger quad even with one fan less. With that said, it's still far from saying the 480GTX is only good with high speed noisy fans. Yate Loon rates it's D12SL12 as a "slow speed" as designated by the "SL". This slow speed fan runs at 1350 RPM +- 10%. If you compare the two radiators and select one lower fan selection for the 480, you can almost get an equal performance. For example the 480GTX running slow speed yate loons at 1350RPM can dissipate about as much as the PA120.3 with medium speed fans. While you do have one more fan making more noise, that's a notable benefit especially if you compare 3000RPM compared to 2000 RPM Ultra Kazes. The 480GTX is obviously optimized with it's higher fin density to make the most out of stronger fans, however the larger size (quad vs. triple) allows it to continue it's lead clear down to the 1120 RPM mark where they are equal.

The last point I wanted to make with this is a normal review could have only tested one fan type. Had they tested at 1000RPM only, they would have said the PA is better and could have misled you into thinking that's consistent with all fans or for everything but high speed fans. This is where the "Scientific" method helps sort out the trend and give you the full picture. We really have two winners here, each in their own respective areas. More so I hope I can provide you with something you can use to design your system and help you see the substantial influence fan selection has. Obviously this again compares a quad sized rad to a triple which many feel is not right, but I think that depends on your space circumstances. If you can fit a quad sized radiator and have enough budget to afford either of the above two radiators, I belive the comparison is very useful and presents the bottom line well. If however you have only enough space for a triple sized radiator, you can't expect the same performance as the top of the line 480GTX. It should be approximately 25% lower although testing would be needed to really sort out that difference.

More importantly, I can't stress enough how important fan selection plays a role here. For the 480GTX you get nearly 5X the performance using the high speed Ultra Kaze 38mm fans at 3000RPM than you do with the yate loons at 1000RPM. Fan selection is EXTREMELY INFLUENTIAL on performance.

How does it compare to the PA120.3 with some estimated heat loads

Ok, so c/w ratios are fine, but what does that mean in terms of degrees difference you might ask. To estimate that, I've selected several random heat load scenarios that might paint that picture a little better. Here they are:

CPU only loops - If you want extreme water-cooling performance you'll see that both radiators can do so admirably, and the amount of heat generated by a single cpu is small enough to afford you very very small deltas even with slow speed fans you can get down below a 3C delta. More importantly you can see in a CPU only loop, the degrees gained by going to much higher speed fans is much smaller. I would question the value of using the very aggressive and loud 3000RPM Skythe Ultra Kazes over silent 1350RPM yate loons on a dual core processor for only a 1.4C gain...etc. Either way you can see the difference between the two radiators is very small when you are looking at small heat loads, they would both do extreme performance under 4C with just 1000 RPM Ultra Low Speed fans.

CPU and GPU loops - This is where fan selection becomes more critical and also where the larger size of the 480GTX really starts to show it's strength. In a quad core and SLI setup, I would probably recommend the 480GTX with a 2000RPM 38mm fan similar to the Ultra Kazes, that would afford you about a 6 degree delta which is very good. That's simply too much heat to use a slow speed fan on either radiator.

Ok enough with that, what I wanted to really point out is not so much the differences between the PA120.3 and 480GTX, but that heat load and fan selection again are very very important. With really small heatloads like a CPU only type of setup, you're really not going to see much of any difference between the rads or between the fans selected. This is where "Silence is bliss" you can have extreme performance water temperatures without extreme noise levels, for those that want both...you need ALOT of radiator and either one will do just fine with slow speed fans even on a really hot quad core keeping water in the 5C extreme performance mark or close to it. Second I wanted to show how fan selection becomes extremely important as the heat loads rise. Even though the PA120.3 is optimized for ultra slow speed fans, putting a 570watt heat load on it with only 1000RPM fans isn't going to perform acceptably well, you will need to consider stringing several PA120.3's together with that amount of heat, OR use stronger fans...the option is yours.

If you are trying to understand how the 360GTX compares to the PA120.3, please don't use the above comparison as a reference. You will still see each radiator more optimized each area, but the crossover point will be different and obviously higher with the smaller radiator size.

So the amount of heat and the fans used are BOTH critical to your decision around the radiator component. Small heat loads are easy, high heat loads need strong fans or multiple radiators. I would suggest for extreme performances that you shoot for a 5C delta, the lower the better.

Pros

KING OF HEAT DISSIPATION!, capable of dissipating over 800 watts with 3000RPM fans at a high performance 5C water/air delta, that is some heat!!

The most thermally efficient single radiator tested to date on Slow, Medium, High, & Ultra High speed fans.

High Quality Paint gloss finish

Quality construction - Extremely high attention to detail and the highest quality present in finishing.

Standard G1/4 threading - I would recommend Danger Den's FatBoy Barbs (My favorites and used for all testing)

Non-Louvered fin design will not trap dust internally as much

Low cost compared to the competition

Cons

Size may be challenging to fit internally on many cases without significant effort. (Time to break out the dremel!)

Thickness, double radiator thickness may also cause installation difficulties.

High hydraulic restriction (pressure drop) reduces system flow rate slightly. (Check to ensure 1GPM is maintained)

High fin density may require more dust cleaning maintenance (although fins are not louvered, so better internally)

Not the most thermally efficient single radiator for Ultra Low Speed fans less than 1120RPM

Hardware Labs Black Ice 480GTX Radiator Full Review