Scythe Fuma 3

[crazyeyesreaper]

The Scythe Fuma 3 stands out as a compact dual-tower cooler, boasting solid performance with a max height of only 154 mm granting it better compatibility with smaller or older cases. The Fuma 3 shines for mid-tier builds where extreme air coolers or all-in-one liquid coolers don't make sense.

Show full review

 

[crazyeyesreaper]

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

 

[lawood]

Thank you for the amazing review.

I was wondering if you could down the line, maybe, review the Artic freezer 34 duo.
It's a very popular cooler in europe and us and is kinda cheap.
Really curious how it would stack up with all this new competition.

 

[AnotherReader]

Thanks for the review. In addition to its relatively low height, it's also much cheaper than Noctua or BeQuiet. As for the graphs, I prefer the new scheme where the reviewed product stands out.

 

[Onasi]

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

I would maybe prefer just a tad more vibrance on the compared coolers. The current scheme is fine, but, at least on my devices, seems a bit TOO muted. Maybe a middle ground between new and previous one? Should improve at a glance readability.

 

[Chaitanya]

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

Older ones were easy to read compared to new ones on small screen devices on PC it doesnt matter which colour scheme is used.

Also its surprising fans have gotten louder while their RPM range is quite low.

 

[Footman]

Pretty good value to performance metric. Nice review.
I am still waiting to see how the Thermalright Phantom Spirit does on your testbed for $34....

 

[Event Horizon]

Appreciate and respect the hard work. If we're comparing air coolers then adding Thermalright products would be great because many reviewers have found them to be extremely good performers.

 

[bug]

Are these dual-fan designs really needed? I mean, I have a single fan design, but the hot air doesn't seem to be loiter 2cm away from the fan because of lack of a second fan to push it farther.

 

[Icon Charlie]

Thanks for the review, but you already know that if I get a CPU like the 7900 which is only 65 watts. I going to be using my good old Cooler Master Hyper T2 that I am currently using on my 5900OEM.

I'm so not going to pay $50.00+ on any glorified heatsink when I already know that at this moment there is no real advancement in cooling your rig.

I could argue that Cooling Tech has gone backwards because of EYE CANDY.

There is a reason why I have new Corsair Carbide 400R and 500R cases in storage. That style of case was premier back then and it works great now for helping to cool off your components.

There is a reason why I keep my old CPU coolers. Because they still work and work well.


But anyhow. Thank you for your hard work on the review as I know how time consuming (professionally) this is.

 

[HairyLobsters]

I thought they said they were making all black and white variants for the Fuma 3.

 

[tami626]

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

I prefer the new one.

Older ones were easy to read compared to new ones on small screen devices on PC it doesnt matter which colour scheme is used.

I am looking at it on my phone and everything looks clear to me.

 

[MachineLearning]

Thank you for the review.

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

It looks completely fine either way IMO. I think I slightly prefer the current colors, but it's a slim margin.

Also, I know this has been asked many times in the past so sorry if you're sick of hearing it. But any chance for a Thermalright cooler review? Preferably Peerless Assassin 120, and/or Frost Commander 140. Their value is virtually untouchable although Scythe can come close.

 

[ir_cow]

Appreciate and respect the hard work. If we're comparing air coolers then adding Thermalright products would be great because many reviewers have found them to be extremely good performers.

You talking about the retention bracket?

 

[ADB1979]

Looking for feedback from our members: do you prefer the older color scheme for the charts which were darker, or the more muted colors in the charts here which help the reviewed product standout a bit better compared to the other colors used for directly compared coolers?

The only way to truly ask your readers this question is to do a dummy article with a few populated graphs, the same graph content but the old and the new color scheme, or a multiple choice, go nuts, have 5 choices shown with 5 different sets of content As it's not a review all answers should be about the colors, or put it to a vote.

I am sure many people will have the horrific memories of that stupid (and fake) row about whether that damned dress was blue and black, or white and gold, and the answer depended on the persons monitor / phone / TV / Holo-Deck settings, and you are going to have the same situation here (apologies for bringing this up). I have my monitor set with reduced brightness (but I have done that to every monitor I have ever owned) and in the case of this particular monitor I have reduced the contrast. Someone else (I hope they are a not a Clone of me) will want different setting on an identical monitor, GPU, other PC settings, seating position, monitor angle, lighting etc. There are FAR too many variables and so the only (simple) way is to have side by side comparisons.

