Xigmatek HDT-S1283 - heatpipes for extreme loads

halcyon · Post by **halcyon** » Thu Jan 24, 2008 4:51 am

Frostytech has reviewed Xigmatek HDT-S1283 heatpipe heatsink for K8/LG775 CPUs.

http://www.frostytech.com/articleview.c ... 233&page=5

The results for 150W and 85W loads with low fan speed are quite promising.

Anybody tried this?

Bluefront · Post by **Bluefront** » Thu Jan 24, 2008 5:46 am

Impressive....if their testing method translates the same in a real-world test. It actually cools very well at the lower fan speed.....indicating it might do well in a very low airflow setup. The fins do look close together though. Worth more testing/reviews for sure.

Felger Carbon · Post by **Felger Carbon** » Thu Jan 24, 2008 11:37 am

Bluefront wrote:Impressive....if their testing method translates the same in a real-world test.

Frosty's testing procedure has two gotchas: first, they use a heat source that's the same size as the IHS. This causes spectacularly optimistic results on HSFs with aluminum CPU blocks such as the Arctic Cooling Alpine 7 and the Evercool Buffalo. With copper blocks, the results are merely optimistic, but comparisons are OK - the results are equally optimistic.

The second problem is limited to AMD heatsinks. Frosty does not alert the reader to which HSFs must vent hot air into the PSU instead of the exhaust fan on 98+% of the AMD mobos.

Frosty does allow the occasional typo to creep in, and never fixes it. Such as reporting the same fan noise for high and low fan speed, despite a notable difference in temperature. But there aren't a lot of these.

Relleum · Post by **Relleum** » Thu Jan 24, 2008 12:59 pm

Well this sucks badly.

I just bought the Noctua NH-U12P from accusticpc.com for 59.99 plus 7 shipping. And this is AFTER I spent hours (or days) researching on which cpu cooler would meet my top requirement, silence -- and still be able to knock the temps down as low as possible.

Somehow I missed this Xigmatek HDT-S1283, which cools better than anything else out there (including the mighty TRUE), comes with a PWM connector (My noctua has a 3 pin), AND costs around 45 bucks shipped. That is over 20 bucks cheaper than my new noctua, for better performance and nearly the same accoustics at low speeds with the adjustable PWM connector.

And now I'm faced with non-refundable shipping charges, a 20% restocking fee, and more shipping charges to send it back. I am basically screwed!

--Relleum

Relleum · Post by **Relleum** » Thu Jan 24, 2008 1:06 pm

Oh, and I almost forgot to mention that it comes with a mounting system that doesn't require the motherboard to be removed.

I made the wrong choice.

Felger Carbon · Post by **Felger Carbon** » Thu Jan 24, 2008 1:53 pm

Relleum wrote:Somehow I missed this Xigmatek HDT-S1283, which cools better than anything else out there

It does?

If you're talking about the Frostytech review, and if you agree that the measure of good cooling is a low temperature rise combined with a low fan noise, the the Silentator whomps the S1283 something fierce (for example), as I interpret the data.

Entropism · Post by **Entropism** » Thu Jan 24, 2008 1:56 pm

I just installed the Xigmatek, using the thermalright bolt through kit, and have a 5v Yate Loon on it. It lowered my temps drastically over the Ninja I had installed before, same bolt kit, same exact fan. Definitely some quality stuff, though it might not be the BEST, it's definitely a TRUE competitor for $40 shipped.

Felger Carbon · Post by **Felger Carbon** » Thu Jan 24, 2008 1:59 pm

Relleum wrote:Oh, and I almost forgot to mention that it comes with a mounting system that doesn't require the motherboard to be removed.

Directron (for one) carries the Revoltec direct OEM equivalent HS for $30. I read a review of the Revoltec yesterday and the review pointed out that it was very very difficult to install the HSF on a built-up computer. It was highly recommended by the reviewer that the mobo be removed. The difficulties were described in detail.

There's a lot of data out there. It might pay you to do a little research before pulling the trigger or pushing the abort button.

edit: you don't have to look too far. On this very SPCR forum and page, you'll find this info on mounting. This Kingwin RVT is identical except for fan and box labels to the S1283.

omgwtf · Post by **omgwtf** » Thu Jan 24, 2008 2:00 pm

This cooler look like... Pentagram HP-100 AlCu Lite. Heatpipes are directly to processor, like in Pentagram...

Luminair · Post by **Luminair** » Thu Jan 24, 2008 8:37 pm

Relleum wrote:Well this sucks badly.

That is excessive! Ninja, Ultra 120, these other mentioned... all of these high end coolers are the top of the top, the cream of the crop. They are in fact borderline overkill even on 3ghz Q6600s.

