Retesting Cooler Master's MasterAir G100M: Last Flight to Infamy

At first glance, our initial test of Cooler Master’s MasterAir G100M revealed a good-looking CPU cooler that just couldn’t take the heat. However, our readers quickly pointed out that while the G100M is rated at 130W, the CPU in our overclocked, full-size testing rig was rated at 140W at stock settings. With that in mind, we decided to retest the MasterAir G100M on our 84W-rated Mini-ITX testing rig to see if it could earn some redemption. Unfortunately, further testing only revealed more problems.

Features


First, a quick recap. The Cooler Master MasterAir G100M is a compact, downdraft style CPU cooler whose form bears some resemblance to that of a classic flying saucer UFO. Its design features a solid copper base that conducts heat from the CPU up into a copper vapor chamber that in turn dumps heat into an array of fins surrounding the core. Then, the cooler's 90mm fan removes the heat. Said fan features built-in RGB LED lighting that accepts control via both motherboard software and an included controller. The G100M has a thermal design limit of 130W and supports all major CPU sockets, including Intel 775, 115x, 1366 and 2011x as well as AMD FM1, FM2(+), AM2(+), AM3(+) and AM4.

To install the MasterAir G100M LGA 1150 you press a set of anchor nuts into the cooler’s backplate and then place the backplate behind the motherboard. Included standoffs, which also function as attachment points for the cooler, secure the backplate in place. Once the backplate is in position, you apply a thin layer of thermal paste to the CPU's heat spreader. Secure the body of the cooler to the motherboard with the provided hardware. Finally, install the cooler and motherboard into the case, followed by the rest of the components - provided the cooler isn’t in the way.

In our original review we discovered that the G100M’s not-so-compact size caused it to encroach on the motherboard’s RAM slots, limiting users to the use of low profile memory. This is a common issue among CPU coolers, and while it can be a bit of an annoyance to builders, especially those who are building a PC for the first time, it’s rare that we find a cooler that completely blocks the installation of system components. However, that isn’t the case here as the G100M was large enough to completely block access to the PCIe slot on our testing motherboard.

It’s worth noting that the placement of the CPU socket (in relation to the PCIe slot) matters here and that other boards from other manufacturers may not have this issue. That said, since we don’t use the GPU in our CPU cooler tests and it doesn’t significantly contribute to the overall results, we’re going to press on with retesting the G100M to see how well it can handle a mid-tier CPU.

Comparison Products

We retained the hardware configuration from past compact cooler reviews while comparing the MasterAir G100M to its compact downdraft-style rivals, the Reijintek Pallas, RC-1001 Brontes, Shadow Rock LP and NH-12.

Specifications

Height2.99 inches / 75mm
Width5.7 inches / 144.5mm
Depth5.7 inches / 144.5mm
Base Height0.25 inches / 6.5mm
Assy. Offset0 inches (centered)
Cooling Fans(1) 90mm x 25mm RGB
Connectors(1) 4-pin
Weight24.1 oz / 684g
Intel Sockets755, 115x, 1366, 2011x
AMD SocketsFM2(+), FM1, AM2(+), AM3(+), AM4
Warranty2 years

As always, we maintained the ambient temperature of the test at 26°C (78.8°F) and recorded the noise levels 0.25m from the case’s front corner on the side that opens. Next, we corrected them to the 1m industry standard by subtracting 12 decibels.

Test Results

Now that we’ve moved the MasterAir G100M into our 84W Core i5-4690 testing rig, you’d think it might be able to keep things cool long enough to complete the test, right? Well guess again. Despite its 130W power dissipation rating, the G100M performed so poorly that the testing CPU reached its 100C thermal limit and throttled down to 3.1GHz in less than a minute.

In an effort to ensure that we didn’t make any mistakes during the installation, we uninstalled, reinstalled and retested the cooler two additional times. We even re-applied thermal paste and doublechecked all of the mounting hardware. Still, the results remained the same.