There are of course more complex and better options but that is going to need more work on your end by producing two sets of every chart and have a toggle button for the end user, or have it automatically show the alt chart when someone mouses over that chart, and they are just off the top of my head. Certainly for now I would go with giving people dummy charts and either put it to a vote or just read all the comments. Soon enough you will have wished you never asked the questions

 

[Event Horizon]

You talking about the retention bracket?

I mean the coolers. Probably too late for the bracket at this point.

 

[crazyeyesreaper]

I would maybe prefer just a tad more vibrance on the compared coolers. The current scheme is fine, but, at least on my devices, seems a bit TOO muted. Maybe a middle ground between new and previous one? Should improve at a glance readability.

I prefer the new one.

I am looking at it on my phone and everything looks clear to me.

Thank you for the review.

It looks completely fine either way IMO. I think I slightly prefer the current colors, but it's a slim margin.

Also, I know this has been asked many times in the past so sorry if you're sick of hearing it. But any chance for a Thermalright cooler review? Preferably Peerless Assassin 120, and/or Frost Commander 140. Their value is virtually untouchable although Scythe can come close.

The only way to truly ask your readers this question is to do a dummy article with a few populated graphs, the same graph content but the old and the new color scheme, or a multiple choice, go nuts, have 5 choices shown with 5 different sets of content As it's not a review all answers should be about the colors, or put it to a vote.

I am sure many people will have the horrific memories of that stupid (and fake) row about whether that damned dress was blue and black, or white and gold, and the answer depended on the persons monitor / phone / TV / Holo-Deck settings, and you are going to have the same situation here (apologies for bringing this up). I have my monitor set with reduced brightness (but I have done that to every monitor I have ever owned) and in the case of this particular monitor I have reduced the contrast. Someone else (I hope they are a not a Clone of me) will want different setting on an identical monitor, GPU, other PC settings, seating position, monitor angle, lighting etc. There are FAR too many variables and so the only (simple) way is to have side by side comparisons.

There are of course more complex and better options but that is going to need more work on your end by producing two sets of every chart and have a toggle button for the end user, or have it automatically show the alt chart when someone mouses over that chart, and they are just off the top of my head. Certainly for now I would go with giving people dummy charts and either put it to a vote or just read all the comments. Soon enough you will have wished you never asked the questions

The thought process on the color change was to make the green standout more compared to all the other darker colors typically used. This is due to the fact that rather than comparing say a single 4090 AIB model to the FE version when it comes to my charts I have more coolers highlighted. Meaning with all the more vibrant colors it can be a bit more confusing. The idea here was that by keeping the Green as is and using more subdued colors the reviewed cooler would be easier to find in the charts meanwhile similar products would be highlighted but also less in your face so to speak. So direct competitors can still be picked out easier but also makes things look a bit cleaner. Experiments will continue as we get more feedback no permanent changes to be made yet but it was something we wanted to look into.

 

[ADB1979]

The thought process on the color change was to make the green standout more compared to all the other darker colors typically used. This is due to the fact that rather than comparing say a single 4090 AIB model to the FE version when it comes to my charts I have more coolers highlighted. Meaning with all the more vibrant colors it can be a bit more confusing. The idea here was that by keeping the Green as is and using more subdued colors the reviewed cooler would be easier to find in the charts meanwhile similar products would be highlighted but also less in your face so to speak. So direct competitors can still be picked out easier but also makes things look a bit cleaner. Experiments will continue as we get more feedback no permanent changes to be made yet but it was something we wanted to look into.

Thank you for the reply.

Clearly indicating the actual product under review is IMHO priority 1, nothing makes sense otherwise.

Now that I see your old charts, side by side with the new ones, the chart on the left is clearly the better IMHO.

