Cooling Thermalright Silver Soul 135 - quick n dirty initial tests in Cerb X

nubbins

Caliper Novice
Original poster
Feb 17, 2021
23
34
I've been very intrigued by the new Silver Soul 135 cooler ever since it appeared on Thermalright's website, and picked one up on the weekend from AliExpress. Thought I'd do some simple tests comparing it to the C14S, nothing too scientific but enough to see which cooler I want to use.

It seems like the perfect cooler to use for a windowed Cerberus X without resorting to water (technically doable on air but at what thermal cost?)
  • 5900X: 130/85/120 1x +200MHz, -2/-5/-30/-10/-20/-15/-20/-20/-20/-20/-20/-20
  • B550 Unify-X: A.30
  • 32GB 3800 14-15-15 @ 1.52V
  • RTX 2060 Super FE @ 0.987V
The C14S is paired with the vented panel and draws air through the panel, with 1 x NF-A9 Chromax in the back of the case as exhaust. The SS135 is paired with the windowed panel. The same 2 x NF-A14 Chromax in the bottom of the case, but running about 25% higher rpm than usual for the SS135. The NF-A9 also running about 20% higher rpm for the SS135.




Peak temp readings from Tctl/Tdie, at 500ms polling rate:

Flat 45% fan curve on C14S until 80C, flat 70% fan curve on SS135 until 80CNH-C14SSS135
CB R23 MT70.1C72.5C
CPU-Z stress 15min70.6C73.9C
Genshin Impact 3 x 1hr sessions64.9C69.8C
MW2019 custom infected 30min78.9C (average ~72C)85.5C (average ~79C)

There might be a few ways to improve its performance, but I can't test them - perhaps someone else can test:
  • Reverse airflow: reverse the rear NF-A9 as intake and get air flowing back to front through the SS135. But the NF-A9 starts to howl like an animal when facing that way. Also, there's no way to remove hot air that way - the exhaust air from the heatsink is really warm and would be detrimental to RAM temps, which are already on the brink with the SS135+Vengeance cooler (47C in TM5). Unify-X VRM heatsink ensures there is no space up top for a top bracket-mounted exhaust fan.
  • Push-pull (pull-pull): I tried two redux P12s but couldn't fit the left fan because of the I/O shroud. There is no room for push-push because without the C14S, I need direct airflow for 3800CL14. A single P12 redux makes no difference, the stock Thermalright 120-but-not-120 fan is already capable.
  • Side-mounted intake: like the M1, the Cerb [X] comes with a side bracket for mounting fans. Unfortunately, it's hard to do with the SFX internal layout and the way I've routed my 24-pin. With the U9S in the M1, side mounted intake only ever helped my RAM temps slightly, so not expecting miracles out of this one.
  • Reverse airflow + side-mounted exhaust: everyone's favourite M1 config, but again, it's not possible for me to test.
So all in all, not overly disappointed given that it was up against a maxed-out/push-pull C14S, but not impressed either. From others' rave reviews of the Silver Arrow 130 and how it's on par with the NH-C14S, I was (maybe unreasonably) expecting a little more from the Silver Soul 135. Perhaps most of the C14S-SA130 "comparisons" were conducted in the Ncase M1, with the NH-C14S in low profile mode which hobbles its performance.

If one doesn't have any RAM cooling needs, I can definitely see a 1-2C improvement at least if mounting push-pull fans. But without very low profile RAM, mounting a 120mm fan would probably run into the same clearance issues in a M1 or Cerb. I still wouldn't expect too much, since it seems like the SS135 might suffer from the same problems as the D9L, which can't be solved simply with moar fans.

Seems like a dual fan NH-U9S might offer similar or better performance with less headache, at the expense of more noise.

===============================================================================

My SS135 sample had some interesting quality control quirks. Not what I expected from an old-timer like Thermalright.
  • The fitment of the top plastic plate for both heatsink banks is amusingly subpar and all over the place.
  • One of the heatsink banks is crooked. A closer look will irk OCD sufferers.
Most importantly, there are a lot of tiny pockmarks on the coldplate. These ended up scratching my 5900X's IHS accordingly, mostly in the center around the RYZEN logo. Nothing extreme, but enough to ensure that the SS135 will never touch my CPU again until I get around to lapping the coldplate.

