Pressurize with All-in-One Water Cooling! What kind of cooler do I need for M.2 PCIE Gen5 SSDs?

This time with T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB SSD to test, bring you to see M.2 PCIE Gen5 x4 SSD what kind of cooler is needed to use will be better, Gen5 SSD heat generation can not be suppressed through the motherboard thin cooling pads, so in addition to using the motherboard's own cooler, also with passive Jonsbo M.2 SSD hard drive cooler, active air cooling small tower cooler T-FORCE DARK AirFlow I SSD Coole, an all-in-one water cooler T-FORCE DARK AirFlow I SSD Coole, and a water cooler. Therefore, in addition to using the motherboard's own cooler, we also used the passive Jonsbo M.2 SSD cooler, the active air-cooled T-FORCE DARK AirFlow I SSD Coole, and the water-cooled T-FORCE SIREN GD120S AIO SSD Cooler to test the cooling performance and analyze the advantages and disadvantages for your reference.

T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB Solid State DriveSpecifications:

Interface: PCI-Express 5.0 x4
NVMe: 2.0
Interface format: M.2 2280
Continuous read speed: 12400 MB/s
Continuous Write Speed: 11800 MB/s
Capacity options: 1TB / 2TB / 4TB
Controller: Phison PS5026-E26-52
NAND Flash: Micron Micron 232L 3D TLC chips
Cache: Hynix SK hynix LPDDR4 4GB
Dimensions: 80 x 22 x 3.7 mm
Warranty: 5-year limited warranty
TBW Durability: 1200 TB

 

T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB SSD Unboxing

The T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB is used as the test subject this time. The capacity models are available in 1TB, 2TB, and 4TB specifications, and they use PCI Express (PCIE) Gen5 x4 transmission bandwidth, so the maximum sequential read/write performances are 12400 MB/s and 11800 MB/s, while the 1TB version will perform a bit worse. The 1TB version has a lower performance.

The 2TB version has a TBW (Tera Bytes Written) of 1200 TB, which translates to 2048 GB (2 TB), the DWPD_Drive Writes Per Day (DWPD_Drive Writes Per Day) would have to be 657.5 GB per day for 5 years to exceed the nominal TBW value. The TBW value is standardized.

The thin graphene heat sink included in the box features better heat dissipation performance and is compatible with all types of heat sinks.

∆ T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB.

∆ Back of box.

∆ Inside the box, there is a thin graphene heat sink in addition to the main body.

 

The Z540 also uses a combination of Phison PS5026-E26-52, SK hynix LPDDR4 4GB (H9HCNNNCPUML) DRAM Cache, Micron 232L 3D TLC solution, which is very common in today's PCIE Gen5 x4 SSDs, for example. Previously unboxed AORUS Gen5 10000 SSD 2TBOr other Seagate FireCuda 540, MSI SPATIUM M570/580 PRO FROZR and other PCIE Gen5 x4 competitors sold in Taiwan, are using this solution, the premise of course, or if all according to the unboxing text of the combination of the principle of the future is not updated la.

The other big M.2 PCIE Gen5 x4 option is Micron's T700/T705, which also uses the Phison PS5026-E26-52 as its main controller, but the DRAM cache is Micron's own LPDDR4 instead of competitor SK hynix's, so the overall combination of the T700/T705 is a bit different from the other models. So the overall combination of T700/T705 is a bit different from other models.

∆ Common Gen5 M.2 SSD solution.

The ∆ DRAM Cache uses a single SK hynix 4GB LPDDR4(H9HCNNNNCPUMLXR-NEE)_4266Mbps.

∆ Dual-sided die layout, Micron 232L 3D TLC NAND Flash die.

 

T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB Solid State Drive Performance Testing

The test platform uses an AMD Ryzen 9 7900 processor with a GIGABYTE B650I AORUS ULTRA motherboard and a T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB installed in the first slot of the motherboard, M2B_CPU, which is provided with full PCIE Gen5 x4 bandwidth by the processor's direct connection channel. The performance of this test is based on the full PCIE Gen5 x4 bandwidth provided by the processor's direct connectivity channel. We would like to remind you that the test performance may vary depending on the firmware version of the SSDs, system hardware configuration, and other factors, and therefore the results are provided for reference only.