With a maximum speed of 2200 RPM, faster than most of the coolers in the comparison, the G100M should at least be able to keep the test CPU under 100°C. Unfortunately, that wasn't the case with the samples we received.

Unsurprisingly, the G100M’s small 90mm fan and high fan speed made it the loudest cooler, if only by a bit, in our comparison.

A combination of sky-high temperatures and almost-as-high fans speeds put the G100M at a large disadvantage in our efficiency benchmark.

Even with its $40 price tag, overall inferior performance ultimately landed the MasterAir G100M in last place in our performance-per-dollar value benchmark. Interestingly enough, the highly regarded Noctua NH-L12 found itself in nearly last place due to its higher $60 price tag, despite its outstanding performance.

With a cool design, RGB lighting and low, $40 cost, the MasterAir G100M started out as a promising CPU cooler with the potential to be one of the best compact coolers we’ve tested in a long time. Instead, what we found was an oversized, UFO-shaped disaster that seemed incapable of doing its one and only job, keeping your CPU cool.

That said, builders seeking a compact CPU cooler that performs as well as it looks should consider something along the lines of be quiet!s’ Shadow Rock LP. And if you require something even more compact, we recommend the Reeven Brontes.

MORE: Best CPU Cooling

MORE: How To Choose A CPU Cooler

MORE: All Cooling Content

This thread is closed for comments
19 comments
    Your comment
  • Dark Lord of Tech
    That type of cooler design will never yield favorable results.
  • almarcy
    I have been using my Cooler Master CPU heat pipe, fin stack and fan without failure for nearly seven years. It looks like many rectilinear designs. It works. My HWMonitor tells me my old AMD Phenom II X4 850 stays within 15C of ambient. Close enough for jazz, in my Universe. YMMV. The unit described in this review is a foolish styling exercise that never made it. Mktg is not a good guide to anything, except nonsense. I imagine some folks buy it. Sad, but accurate.
  • honkuimushi
    Much better tests. It’s a shame performance is so lacking. Still, I would have liked to see the stock fan and maybe some other 90 mm coolers in here like the Noctua L9x65 and the Cryorig C7. The first question for any third party cool has to be how much better is this than the cooler I get for free? Unfortunately, this looks like it might actually be worse.
  • stdragon
    I never recommend matching a CPU's TDP to a HSF. If the CPU is rated for 130, go with a larger capacity HSF for extra headroom to account for increases in ambient temps.

    IMHO, the Shadow Rock Slim (160W TDP) would have been a perfect match for that CPU.
  • cryoburner
    276663 said:
    That type of cooler design will never yield favorable results.

    This one is a bit different from most round cooler designs in that it has a large vapor chamber filling the entire center of the cooler. I'm curious how much that actually assists cooling performance for this heat sink though. This page has some photos of the vapor chamber, which is a large cylinder roughly 41x47mm in size...

    http://www.expreview.com/58531-all.html

    It's worth noting that AMD's Wraith Spire bundled cooler also includes a similar kind of copper vapor chamber in the center, though in that case it's surrounded by an extruded bock of aluminum with thicker fins. It's not just a copper base, but rather a large vapor chamber that extends all the way up to just under the fan.

    But why make this cooler nearly 145mm in diameter when it's only going to use a 92mm fan? They could have easily fitted it with at least a 120mm fan, and probably got better cooling performance. Though perhaps not much heat is getting out to the ends of those thin fins as it is. This cooler is around 75mm tall by 145mm wide, while AMD's Wraith Spire is just around 70mm tall by not much more than 100mm wide, with the same size fan. And yet the Wraith Spire outperforms this cooler by a decent margin, judging by this review of the G100M...