 
Last edited:

tinyitx

Shrink Way Wielder
Jan 25, 2018
2,069
2,000
I think a difference between using a side vented panel (for C14S) and using a windowed panel (for SS135) skewed the comparison a bit.
Could you repeat the test with ''both coolers using vented panel'' and ''both coolers using windowed panel''?

C14S has a clear and direct fresh airflow path with a vented panel while the SS135 does not with a windowed panel.

Lastly, is there any reason why you did not lap the coldplate to take care of those marks before using it?
 

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
Interesting. C14S still seems to be a great choice and one of the best top down coolers. Would be nice to see the results of the SS135 with a vented panel as well!

I gotten a aftermarket graphene coated SS135 from a Taobao vendor, but am still too lazy to get around testing it out.
 
  • Like
Reactions: nubbins

nubbins

Caliper Novice
Original poster
Feb 17, 2021
23
34
so 1 fan c14s is still as good as ss135?
Also what will happen if you throw a phanteks t30 inside ss135?

Push pull C14S only gains about 2C in sustained all-core and 1C in games. I do use iPPC fans but the only difference to the regular A14 is the rpm range, at the same rpms they are the same.

So I'd still expect the regular C14S to come out on top. Unless you run it with a bottom fan only and starve it with the windowed panel. SS135 surely would improve on test bench, but so does the C14S.

As to fans, I think the only way to get tangible improvement is to run 2 fans.

I think a difference between using a side vented panel (for C14S) and using a windowed panel (for SS135) skewed the comparison a bit.
Could you repeat the test with ''both coolers using vented panel'' and ''both coolers using windowed panel''?

C14S has a clear and direct fresh airflow path with a vented panel while the SS135 does not with a windowed panel.

Lastly, is there any reason why you did not lap the coldplate to take care of those marks before using it?

Hence the quick and unscientific test, it's unscientific as hell XD

I initially thought they were surface scratches. I could barely feel them with fingernails, so I thought it wouldn't make a mark on my CPU, as I've never had it before. I don't think the marring affects the temp results - the core deltas are better on the SS135 than the C14S, actually.

I thought about the vented panel a lot. I admit, I probably should have tested it, but I can't anymore now that I know the cooler needs lapping.

In theory, it would seemingly have better airflow, but from experience in the Cerb, Cerb X, and M1 (all same layout), there are a few reasons I think it might not:
  • The bottom A14s push a lot of air as it's pretty unrestrictive, so it's a lot of positive pressure. As a result, the cooler should get minimal air from the vented panel. It was a similar phenomenon on the M1, where if you choose to exhaust out the back with a tower cooler, the side venting doesn't help CPU temps much, cooler is about as starved as before.
  • It might actually be better in the M1v6 (not v5). In the Cerb there are no vents where the FE 2060S is, so the hot air is just hitting the side panel and spreading.
  • Obvious solution to that is to use the side bracket and mount a 120mm intake, but I can't as there is no space between the cooler height, PSU, and 24-pin.
  • In the M1v5, even the side 120mm had a minimal impact on CPU temps under NH-U9S. It helped RAM a few degrees though.
  • I'm thinking the best setup would be rear 92mm intake, cooler facing backwards, then side bracket 120mm exhaust or top bracket 120mm exhaust. All of which are impossible for me lol
Don't pay too much attention to the gaming tests, they don't mean much, they're just for me. You can see that the gap widens because the C14S draws in its own air versus the SS135 choking on the GPU - it's not a fair comparison.

But the CPU-Z 15min stress results show what it's like with the GPU idle and basically airflow solely feeding the SS135. There's a lot of air going in the case and a lot of air exhausting during that test (and that exhausr air is very warm).

But I definitely think the SS135 holds huge potential for the NCASE M1 because the case holds the C14S back on account of clearance. Especially back to front airflow.

