Thermalright Silver Arrow Heatsink and Fan Unit Review @ [H]

FrgMstr

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,634
Thermalright Silver Arrow Heatsink and Fan Unit Review - Thermalright's new processor air cooler is surely a no-compromise unit that will not be for everyone, much less every enthusiast. It is bigger than Texas, or Alaska for that matter. The Silver Arrow is destined to be displayed in many a chassis' clear side panel. Does it have what it takes when it comes to performance?
 
This might just be my next cooler. I've been wanting to kick my stock cooler to the curb since I bought my Core i7 920. Looks like I'll need to find some good low profile RAM so I can use all my DIMM slots, though.
 
Hmmm... it looks very similar to the Noctua NH-D14, I wonder how it performs against it?!
 
I saw a review on xbit that compared them. On their review, the silver arrow nudged past the noctua.
 
dracos said:
Hmmm... it looks very similar to the Noctua NH-D14, I wonder how it performs against it?!
Click to expand...
Cooler and quieter:

24_diagrt(xbt).png

27_dnoise(xbt)_small.png


[H] chose a good HSF to review.
 
FYI, in your chart comparing Sound Pressure Levels, you label the Silver Arrow as "thermaltake" instead of thermalright
 
Emphasis mine said:
The base of the Silver Arrow is free from scratches or dents but upon closer inspection is not truly flat. It is concave. Thermalright does this on purpose in connection with its mounting system to apply more pressure to the center of the CPU.
Click to expand...
I think you probably meant convex. Previous TR heatsinks were convex-based, the image appears to show the reflection be bowed out, and this would mean the centre of the heatsink base would contact the centre of the IHS before the edges (and put more pressure on the centre for better contact in case the IHS is concave).
 
So how does the PWM for the fans work? Will these daisy chain like arctic cooling PWM fans? Just wondering how this works with dual fans and only one PWM header.
 
Does the size of the unit effect the internal case temperature?

Looking at it one way, it's large and could significantly impact/restrict airflow.

On the other hand, it's adding a few large fans that might either push alot of case heat out, or pool cooler air in...
 
Probably because its quieter.

I passed on this for a few reasons:

1. Its going to limit ram choices.
2. I personally thing the large size and overall appearance are not pleasing to the eye.
3. I can tolerate some iffy colors on a board, but not on a fan. Baby blue? Really?

So I went with the venemous and my good please thermaltake 2018.

2010-10-16151241-1.jpg
 
My thoughts exactly~ Isn't this getting into H50/H70 prices? If so, how does it fare against them?
 
DWolvin said:
My thoughts exactly~ Isn't this getting into H50/H70 prices? If so, how does it fare against them?
Click to expand...

as far as I remember a dual fanned H50 only comes close to the performance of a good air cooler. I don't recall the H50 ever beating a TRUE/VX.
 
Best tested temps AND noise at the same time?

Worth the extra money and size to me. Just have to make sure my next case will fit it...
 
diminishing returns is right. We're reaching the very limits of ambient air-cooling and I don't expect to see anything significantly better. Just see how the venomous X is still right there with this monster despite being a decent amount older (which as a VX owner makes me somewhat relieved). I'm guessing the next product we see from corsair in order to compete is a double rad self contained unit since they're also pushing the envelope with their current form factor (H50/70)
 
Wow, looks like we have a new king of the hill here in a very competitive industry.

This just might be the first and only card i'll excuse for causing you to lose the use of that first DIMM slot as this is incredible performance. It shocks me that this is so quiet based on the testing I see.
 
the thermalright version is a bit pricier, personally I'd go with the cogage version since I'd replace those fans with a single san ace 9G1212H1011 any way.
 
I'm thinking I might get one of these, use it in "suck" mode and fashion some sort of ducting out the rear exhaust fan hole, and replace the exhaust fan all together...

Fewer fans = less noise = happier me
 
Zarathustra[H];1036383955 said:
I'm thinking I might get one of these, use it in "suck" mode and fashion some sort of ducting out the rear exhaust fan hole, and replace the exhaust fan all together...

Fewer fans = less noise = happier me
Click to expand...

a 38mm fan fits perfectly in the middle, and performance pc has one of the best on stock normally (the san ace I mentioned above).
 