On a different note (Mods, please feel free to move this sentence to a new thread). Your testing methodology is fantastic in ways, and not all useful in others and needs IMHO a total reformation, mostly to do with what coolers are included in your table. Yes it is perfectly good and fair to compare a dirt cheap air cooler to a very expensive water cooler to give people an idea of value to performance, but to do so there needs to be a "baseline" Default packaged AMD/Intel cooler is the literal bottom end, and then include a top notch water cooler for the "pinnacle". What then to include in the middle as a product "for the rest of us".? Including 100 products is insane, including 5 is weak, 10 is pretty much cool with everyone. So how to compare in this chart.? Do you have (as always the pinnacle and the weakling) and the top 9 plus the review, or do you have the 9 nearest competitors by price, by air/water only, by overall product size etc.? It is IMHO as unfair to compare an air cooler to 9 water coolers when the end user doe not want water cooling and doesn't have the space to do so either, which makes any review for them entirely dependent on what is practical, i.e. physically fits their needs/case.

There are many ways to slice this cake, but this cake is yours not mine and I wont tell you how to cut your own cake but I am more than happy to help, as I am sure are dozens of others.

 

[Lesha]

The thought process on the color change was to make the green standout more compared to all the other darker colors typically used. This is due to the fact that rather than comparing say a single 4090 AIB model to the FE version when it comes to my charts I have more coolers highlighted. Meaning with all the more vibrant colors it can be a bit more confusing. The idea here was that by keeping the Green as is and using more subdued colors the reviewed cooler would be easier to find in the charts meanwhile similar products would be highlighted but also less in your face so to speak. So direct competitors can still be picked out easier but also makes things look a bit cleaner. Experiments will continue as we get more feedback no permanent changes to be made yet but it was something we wanted to look into.

Prefer the old one, the green stands out. Someone mentioned that the new scheme is too muted and I agree.

 

[docnorth]

The thought process on the color change was to make the green standout more compared to all the other darker colors typically used. This is due to the fact that rather than comparing say a single 4090 AIB model to the FE version when it comes to my charts I have more coolers highlighted. Meaning with all the more vibrant colors it can be a bit more confusing. The idea here was that by keeping the Green as is and using more subdued colors the reviewed cooler would be easier to find in the charts meanwhile similar products would be highlighted but also less in your face so to speak. So direct competitors can still be picked out easier but also makes things look a bit cleaner. Experiments will continue as we get more feedback no permanent changes to be made yet but it was something we wanted to look into.

The new layout on the left looks more convenient. Just me.

 

[claes]

I too prefer higher contrast

 

[crazyeyesreaper]

Thank you for the reply.

Clearly indicating the actual product under review is IMHO priority 1, nothing makes sense otherwise.

Now that I see your old charts, side by side with the new ones, the chart on the left is clearly the better IMHO.

On a different note (Mods, please feel free to move this sentence to a new thread). Your testing methodology is fantastic in ways, and not all useful in others and needs IMHO a total reformation, mostly to do with what coolers are included in your table. Yes it is perfectly good and fair to compare a dirt cheap air cooler to a very expensive water cooler to give people an idea of value to performance, but to do so there needs to be a "baseline" Default packaged AMD/Intel cooler is the literal bottom end, and then include a top notch water cooler for the "pinnacle". What then to include in the middle as a product "for the rest of us".? Including 100 products is insane, including 5 is weak, 10 is pretty much cool with everyone. So how to compare in this chart.? Do you have (as always the pinnacle and the weakling) and the top 9 plus the review, or do you have the 9 nearest competitors by price, by air/water only, by overall product size etc.? It is IMHO as unfair to compare an air cooler to 9 water coolers when the end user doe not want water cooling and doesn't have the space to do so either, which makes any review for them entirely dependent on what is practical, i.e. physically fits their needs/case.

There are many ways to slice this cake, but this cake is yours not mine and I wont tell you how to cut your own cake but I am more than happy to help, as I am sure are dozens of others.

When moving coolers over I had to prioritize ones that actually supported LGA1700.

