What have these reviewers been smoking? No way an SKU900 outperforms an Alpha8045!

What have these reviewers been smoking? No way an SKU900 outperforms an Alpha8045!

I just bought an SKU900A today, because I needed it for a DFI Lan Party II board which does not have mounting holes for the Alpha. I paid $44 for this highly regarded piece.

I did a direct back-to-back comparison on the exact same system: 8RDA+ running 2400+ at 1.875 vCORE, which puts out a lot of heat. An Alpha8045 running a Vantec Tornado 80 mm fan (84 cfm) was netting me 44.5 C stress temps. I put on the SKU with a regular thickness 92 mm Panaflow fan: 55 C stress, horrid! I then replaced the fan with a 92 mm Vantec Tornado (119 cfm), result? 50.7 C stress. It's not even close, fellas. I put the Alpha back in and sure enough, 44.5 C stress.

And it's not even a fair comparison, as the Alpha runs an 84 cfm 80 mm Vantec and the SKU a 119 cfm 92 mm Vantec, which is quite a bit louder. I had the fan blowing on the Alpha which I found is better than sucking.

From the design, can't imagine how the SKU should cool better, as it is all copper, which absorbs heat well but does not give it off as well as aluminum (this is a point of controversy). There is a lot of surface area, but the fins are so close together it is difficult to force air through. The Alpha uses a copper core to draw heat away, fused to forged (dense) aluminum rods to give off heat. Moreover, the rods have a flat black surface to dissipate heat effectively and are spaced appropriately.

The SKU900A uses a standard clip and is a ***** to hook up with a screwdriver, which slips off and may gouge the motherboard. Once the standoffs of the Alpha is installed, the sink just screws in.

Don't mean to flame the SKU guys but these are my findings. It's not even close. I believe the 900A is a bit more compact than the 900U but doubt the difference in size will make up for that much of a difference.

mind you, the 8rda+ reads temps between 10-15 degrees lower than actual. that 44degree load? no way, esp. with 1.875v... more like 60C.

nevertheless, I think your findings are valid, but I don't know what the margin of error for the thermistor is in each individual case.

let's wait until someone else comes along. meh.

Nope, I am using a digital temperature probe touching the CPU. I have tested maybe 60 motherboards so far, and you cannot trust the temps in the BIOS. I stressed the system by running Memtest. Moreover, since I am using the exact same temp probe even if it is inaccurate, you can do a relative comparison. The entire testing scheme was done is the same 1 hour or so block of time.

Has the temp probe changed position?
Has the temp probe been crushed?
... or have other variables been altered?

I tape the probe to the cpu matrix using 3-4 strips of strapping tape. In this particular case, I could not jiggle the probe. Also, I am pretty sure the tip was not crushed when I installed the SKU900. Remember I was originally running the Alpha, took it off, put on the SKU900, took that off and put the Alpha on. If the probe was crushed it should have been crushed with the Alpha also. But no, I'm sure the probe did not shift. After all, I did not replace the CPU.

Also the temp drop from 55 C of the 2800 rpm (I think), 92 mm Panaflow to 50.7 C with the 5700 rpm, 92 mm Vantec seems plausible.

um dude wtf? after using hte alpha since u used it first shouldt u let the cpu rest or else it will stay hot... turn off the thermometor and the comp and let that crap rest it will be scorching wiht heat redo this test then say the slk 900 u is bad and also try thr new on sp 94

Does this mean your alpha is better then all hsfs?

um dude wtf? after using hte alpha since u used it first shouldt u let the cpu rest or else it will stay hot... turn off the thermometor and the comp and let that crap rest it will be scorching wiht heat redo this test then say the slk 900 u is bad and also try thr new on sp 94

Click to expand...

Nope. Alpha on: 44.5 C, SKU900 on: goes up to 55C, put on 92 mm fan on SKU, temps drop to 50.7 C, put Alpha back on, temps drop to 44.5 C. Obviously I run Memtest a couple of minutes to get temps to stabilize. I mean, it's not rocket science.

Heck, why do you think I put the Alpha back on and am running it now? I got no bone to pick. I can't return the SKU. I'm stuck with it, but who wouldn't prefer 44.5 C temps over 50.7?

Does this mean your alpha is better then all hsfs?

Click to expand...

Nope, according to two reviews I seen, the SKU900 (-U anyway) is supposed to be the best there is. And they tested the Alpha8045 too. There is a thread here about heatsink reviews being biased . . .

I have no reason not believe your review. You seem to know what u are doing. How about a few more tests to see if in fact your Alpha is truly performing better. U only tested the SLK once so I don't think u can come to such a conclusion that quick. Did u check to see if there were any flaws at the bottom of the heatsink. Perhaps minor scratches? Since u only tested it once I may have to question the seating of the heatsink at the time of your test. If u can repeat the same tests and still prove the Alpha is still better by about the same margin than u have me sold.
CHEERS!

TheGr8s1

Nope sorry, I'm not putting the SKU900A back on the board. It was a ***** as that clamp-on bracket is a hassle, and there is not much working space with the motherboard installed in the case. The screwdriver slipped a couple of times while putting pressure on the clamp so I risked damaging the board. But with the sink on, I wiggled it to be sure it was seated properly.