And from those frostytech numbers, the variation between the top products is incredibly small. Fractions of a degree. Test error could be greater than that.

Using one over another certainly doesn't suck badly.

burebista · Post by **burebista** » Fri Jan 25, 2008 12:52 am

One of our Xigmatek tests on S-1283 and family. It's in romanian but it has some pictures.

All I can say is that I'm impressed by performance/price ratio. Performance wise it's on par with a lapped TT BigTyphoon.

Bluefront · Post by **Bluefront** » Fri Jan 25, 2008 3:29 am

Humm.... " it's on par with a lapped TT BigTyphoon". If that's as good as it gets, I'm not impressed. My Ninjas easily beat my one Big Typhoon in their cooling ability. Hard to figure....

burebista · Post by **burebista** » Fri Jan 25, 2008 5:07 am

Bluefront wrote:My Ninjas easily beat my one Big Typhoon in their cooling ability.

I own a Ninja too on my X2 and it's awesome, but try to cool a E6850 @4.4GHz 1.5xxV Vcore with it.

And here Ninja is $56.3 and S-1283 is $45.4 so that's I said that I'm impressed by performance/price ratio.

Gudas · Post by **Gudas** » Sat Jan 26, 2008 3:17 pm

I decided to swap my TT BigTyphoon (also lapped) with this cooler.
It is a little better now, the hot air don't stop now in the case (TT BigT - hot air has concentrated in the area between PSU and 5,25" drive bay) and the temps are a little lower. I think a bit lapping will bring some small diferrence.
But the fan that comes with it is loud (for me), i use now Noctua NF-S12 @ 7V - the temps are higher but i cant hear it now

Relleum · Post by **Relleum** » Mon Jan 28, 2008 4:22 pm

Well, after installing my expensive noctua, my temps dropped 15C in idle, and 22C drop in load. Top that off with the fact that my comp is ridiculously silent, and I really can't complain!

I might have paid a bit more for slightly less efficiency, but at least I'm satisfied and happy with the performance.

--Relleum

celeron · Post by **celeron** » Mon Mar 10, 2008 11:39 pm

What kind of temps are you people getting compared to scythe infinity? I'm thinking about changing to HDT-S1283. Reason for that is infinitys almost impossible assembly/disassembly with gigabyte S775 mobos. Especially 965P-DS4.

Post by **MikeC** » Tue Mar 11, 2008 12:55 am

Felger Carbon wrote:Frosty's testing procedure has two gotchas: first, they use a heat source that's the same size as the IHS. This causes spectacularly optimistic results on HSFs with aluminum CPU blocks such as the Arctic Cooling Alpine 7 and the Evercool Buffalo. With copper blocks, the results are merely optimistic, but comparisons are OK - the results are equally optimistic.

The second problem is limited to AMD heatsinks. Frosty does not alert the reader to which HSFs must vent hot air into the PSU instead of the exhaust fan on 98+% of the AMD mobos.

The other problem with the frostytech test platform is that it probably allows for virtually "perfect" mounting pressure -- they never mention how the HS mounted, check every review. Perhaps this is because often they cannot use the stock HS "clip".

Look at the pics in their test methodology article (which never mentions installation of the heatsink either -- other than TIM). The AMD platform appears to have only the stock heatsink retention bracket, while the 775 socket has threaded screw inserts. So how do you think heatsinks get mounted on these? Think about the range of HS mounting schemes with today's HS -- no way you could use the supplied "clips" on a whole lot of HS with these platforms.

Am I missing something?

oh and here, btw, is a somewhat different assessment of the same HS -- http://overclockers.com/articles1503/

FartingBob · Post by **FartingBob** » Tue Mar 11, 2008 3:34 am

Seems to do pretty well. Not the best around, but certainly in the top grouping. Anybody compared this to the TRUE or ninja with the same fan to directly compare them?

vaio · Post by **vaio** » Tue Mar 11, 2008 3:35 pm

Should you not be holding it up in comparison with it's competitors in the same price range?

For example, wether or not it beats out the TRUE is moot given the major price differential.

I am not advocating the Xigmatek, just trying to keep perspective.
This is a circa Â£20/$40 sink so to even be mentioned along with the "big guns" is notable.

Tkpenalty · Post by **Tkpenalty** » Mon Mar 17, 2008 4:31 am

I'd make a beeline for the Xigmatek-in fact I have it. Suprisingly enough, it runs pasisvely, still outdoing the TRUE. I've owned the TRUE and HDTS1283, and the xigmatek was almost half the price of the TRUE.

I have to say the days of the TRUE are almost truly over.

With the Xigmatek, the stock fan is pretty loud to be honest, but with the PWM, any case with decent steel thickness will render the cooler inaudble. And performance still kicks the TRUE. If the stock fans are too loud... then replace them! On this note replacing the Xigmatek's stock fans will still make the combo cheaper in total.

