Xigmatek Dark Knight 1283V CPU HSF review

[K404]

Xigmatek Dark Knight CPU HSF Review

Intro.

Xigmatek are a relatively new name to the world of PC Hardware cooling but have built up a wide range of products and have been given positive reviews by reviewers world-wide. For a new company, availability is good, which is no small feat- especially in this economic downturn.

Xigmateks focus is on aircooling, where they have no less than 13 CPU coolers. In addition to quantity, they also have innovation- Their HDT (Heatpipe Direct Touch) design allows for the best possible thermal transfer from CPU to heatsink, which should provide that last scrape of stability when running on the edge of what aircooling can offer- especially useful for high heatloads like Intels I7 chips.

On the test-bed in this review is their Dark Knight (1283V) CPU cooler.





The Dark Knight

This cooler comes in a conventional box-y box with a perspex window to show off the fan. Artwork includes a sword and a picture of the complete HSF unit. Core I7 compatability is explicitly mentioned.












The box contents are: the heatsink itself, a 120mm Xigmatek fan, 4 springs, 4 threaded bars, backplate + insulation, AM2 crossbar, 2x mounting arms for Intel solutions, 5 fan mounting grommets, an instruction leaflet and some thermal paste.
















The HSF comes as standard with a sprung bolt mounting system that replaces the twist lug method used in Intels stock retention mechanism. An electrically insulated backplate is included as a screw-down base. The heatsink is tightened using four Phillips screw heads. The springs are fairly tense- not too much effort will be required to tighten them, although some effort is needed to push the threaded section down far enough to make contact with the backplate. It's still far preferable to Intels stock approach though.







The fan is a smokey, transluscent colour. A mix of dark brown and grey. There are 4 white LEDs in the frame that give the unit a unique look when powered up. At full speed its... noticeable. For an overclocking session, id consider it quiet compared to many dedicated fans. For gaming it will also be fine but for watching DVDs or office work, it might annoy slightly, depending on the mobos implementation of E1ST or CnQ

[K404]

Regarding mounting

Bare HSF dimensions: 120mm wide, fin height of 110mm and depth 50mm.

Distance from mobo to bottom fin:
775: 39mm
1366: 38mm
AM2: 50mm

For I7 and 775 mounting, the cooler can be set up either horizontally or vertically- both orientations have points to be aware of.

For the horizontal setup, theres a high risk of the cooler not fitting around RAM or PWM cooling. For vertical attachment (more popular,) the fan must be pushed FULLY onto the rubber mounts, else, depending on the mobo layout, there might be problems with giving RAM cooling solutions enough space (as demonstrated below)

Additionally, for vertical mounting, if the CPU socket centre is less than 60mm away from the edge of the board, the heatsink will overhang the edge, which might be a problem depending on the case used and the space available.

For AM2 mounting, because of the size of the crossbar latch, it has to be positioned on the side opposite to the fan, else the fan will not fit.

The most important thing needing said about AM2 mounting is that due to where RAM slots are typically placed in relation to the CPU, RAM with tall heatspreaders will probably NOT BE COMPATABLE with these coolers. As pictured below, there is a hefty overlap from the fins. I suspect this will affect most HSF combos, so Xigmatek should not be heavily penalised for this, but a scalloped design would give them a marked lead over most of the competition for AM2 support.

[K404]

Test setup, notes, assumptions + drawbacks


CPU+ Mobo combos:

s775: Intel C2D E86 + Asus X48 P5E64 WS Evo

s1366: Intel I7 920 C0 + DFI X58 EH6 w/ 3x2GB Crucial Ballistix

s940: AMD 5600+X2 + XFX 8200 w/ 2x1GB Crucial Ballistix

The AM2 setup is my 24/7 computer. The components picked for that are not really great overclockers.

The I7 setup had settings picked to keep the RAM and Uncore MHz fairly consistent. Uncore was 3200-3300MHz, RAM was kept at 800-825MHz. RAM timings were 8-8-7-1T, 1.65v vTT was 1.44v

The I7 mobo could not set vCore lower than 1.00v. 1.02v was the lowest voltage used.

The I7 was not tested with a stock cooler.