Also, if the probe is crushed, you usually get an FFF error on the LCD readout.

The heatsink was brand-new, and I had just removed the tape. When I removed the sink, I looked at the imprint on the bottom where the cpu touched it, and it looked OK. I am using AS3.

I WISHED the SKU900A worked better, or I would be running it now. But what can I say, I just feel it's fair to Alpha to post my results. Again, from the design, I have always been skeptical of all copper heatsinks.

All materials release heat at the same time its just denser objects, Copper, have more heat to dissapate. Also the black has such a minute amount of heat loss through radiation you shouldnt consider this a facto in the temp difference. The reason i think you got better temps was becaus the AS3 was burned in, The alpha is mounted better, the slk-900 was scratched. I highly doubt review sies are lieing to us since the ALpha isnt even competitive with the Swiftys which the Slk beats

oh my god dude, stop calling it a bloody SKU, its SLK!!
Say it with me now, SLK .

Not saying your results are incorrect, just not enough trials through the process to make an assumption.

Possibly you are not achieving the same level of pressure with the 900 as you were with the alpha. (most reviewers review the bolt on, not clip on)

Thermal past application can attribute some to the difference.

The slk using clips on a vertical motherboard (tower case) could possible cause part of the heatsink to be pulling off the core, and part to be pressing on due to the 900's weight.

The 92mm fan may be decreasing your temps due to its larger "dead spot"

indeed... considering the number of ppl that have tested and believes in thermalright heatsinks it kinda overshadow your own review and makes it seem as if it was human error and not a fault of the heatsink

well if there is any margin of error in his testings, it was marginal for both the alpha AND the SLK, so either way in directo comparison, his alpha works better for him.

I have heard that ThermalRight produced some sub par slks. 800s and 900s not performing what they used to, so maybe the heatsink previously reviewed was good quality, but yours that youve gotten is of sub standard?

As mizzery stated, the clipped heatsink could perhaps be pulling itself off the core. The more pressure you have between the 2 surfaces the better, but you dont want to crush the core. bolt ons definitely sustain constant even pressure if mounted correctly, so some degrees may be lost due to that.

now you said your slk doesnt have any scratches, but is it flat? perhaps to get your slk working better you should lap it anyways to remove any scratches that may be hard to see AND to make it flat. I know EVERY heatsink ive gotten comes concave, so a good lapping may do alot of good.

but hey, if all else fails and your slk still doesnt work as good as your alpha, you can count yourself lucky that you somehow got a super alpha that outperforms the #1 rated heatsink on the market right now. That's pretty cool if you ask me =P

*edit* also i noticed you said for your alpha you stated that you have the air blowing on the heatsink because it works better that way, Please tell me you have your fan blowing down onto your heatsink with the slk as well? both should be pushign air ONTO the heatsink. If you already have it like this, then just excuse this addition. it just seemed weird that you made an addition to one of your posts to state that the alpha was different by pointing the fan to blow air onto the heatsink.

If anything, the Alpha's clamping force is suspect. Because it uses standoffs and spring tensioned bolts, it had to be designed so undue force is not applied by the sink to the board. I think the spring force is maybe 1/3rd of that for the Alpha8942, used for Intel socket 478 apps. Sometimes I wonder if the sink is even making contact.

The SLK-900A (got that right now) on the other hand, needed a heck of a lot of force to clamp it on. There is no doubt in my mind adequate contact was made! As I mentioned, I wiggled it between fan changes to feel if the sink was cocked on the chip or not. And my XP2100 is pretty new so the foam pads are not shot or anything.

As for AS3 burn in, I've seen it at maybe 10 minutes. I ran each fan on the SLK-900A for 20' each. Sink was brand new, so there could be no scratches. Feeling the sink, it does absorb a lot of heat, but it does not seem to get rid of it efficiently. What it might need is a fan that has stronger blowing force to get air through the fins, but you'd have to check the fpm of the smaller 60-80 mm fans to try to find the best one. Still, a 119 cfm fan is a screamer.

The poor performance of the Alpha on previous reviews may have been due to the fans used. Ideally, comparisons should be made between heatsinks using the same cfm fans. I am using a huge Vantec 80 mm Tornado on my 8045, but most of the old reviews of the Alpha tested with regular size 80 mm fans. As I mentioned, I always suck with the Alpha (as they recommend), but with the huge Vantec, I've found blowing was good for a 3-4 C reduction. I hate heatsinks that blow as you are spreading hot air on to the mosfets, caps, and ram.

The next time I use the SLK-900A, I might try that 6500 rpm 60 mm Delta. Something that blows air with more force at the core. I also have some 5500 rpm 80 mm fans, although the 92 mm Vantec Tornado runs at 5700 rpm.

I'd believe you. Swiftys and Alphas are still great coolers.

when you took off that sticky plastic on the bottom of the heatsink did you clean the base with isoproyl alcohol? the sticky plastic leaves a glue on the bottem, that might be the problem.