Testing Platform

Processor: AMD Ryzen 9 7900 (PBO AUTO)
Cooler: AMD Wraith Prism
Motherboard: GIGABYTE B650I AORUS ULTRA (BIOS Version: F32b)
Memory:G.SKILL Trident Z5 Neo RGB DDR5 6400 MT/s 32GB (2x16GB)
Graphics: Internal
Operating System: Windows 11 Professional 23H2
System Drive: Kingston A2000 NVMe PCIe SSD 500GB
Test Drive: T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB (formatted empty)
Power supply:FSP Hydro PTM PRO ATX3.0 (PCIe5.0) 1200W
Case: STREACOM BC1 Benchtable V2
Graphics Card Driver: GeForce Game Ready 555.99

 

The first step is to check the basic information of T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB via CrystalDiskInfo software. It adopts PCIe 5.0 x4 transfer mode and NVM Express 2.0 standard, and the supported features include S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology), TRIM, and VolatileWriteCache. Supported features include S.M.A.R.T. (Self-Monitoring, Analysis, and Reporting Technology), TRIM, and VolatileWriteCache, and the firmware version tested by the author is EQFM22.3.

∆ CrystalDiskInfo software view information.

 

Use with empty disk CrystalDiskMark The software was used to test the read/write performance of the T-FORCE Z540 M.2 PCIe Gen5 x4 SSD 2TB in NVMe SSD configuration mode with the file size set to 1GiB, and the sequential read/write speeds measured in the preset profiles are 11956 MB/s and 10407 MB/s.

The Q8T1 test result in the first column represents a thread with eight queue depths, which means that there are eight 1MB-sized accesses in the worklist waiting to be accessed. This corresponds to a real situation, such as reading or writing eight 1MB files from or to a hard disk at the same time, which is generally less likely to be done in general.

The random read/write performance of the RND4K Q32T16 project is 5481 MB/s and 4143 MB/s.

The RND4K Q32T16 operates with 16 queue depths out of 32 threads, and the file types are random access 4 KB size files written or read into the SSD.

∆ CrystalDiskMark Settings:NVMe SSD / Profile:Default, Sequential Read/Write and Random Read/Write test scores.

∆ CrystalDiskMark Setting:NVMe SSD / Profile:Best Performance, Continuous Read/Write and Random Read/Write test results will be closer to the official specification.

 

For daily use or for gamers, the QD1 to QD4 range is a good reference. If we switch the CrystalDiskMark profile to real-world performance, the first column will be Q1T1. Although the result will be a bit lower than the Q8T1, it will be more in line with the actual performance of the daily use experience.

The reason for this is that most of the daily operating systems use the Q1T1, which is a queue-depth mode with one thread, so the Q1T1 is more suitable for daily use than the Q8T1. The read and write speeds of the Q1T1 are 8,194 MB/s and 9,316 MB/s with the test file size set to 1GiB in the Q1T1 mode.

∆ CrystalDiskMark Settings:NVMe SSD / Profile:Real World Mode 1GiB profile test results.

 

3DMark The storage benchmark test is conducted by game startup loading, copying game files, game archiving, installing games, and OBS game recording. The scenario mainly uses a number of games to conduct the actual test, so that gamers can clearly refer to the performance of the hard disk in the use of the game, and the above test is recorded in terms of time units, but the final results are calculated using the bandwidth and the average access time. The higher the score, the better.

∆ 3DMark Storage Benchmark Test.

 

以 PCMARK 10 To test storage performance for real-world applications, the Full System Drive Benchmark uses a range of everyday applications and software, including Windows 10, Adobe series software, game launcher, Microsoft clerical software and related applications, to test hard drive performance in real-world usage scenarios, and achieves a bandwidth of 69504 MB/s and an average access time of 37 µs. In the test, a bandwidth of 695.04 MB/s, an average access time of 37 µs, and a total test score of 4,438 points were achieved.

Data Drive Benchmark (Data Drive Benchmark), is the main test object for data disks used for storing documents, and it can also be used to test NAS, UFD and memory cards and other related types of storage devices. In this test, 339 JPEG files (totaling 2.37 GB) are copied to the SSD, and then a copy of the JPEG files is made, and finally 2.37 GB of JPEG files are copied to another hard drive to complete the three-step write-read-write-read test. In this test, 339 JPEG files (totaling 2.37 GB) were copied to the SSD, then copies of these JPEG files were made, and finally 2.37 GB of JPEG files were copied to another hard drive, completing the entire write-read-write-read three-step test, which resulted in a bandwidth of 1114.73 MB/s, an average access time of 19 µs, and a final total test score of 7843 points.

∆ PCMARK 10 Full System Drive Benchmark.

∆ PCMARK 10 Data Drive Benchmark.

 

A look at the M.2 SSD coolers used in the test

First of all, the GIGABYTE B650I AORUS ULTRA motherboard has the original third generation thermal armor supplemented with an active cooling fan, which is the standard configuration of the original cooling solution on the author's motherboard. Since the motherboard itself comes with a cooling solution, the M.2 SSDs can be installed directly without the need for too many procedures.

∆ The original third-generation cooling armor on the motherboard is complemented by active cooling fans.

There is a small active fan inside the ∆, with a special cooling groove design facing the processor to further utilize the airflow inside the chassis and enhance the efficiency of heat convection.