    https://www.tech-critter.com/review-cooler-master-masterair-g100m/

    If it's significantly larger and has all sorts of clearance issues, yet doesn't perform as well as AMD's mid-range stock cooler, then what's the point? And if the Wraith Spire is considered a "95 watt" cooler, then what's that say about the G100M's supposed "130 watt" capacity? It's probably an improvement over Intel's stock coolers, and perhaps even over the smaller, "65 watt" Wraith Stealth, which is only 54mm tall and lacks the Spire's vapor chamber, but it certainly doesn't seem like a $40 cooler from a performance standpoint. It performs more like a $20 cooler, with another $20 going toward a fancy-looking design and RGBs. And I do like how the cooler looks, but it definitely seems like a case of form over function.
  • Nintendork
    I would only pick CM for their bread and butter Hyper 212 line, for others coolers, Noctua.
  • gaborbarla
    I wonder how this compares to a X Dream i117 which I have seen working on a 8600k very happily.
  • mattkiss
    In regards to the "Temperatures Over Ambient" slide, what exactly are the blue and green columns, PWM delta T?
  • Rexer
    205977 said:
    I would only pick CM for their bread and butter Hyper 212 line, for others coolers, Noctua.


    I owned a CM 212 for a few years, too and that was a really good cooler. I switched to a Corsair H 100i water cooler back 2014 which was really good, too. But returned to back to air cool because of a mess a friend of mine had when his water cooler leaked. I purchased a Noctua NH-U12S and have excellent results. Don't get me wrong about the H-100i water cooler. My friend has a very elaborate set up which he cools a video card with his cpu. I may return to water cooling but right now, I feel safe with air coolers and that's just me.
  • Rexer
    276663 said:
    That type of cooler design will never yield favorable results.


    That is a curious design. Sorta looks functional. Most PC cases are aerodynamically dysfunctional so configuring a how a case and cooler work together to get the maximum benefit of air flow is an art. I recall employing an air scoop (taping a cone facing front case fans) on a reference video card when I found out the radial fans can push more air than it's able to grab. I got a few degrees cooler, not much more. Lol, got the idea from an old guy who raced a dragster in the 60's. Said he used scoops on his engine's fuel injectors to help ram air. Sort of prevent vapor lock when a vacuum is present.
  • stdragon
    1293170 said:
    That is a curious design. Sorta looks functional. Most PC cases are aerodynamically dysfunctional so configuring a how a case and cooler work together to get the maximum benefit of air flow is an art. I recall employing an air scoop (taping a cone facing front case fans) on a reference video card when I found out the radial fans can push more air than it's able to grab. I got a few degrees cooler, not much more. Lol, got the idea from an old guy who raced a dragster in the 60's. Said he used scoops on his engine's fuel injectors to help ram air. Sort of prevent vapor lock when a vacuum is present.


    Scoops are mainly about ingesting cooler air instead of the pre-heated warmer air found under the hood. The colder the air, the more dense. The denser, the better and more HP you'll generate.

    CPU's are not that much different in regards to delta T. Both engines and CPUs benefit from the largest temperature differential.

    Air scoops can also act as ram-air if shaped properly. But their only good for high speed racing as it takes a good amount of speed before you can get an increase in positive air pressure being forced down a naturally aspirated engine. But that's not applicable for computers for obvious reasons. What you really want is the former. You want to channel the coldest ambient air possible to have a direct path to the HSF. Otherwise, pulling air from the case is already pre-heated from the MB VRMs, GPU, and other system components.
  • Goku Kakkarot
    i hate this desgin! and also it's not reliable and looks ugly too... And results are bad either. Also it is covering the RAM's sockets.. So many disadvantages...
  • Crashman
    2332837 said:
    In regards to the "Temperatures Over Ambient" slide, what exactly are the blue and green columns, PWM delta T?
    The voltage regulator. I've given up on the term VRM because it properly refers to the removable resistor pack (module) used on Pentium Pros. Modern CPU voltage regulators are PWM plus filters.
  • stdragon
    8708 said:
    2332837 said:
    In regards to the "Temperatures Over Ambient" slide, what exactly are the blue and green columns, PWM delta T?
    The voltage regulator. I've given up on the term VRM because it properly refers to the removable resistor pack (module) used on Pentium Pros. Modern CPU voltage regulators are PWM plus filters.


    Say...what??!