Guys, note that some review sites are very dodgy with their results, only testing it once and drawing a conclusion.

I'd only use frosty tech to draw conclusions for the performance of coolers. And so far they have been extremely accurate. I got virtually the same temp drop with the xigmatek vs TRUE.

Its natural that the Scythe coolers are inferior, with their super thick copper bases in the way.

halcyon · Post by **halcyon** » Tue Mar 25, 2008 12:58 am

Some more reviews:

Relative performance against Thermalright Ultra 120 Extreme at low/high fan speeds
http://www.xbitlabs.com/articles/cooler ... html#sect0

Load temps with overclocking against watercooling kits
http://www.techpowerup.com/reviews/Xigm ... ler/4.html

Against Samurai, Noctua U12F, Infinity, Mine, Thermalright 90 Ultra...
http://www.technic3d.com/article-550,4- ... bruder.htm

Against Thermalright IFX-14, Ultra 120 Extreme, Infinity, Ninja CU, Zipang, Samurai, Ninja mini, etc. at 12/5V fan speeds
http://www.pc-experience.de/wbb2/thread ... adid=25455

Against Orochi, Silenator, Samurai, Noctua U9 with Quad core loading compared to acoustic load at various fan speeds
http://www.au-ja.de/review-cooler2007qxee-19.phtml
http://www.au-ja.de/review-cooler2007qxee-21.phtml

On a related note, it is interesting to see how Orochi can perform very well or quite badly, depending on fan speeds/load/fan combinations.

It is very difficult to compare the performance with other coolers, due to changing test setup / methodology, but the above shed some light on the performance.

Having said all that, I think the heatsink looks very promising and I've selected it for my next build (if it fits my mobo).

VanWaGuy · Post by **VanWaGuy** » Tue Mar 25, 2008 3:03 am

Tkpenalty,

Are you in any way affiliated with Xigmatek? Your recommendation is pretty positive.

Any scientific reason to only use the Frosty tech review and ignore others that do not rate the Xigmatek the top? With no reason, your recommendation sounds as dodgy as you are accusing the other reviews to be. (Not saying they are, just some data would be nice.)

And why is it natural the Scythe's would be inferior? A thick copper base in the way? How much of a problem is that? Copper is an excellent conductor of heat. The biggest concern in cooling is making sure that there is no air gap between the CPU heat spreader and the heatsink base. Air is not a good conductor, and you want as much metal to metal contact there as possible.

If you have a good spreader to heatsink connection, then heat will be conducted to the heatsink base at a rate controlled by the temperature difference, and that temperature will be controlled by how much of that heat can be removed from the base. That will be how much the base radiates plus how much the heatpipes can carry away.

Maybe a few controlled experiments will prove me wrong (won't be the first time), but I don't think that a flat smooth copper base hurts at all. So far, I think that the direct contact is visually compelling and makes for good marketing material, but have not seen test results or physical evidence that it works substantially or even any better.

VanWaGuy · Post by **VanWaGuy** » Wed Apr 02, 2008 7:34 pm

Well, experiments were run, and it didn't own the TRUE, but did very respectably.

My comments about the direct touch only being a marketing gimmick seems to have been proven wrong.

Immortals · Post by **Immortals** » Wed Apr 02, 2008 7:54 pm

VanWaGuy wrote:Well, experiments were run, and it didn't own the TRUE, but did very respectably.

My comments about the direct touch only being a marketing gimmick seems to have been proven wrong.

I would have guessed marketing gimmick too. I mean why didn't someone think of this before?

halcyon · Post by **halcyon** » Fri Apr 04, 2008 12:24 pm

Considering it bested Ninja CU and TRUE at 5V, I think it is the new champ. At least, until the next one :)

However I do believe that current heatpipe based basic designs are nearing their performance ceiling. I think a degree or two may be able to be shaved of and that's about it. It probably requires a different type of technology to achieve something significantly better.

bgavin · Post by **bgavin** » Sun Apr 13, 2008 10:30 am

I find FrostyTech to be quite illuminating.

Even if their test method is not "real world", it is highly consistent. One can certainly make educated decisions based upon their tables of consistent data between coolers.

I put their data tables into a spead sheet for further analysis by comparison. For example, I want a base that is screwed to the clips and does not swivel/rotate like the Thermalright do. Back-plate mounting is an absolute requirement. A flat, smooth, non-oxidizing plated base is a plus.

Sorting the Frosty data for 85w temp rise less than 16C, it was easy to find the Zipang is the most quiet down draft, and the Tuniq Tower 120 is the most quiet side draft at the slower fan speed.

The Zipang looks like it will even help cool the stock GA-P35-DS3 north bridge passive sink.