Interesting. C14S still seems to be a great choice and one of the best top down coolers. Would be nice to see the results of the SS135 with a vented panel as well!

I gotten a aftermarket graphene coated SS135 from a Taobao vendor, but am still too lazy to get around testing it out.

Damn already? Isn't the regular cooler barely released??

If I could have had a Silver Arrow graphene I would have just gotten that, but Taobao and proxy buying stopped me
 
Last edited:
  • Like
Reactions: Neathdrawls

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
Damn already? Isn't the regular cooler barely released??

If I could have had a Silver Arrow graphene I would have just gotten that, but Taobao and proxy buying stopped me

Yeah, this vendor has it already. They are pretty quick in getting new stuff and also spare parts too, and so far their quality and customer service has been good for me.

I actually also gotten the option to change out the Grey D12-Pro to the Black one too. Your post made me go and dig the cooler out the check the base plate too, and it seems fine.

 

panda_spud

Minimal Tinkerer
New User
Sep 14, 2021
4
4
Yeah, this vendor has it already. They are pretty quick in getting new stuff and also spare parts too, and so far their quality and customer service has been good for me.

I actually also gotten the option to change out the Grey D12-Pro to the Black one too. Your post made me go and dig the cooler out the check the base plate too, and it seems fine.

Do you have the link to the shop? Cant seem to find it on taobao.
 

nubbins

Caliper Novice
Original poster
Feb 17, 2021
23
34
is it possible to install a better fan in ss135? If yes , how?

I swapped on a redux P12 using the original clips but it's kinda sketchy as they are not designed for normal sized 120mm fans. Lots of force needed, not sure if that could bend/crack a fan in the long run. But in the short term at least it works.

If you have regular 120mm clips from some other cooler (eg. U12S, H7), they might work better. The special thick 120mm clips Noctua sent me for 120mm compatibility on the C14S do not work.

I also don't think you need a "better fan" on the SS135, as that's not the problem. The default Thermalright fan is fine and a P12 offers no improvement, I doubt you'll see any improvement on a single fan with anything less than a iPPC 2000/3000rpm screamer at full speed. What it needs is a 2nd fan for push-pull and plenty of fresh directional airflow in the case, both of which I am unable to test.
 
Last edited:

millerbedicks

Chassis Packer
May 20, 2021
15
2
Can one just buy another clip? Because I am interested in adding a T30 or industrial fan though.


This person did T30 but he just said that the stock clip is not helpful and he only inserted it and it somehow sticks.
 
Last edited:

nubbins

Caliper Novice
Original poster
Feb 17, 2021
23
34
Can one just buy another clip? Because I am interested in adding a T30 or industrial fan though.

This person did T30 but he just said that the stock clip is not helpful and he only inserted it and it somehow sticks.

That's more along the lines of what I'm expecting from a properly configured SS135, wow! Looks like the NCASE M1 has found a real contender. This is what I'm talking about, you shouldn't need another 120 because it's basically already a 2-fan setup in there with the rear intake so close.

I'm not aware of Thermalright selling any clips for this cooler but you could reach out and ask. I wouldn't try to stretch the stock clips anyhow, it's not great.

Judging from the QC differences I've seen here, probably not the best plan to hope to keep a fan wedged in there.
 

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
The cooler itself comes with 4 fan clips (2 sets), so that you should be able to run push pull if your case has clearance for it.

As the fan is uniquely (?) shaped, their mounting holes are not the standard 105x105mm, that's why the supplied fan clips are slightly different and normal 120mm fans will not fit without stretching.
 

tinyitx

Shrink Way Wielder
Jan 25, 2018
2,069
2,000
The cooler itself comes with 4 fan clips (2 sets), so that you should be able to run push pull if your case has clearance for it.

As the fan is uniquely (?) shaped, their mounting holes are not the standard 105x105mm, that's why the supplied fan clips are slightly different and normal 120mm fans will not fit without stretching.
From a scaled drawing, I calculate the distance between mounting holes to be 94 (plus or minus 1) mm. As said, this shorter distance (vs standard 105mm) makes a standard 120mm fan not fitting without stretching the clip.
 