The E86 rig was not tested at stock. Its a sub-zero setup and is "tuned high" for competitive bench runs.

All tests were done on an open bench rig. I appreciate this will have a definite effect on temperatures, but the plus side of this is that other heat sources have less of an effect on CPU temperatures due to less thermal buildup, so results do not become skewed.

This will be the most time i've spent with aircooling since I moved to 775 in summer 2006. My test methodology will probably not measure up to that of the hardcore aircooling guys. I do expect it to be good enough to provide meaningful, comparable information.

A K-probe was tested for a hardware-based temperature result but the results were considered very inaccurate and were not used. The distance between the probe/IHS surface and the core- a distance that could not be reduced- was considered to be the main culprit, so I had to rely on software readings.

OCCT was used for generating a consistent heatload. The test was ran for 30 mins with a large data set under normal priority

Speedfan was used for the software-based temperature readings for AM2 while CoreTemp was used for the Intel setups. The OCCT graphs generated for Intel gave consistently lower temperatures than CoreTemp by typically 5 degrees. I have made reference to both sets of data.

For plotting the temperature data, the worst-case scenario data was used. ie. the lower ambient temperature recorded and the higher CPU temperature recorded.

Arctic Ceramique was used for all testing.

Because of the mixed ambients encountered during testing, all the results have a second data interpretation in terms of degrees Celcius above ambient. I felt this was necessary to prevent the results becoming skewed. However, the graphs are all in terms of the temperatures recorded.

[K404]

Test Results





The BIOS only went up to 1.45vCore and I had no real interest in pushing the board to its limit

[K404]

AM2 OCCT Graphs

Stock Cooler:

(L-R: AM2 stock cooler @ 2.9, 3.0, 3.1GHz)


Dark Knight:


(L-R: AM2 with Dark Knight 2.9, 3.0, 3.1, 3.2GHz)




Intel 775 OCCT graphs


(L-R: Intel stock cooler @ 4.0GHz, Dark Knight @ 4.0, 4.1, 4.2GHz)

[K404]

Intel I7 OCCT Graphs:

(L-R: I7@ 2660MHz, 3.3, 3.4, 3.5, 3.6GHz)


Conclusion.

The HSF looks great, the mounting mechanism is easy to work with and the temperatures are comfortably lower than stock. For AM2 setups, from a performance point of view- a quality aftermarket HSF is maybe going to waste as the AM2 just doesnt kick out enough heat for a new cooler to be worthwhile. Either that or the stock HSF is actually decent! For 1366 and 775 setups- the Dark Knight is a good HSF for squeezing a better overclock, without doubt.

My thanks to Tony @ Xigmatek for the review sample and the other goodies he supplied as well

Pricing: coming in at £37 on Scan: http://www.scan.co.uk/Products/Xigma...-Intel-AMD-CPU Its by no mean the cheapest cooler on the market but the looks, ease-of-mounting and performance make sure its worth serious consideration.

[K404]

Some afterthought testing


All the results in the original review were achieved with the fan set up to push air through the heatsink. I reversed the fan so it pulled air through the heatsink and got a 2 degree drop in temperature.

Pushing air over


Pulling air though

[Nathanial]

Nice review as always thats a cool ass looking cooler, Xigmatek are going to be biig this year i can see it .

[vento1]

Great review Kenny, very nice looking and performing cooler.

[GeorgeStorm]

Nice review
Yet another solid product from xigmatek

[4Qman]

Ken this is a dam right awesome review. You have covered it all mate and its spread out great.

I hope this is now the benchmark for all our reviews, something which shows data and a quality product.

Thanks

[maverik-sg1]

Nice review Kenny mate - seems to be a nice cooler.

Did I miss the estimated street price?

The plating on the heatpipes looks gash (poor finish with blisters) btw

[K404]

Originally Posted by maverik-sg1

Did I miss the estimated street price?

Nice catch Mav Now added to the conclusion


Thanks for the comments guys

[borandi]

Awesome review. The HSF looks identical to the Akasa Nero though?

[K404]

Couple new screenies added to post 7. I swapped the fan direction to see what what happen. A temperature drop happened