Just wondering, how much does the backplate stick off the back of the motherboard at the areas it screws in through the top? Reason I ask is I have a haf 932 with a x58 ud4p, and the little window thats supposed to allow easy access to that area seems to be a bit off; the bottom left hole is outside of that window. I remember a friend had a similar problem with an ocz vendetta 2, he had to flex the whole panel a bit to get the backplate to slide in (it basically pushes against that panel the motherboard is mounted on). I'd be really interested in trying this out especially since I have a cheap but extremely effective ducting system heading into the side of my case; for most of the year the ambient temperature outside is well below what it is inside for most of the day.

o56uj4.png


Heres a picture of what I mean. Blue is the motherboard, black is the top 2 mounting holes, brown is the panel the motherboard is mounted on and red is where the bottom 2 mounting holes are (behind the panel). There is a gap between them obviously (with the standoffs), I am just curious if the space will be enough. Looking at the backplate for it, it appears so but I cannot tell for sure.
 
Last edited:
assuming they use the same backplate as their aftermarket 1366 rev. A/B mounting kits there should be enough clearance for 8mm motherboard standoffs. I'm currently using the 1366 rev. B kit and have clearance w/o the backplate window. if you're asking if there's enough clearance to assemble it w/o removing the board, it may be tight, but you should be able to slide it in.
 
with lower rpm fans they seem to show the same performance, but turn up the speed and the arrow starts to pull ahead.

I think this may be due to the basic design differences.

the nh-d14 uses 6x6mm heat pipes while the arrow uses 4x8mm, while this decreases total cooling capacity (some where ~13%) slightly it gives a spacing advantage on the fin array (less heat contamination between the pipes on the fins). the cooling capacity of the 8mm pipes v. 6mm is roughly 33%, and being that the center of the ihs is where the most heat is generated the arrow has 2x8mm pipes doing the brunt of the work v. the nh-d14's 2x6mm.

also the spacing of the fins on the arrow is tighter, but they're also thinner than the fins of the nh-d14 and there's more of them giving it a greater surface area for cooling.

from all the test info I've seen all this adds up to is only ~2C better temperatures on the arrow over the nh-d14 under overclocked load temperatures apples-to-apples with higher speed aftermarket fans (I'd really like to see some tests done with some sanyo denki fans though... lol).
 
Honestly it seems like this is identical to the D14 but with larger fans, the 2C difference might as well be chalked up to margin of error once both coolers have identical fans on them.
 
Last edited:
It really does look like a D14 clone. Not that I guess there's anything wrong with that...competition and all. It does make me wonder if we've hit the limit for air-cooled tech. Till someone figures out that heatpipes made from unobtanium work 5C cooler, this is probably it.
 
spagthorpe said:
It really does look like a D14 clone. Not that I guess there's anything wrong with that...competition and all. It does make me wonder if we've hit the limit for air-cooled tech. Till someone figures out that heatpipes made from unobtanium work 5C cooler, this is probably it.
Click to expand...

Someone will come up with some new angle, like the move to tower coolers, and it'll drive a new generation.
 
Activate: AMD said:
diminishing returns is right. We're reaching the very limits of ambient air-cooling and I don't expect to see anything significantly better. Just see how the venomous X is still right there with this monster despite being a decent amount older (which as a VX owner makes me somewhat relieved). I'm guessing the next product we see from corsair in order to compete is a double rad self contained unit since they're also pushing the envelope with their current form factor (H50/70)
Click to expand...

There's been some hoopla about a 1-way 'wick' style material that will make its way to heat syncs sooner or later. I vaguely recall Intel investing in it... if this catches on, you might see some new air solutions that are a huge improvement. ..wish I could find the link :(
 
They really need to post temps of the mobo chips that are underneath. If you can't overclock the motherboard because the chip-sets are too hot it defeats the purpose of having a Brontosaurus for a cooler.

BTW, I'm confused by the setup. Kyle lists Push and Pull, but also states they could blow it out the back. Anyone have the ability to draw a little airflow diagram? If the air is blowing in, and front fans are blowing in, even with top fans it seems you'd have a turbulence issue mid-board.
 
In most cases, it outperforms the former leader, the Noctua NH-D14. Not only that, but the Thermalright is also much more quiet.
But the Noctua NH-D14 aare equally massive, slightly louder, and slightly hotter cooler in most cases outperforms the Silver Arrow in extreme overclocks.
 
You must log in or register to reply here.
Back
Top