REVOCCASES

SFF Guru
REVOCCASES
Silver Supporter
Apr 2, 2020
1,534
1,786
www.revoccases.com
good estimate @tinyitx - it's indeed 95mm

just found a shop who is selling the fans separately...

Bild_2021-09-26_140525.png


additional clips could also be bend DIY using some stainless steel wire ;)
 

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
I have just gotten around to testing my Thermalright SS 135 with a 5600X.

Results are here, and will be adding more in the future (eventually when I get around to it), with the solid panel, with and without front 9025 fan.

Short summary, in my testings, the SS135 is really quite capable of a cooler at only 135mm tall.

At ambients of around 30C, it keeps my 5600X at PBO motherboard limits (~PPT 112W, TDC 69A, EDC 113A) at max 80C. In comparison, on an open bench with the NH-D15S, temps and cinebench r23 scores are very similar.

Of course, the SS135 has 3 other fans helping it inside the case, and it is definetly noisier than the D15S.

 

nubbins

Caliper Novice
Original poster
Feb 17, 2021
23
34
I have just gotten around to testing my Thermalright SS 135 with a 5600X.

Results are here, and will be adding more in the future (eventually when I get around to it), with the solid panel, with and without front 9025 fan.

Short summary, in my testings, the SS135 is really quite capable of a cooler at only 135mm tall.

At ambients of around 30C, it keeps my 5600X at PBO motherboard limits (~PPT 112W, TDC 69A, EDC 113A) at max 80C. In comparison, on an open bench with the NH-D15S, temps and cinebench r23 scores are very similar.

Of course, the SS135 has 3 other fans helping it inside the case, and it is definetly noisier than the D15S.

Impressive results! Perfect fit for a 5600X or properly tweaked 5800X, with that direct airflow path.

I thought the SS135 wasn't shabby at all on noise. The D15S is pretty quiet but I have a feeling that would change when it's put into a confined space like the SS135 is. All coolers are whisper quiet on open bench. Have you tried messing with the fan curve a bit? In my short time with the SS135 it was still reasonably quiet at 50-70%, while losing little performance. Especially with all those case fans to help out.

Though, I always have my DT770s on so everything seems pretty quiet to me, maybe my view is somewhat skewed XD
 

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
Impressive results! Perfect fit for a 5600X or properly tweaked 5800X, with that direct airflow path.

I thought the SS135 wasn't shabby at all on noise. The D15S is pretty quiet but I have a feeling that would change when it's put into a confined space like the SS135 is. All coolers are whisper quiet on open bench. Have you tried messing with the fan curve a bit? In my short time with the SS135 it was still reasonably quiet at 50-70%, while losing little performance. Especially with all those case fans to help out.

Though, I always have my DT770s on so everything seems pretty quiet to me, maybe my view is somewhat skewed XD
Thanks! My 5600X hasn't won silicon lottery though, and 5800X runs the hottest in the Zen3 series, but I think the SS135 would still do a good job with it.

Ah, maybe my phrasing was skewed. At max fan speeds, the SS135's D12-Pro at 1850 RPM is noticebly noisier than the D15S's NF-A15 at 1500 RPM.

To my untrained ear, the NF-A15 has a more of a hum/drone at max, whereas the D12-Pro has more of a rattle, but of course they have different max speeds.

Within the case itself, yes, with a fan curve the SS135 is pretty quiet in low to mid loads, in fact I can't notice much noise from the rig when it is around 40% to 50% fan curve for all fans.

I was saying elsewhere though, my standing fan next to me is pretty much the noise ceiling in my room and it can cover up most of the fan noise from my PC. Also, as full speeds, the two rear exhaust is rated at 4800 RPM, and is pretty much a helicopter.
 

Neathdrawls

SFF Lingo Aficionado
Jul 28, 2021
136
174
Are the tops/heat pipe covers removable? I'd like to paint em to match if possible...
Unsure if they can be removed, but if you are wanting to paint it black, Thermalright has released the black versions of the SS135 and SS110.

You might just want to wait for those to be in your hands then?