    VRM = Voltage Regulator Module

    Dual socketed PPros have always had removable VRMs. They were removable because you could add one when adding a second processor later. Though I suppose they were also known for failure, hence needed to be replaced and not the entire MB?

    Anyways, I've NEVER heard the the acronyms of VRM and PWM get interchanged. VMRs are for voltage regulation, and PWM refers to Pulse-Width Modulation with regards to fan RPM control.
  • Crashman
    2695855 said:
    8708 said:
    2332837 said:
    In regards to the "Temperatures Over Ambient" slide, what exactly are the blue and green columns, PWM delta T?
    The voltage regulator. I've given up on the term VRM because it properly refers to the removable resistor pack (module) used on Pentium Pros. Modern CPU voltage regulators are PWM plus filters.
    Say...what??! VRM = Voltage Regulator Module Dual socketed PPros have always had removable VRMs. They were removable because you could add one when adding a second processor later. Though I suppose they were also known for failure, hence needed to be replaced and not the entire MB? Anyways, I've NEVER heard the the acronyms of VRM and PWM get interchanged. VMRs are for voltage regulation, and PWM refers to Pulse-Width Modulation with regards to fan RPM control.
    Fact: It was modular. It is no more.
    Fact 2: Pulse Width Modulation is not limited to fan control.
    Less apparent was that it was the Voltage Resistor Module before we got all slangy with it.

    Oh, and Level 0 isn't RAID ;)
  • stdragon
    8708 said:
    Fact 2: Pulse Width Modulation is not limited to fan control. Less apparent was that it was the Voltage Resistor Module before we got all slangy with it.


    So what's "PWM delta T" in reference too? I know that PWM fan control is based off Delta T values. Are you saying that's in reference to the voltage regulator? Because if so, that's very misleading semantics aside. At the very least, there should be some upfront clarification on that because I'm betting 99.99% of members on the forum here didn't catch the reference.

    8708 said:
    Oh, and Level 0 isn't RAID ;)


    Not sure what else you're referencing. But in fact there is a RAID0 (striping) as defined by SNIA (Storage Networking Industry Association)
  • Crashman
    2695855 said:
    8708 said:
    Fact 2: Pulse Width Modulation is not limited to fan control. Less apparent was that it was the Voltage Resistor Module before we got all slangy with it.
    So what's "PWM delta T" in reference too? I know that PWM fan control is based off Delta T values. Are you saying that's in reference to the voltage regulator? Because if so, that's very misleading semantics aside. At the very least, there should be some upfront clarification on that because I'm betting 99.99% of members on the forum here didn't catch the reference.
    8708 said:
    Oh, and Level 0 isn't RAID ;)
    Not sure what else you're referencing. But in fact there is a RAID0 (striping) as defined by SNIA (Storage Networking Industry Association)
    It's the voltage regulator. And no, Level 0 is not Redundant, it's jut AID.
  • stdragon
    8708 said:
    And no, Level 0 is not Redundant, it's jut AID.


    :/ Technically, you're correct. Redundancy implies fault tolerance. RAID-0 has neither. But that said, a RAID-0 volume is neither a JBOD or spanned volume. It's stripped, and it's an established technical standard. There is a difference.

    Look, I'm not arguing over technicalities. But, there is an established lexicon within the industy. By all means, make the point to the contrary. But going rogue with the lexicon (and no further context) to a wider audience is confusing as can be. That's all.
  • Larmo-Ct
    I had to replace a liquid cooler system, in a machine that I bought from iBuypower. The original, was a brand I never heard of, which I find can be common in factory builds. I replaced that with a Corsair H 100i water cooler. I installed it a couple of years back, so I'm waiting for that one to fail at some point. In a machine I just built, I have an i7-7800X LGA2066. Instead of liquid cooling, I went with a Cooler Master Hyper212X Turbo cooler. I had heard that the i7-7800X can run pretty hot, but knock wood. The Hyper212 is doing a great job in a mid sized tower case with a total of seven fans running. The Hyper212 just barely fit into the case though! :-)