After sorting through what I had. I essentially picked. a few coolers from each price / performance segment.
Air Coolers:
Entry = ID-Cooling SE-224XTS / CM Hyper 212 Halo Black / DeepCool AK400
Mid-Range= Noctua NH-U12S / Montech Metal DT24 Premium / Endorfy Fortis 5 dual-fan
High-End = Noctua NH-U12A / NH-D15 / NH-D15S / Assassin IV

The above coolers cover the following price points: $35 / $40 / $45 / $65 / $70 / $80 / $95 / $100 / $120 / $130.

As for AIOs, I tested 2 of each size. 2x 240s 2x 360s and 2x 280s

That worked out to 10 air coolers and 6 AIOs meaning I rebenched on 2x systems 2x times for 16 coolers doing multiple runs and remounts since I test at max RPM and a set noise normalized test. that's 4x runs just to get initial data not counting again the remounts and additional runs to verify results.

Previously coolers would take something like 20+ hours per review it may have been more than that I can't remember now. But it was a lot of time invested. The new method is actually a bit quicker but not by a whole lot. Keep in mind i have actual jobs and family to care for.

As for stock coolers Intel doesn't include them with K series and AMD doesn't include them with X series add in Intel has 2x versions and AMD has 3x versions thats 5 coolers to test or 100 man hours of work give or take.

 

[Chrispy_]

I also prefer the more muted charts over the old ones.
I have atypical colour vision even by regular red-green colourblind standards so using contrast to highlight the review product among the pack is welcome.
If people prefer the bolder colours, then another option is to invert the label colours for the highlighted item, or use bolder fonts.

As for the Fuma3, thanks for the review. It seems to be stuck in the middle of pack of decent, but unspecial coolers that are selling for a relatively high price compared to the sweet spot offerings from ID-cooling, Thermalright, Deepcool. Regional pricing and availability plays a big part, of course, as does your platform of choice - given how LGA1700 and AM5 results can differ by so much.

 

[Heiro78]

The thought process on the color change was to make the green standout more compared to all the other darker colors typically used. This is due to the fact that rather than comparing say a single 4090 AIB model to the FE version when it comes to my charts I have more coolers highlighted. Meaning with all the more vibrant colors it can be a bit more confusing. The idea here was that by keeping the Green as is and using more subdued colors the reviewed cooler would be easier to find in the charts meanwhile similar products would be highlighted but also less in your face so to speak. So direct competitors can still be picked out easier but also makes things look a bit cleaner. Experiments will continue as we get more feedback no permanent changes to be made yet but it was something we wanted to look into.

I prefer the new style with the new item being full color. Like Onasi said, the other highlighted items are too muted where it's a little difficult to tell that the be quiet dark rock elite is one. It could just be an issue with that brown as the other colors are easy to distinguish. Thank you for your work.

 

[ADB1979]

When moving coolers over I had to prioritize ones that actually supported LGA1700.

After sorting through what I had. I essentially picked. a few coolers from each price / performance segment.
Air Coolers:
Entry = ID-Cooling SE-224XTS / CM Hyper 212 Halo Black / DeepCool AK400
Mid-Range= Noctua NH-U12S / Montech Metal DT24 Premium / Endorfy Fortis 5 dual-fan
High-End = Noctua NH-U12A / NH-D15 / NH-D15S / Assassin IV

The above coolers cover the following price points: $35 / $40 / $45 / $65 / $70 / $80 / $95 / $100 / $120 / $130.

As for AIOs, I tested 2 of each size. 2x 240s 2x 360s and 2x 280s

That worked out to 10 air coolers and 6 AIOs meaning I rebenched on 2x systems 2x times for 16 coolers doing multiple runs and remounts since I test at max RPM and a set noise normalized test. that's 4x runs just to get initial data not counting again the remounts and additional runs to verify results.

Previously coolers would take something like 20+ hours per review it may have been more than that I can't remember now. But it was a lot of time invested. The new method is actually a bit quicker but not by a whole lot. Keep in mind i have actual jobs and family to care for.

As for stock coolers Intel doesn't include them with K series and AMD doesn't include them with X series add in Intel has 2x versions and AMD has 3x versions thats 5 coolers to test or 100 man hours of work give or take.

Thank you for a comprehensive answer to all points.

You Sir are a champion