 

The second is the Jonsbo M.2 Radiator cooler, this is the author in the shrimp skin above to buy about one or two hundred dollars to get, is the only passive cooling module in this test of the selection, but because of the cheap and the effect is also good, so it seems that a lot of people have to buy this to use.

∆ All-aluminum Jonsbo M.2 SSD drive cooler.

∆ Slightly more troublesome to install and may require a little fumbling.

∆ Motherboard installation demonstration.

 

The third choice, T-FORCE DARK AirFlow I SSD Cooler, is an active air cooling cooler. The concept is to downsize the dual heatpipe air cooling tower and use it on M.2 2280 SSDs by using two Ø4 mm copper heat pipes with cooling fins to cool the M.2 SSDs, and then blowing the heat away directly with a Ø40 mm PWM fan. The heat is carried away directly by a Ø40 mm PWM fan.

The author has checked a little bit, but only Xinya in Taiwan has the name of "free gift", but the price is almost 2,300 dollars. I'm not sure if you have to buy any SSD to get the freebie, or if you can just buy the SSD alone.

∆ T-FORCE DARK AirFlow I SSD Cooler.

∆ The box comes with two thermal pads and a small screwdriver with installation instructions printed on it.

The ∆ T-FORCE DARK AirFlow I SSD Cooler powers the fan via PWM.

T-FROCE logo on the top of the ∆.

 

The T-FORCE DARK AirFlow I SSD Cooler requires the thermal pads to be in contact with the M.2 SSD in order for the heat to escape and complete the cooling process. The two included thermal pads have to be placed on the top and bottom of the M.2 SSD, making the overall installation process a bit cumbersome, so it is recommended to install them before attaching the motherboard to the chassis.

Then we have to mention its biggest disadvantage, the most direct impact is the volume of this thing, because there is a small air cooling tower dispersion in the M.2 SSD side, so for the CPU air cooling tower dispersion compatibility will be a big consideration, and the CPU water cooling piping direction is also surely to change the direction of the direction because of it.

∆ Dual heat pipes in direct contact.

∆ M.2 SSD Installation Demonstration.

∆ The actual use of the scene to show that this floor space ... ITX motherboard directly stuffed well stuffed.

 

This time I also borrowed a rather special cooler T-FORCE SIREN GD120S AIO SSD Cooler, which is an all-in-one water cooling cooler for M.2 2280 SSDs, I know it's not necessary but it's really interesting so I borrowed it to test it out (and play with it).

The overall concept is to transplant the 120mm all-in-one water cooling used on CPUs to M.2 SSDs, and from what I've seen, this model is supposed to use Apaltek's water cooling solution.

The author thinks that the bigger problem is the direction of the water-cooling pipe at the M.2 SSD water-cooling header, because it is in L-shape, although it can be adjusted by a large rotation, but with the common setup of the first slot of the motherboard M.2 SSD slot to support Gen5 bandwidth, the water-cooling pipe of the GD120S will either be stuck on the body of the graphics card, or else it will be stuck on the cooler of the processor, which is the most compromise is that two water-cooling pipes will be squeezed together on the left and right sides and become ugly to place. The best compromise is that the two water-cooling tubes are squeezed together on the left and right sides, which makes for an ugly arrangement.

∆ T-FORCE SIREN GD120S AIO SSD All-in-one water cooling.

The ∆PUMP is mounted on a 120 mm cooling plate, and from the details, it is assumed that it is an Apaltek water-cooling solution.

∆Only supports M.2 2280 SSDs with integrated water cooling head.

∆ Use a copper base with thermal pads to contact the M.2 SSD.

∆ Actual installation demonstration.

 

M.2 PCIE Gen5 x4 SSDs with various coolers in action

Since the Z540 isn't a new Gen5 SSD anymore, I'll just do a few performance tests for your reference, since many of you have unboxed it anyway. This time, the important part is the thermal performance test, which is conducted by installing the platform on the bare test bracket, so the actual cooling effect will be a little bit better than that in the chassis, and setting all the slots in the BIOS to run at full speed during the test. During the test, all slots were set to full speed in the motherboard's BIOS, and the actual test was conducted in a closed room with 24°C air conditioning, as the ambient temperature in a normal room is difficult to control, so this is for reference only.

The test software still uses CrystalDiskMark Settings:NVMe SSD / Profile:Default, but manually adjust the number of times to 9; test file size setting to 64 GiB, some foreign media will set their own scripts to stress test, but the author personally think that Taiwanese gamers are more commonly used. CrystalDiskMark software to test your own M.2 SSDs, so it would be better to continue using the same software to reproduce the tests on your own.

Temperature logging is done by using HWinfo64 software to record the maximum temperature of the hard disk, and manually adjusting the polling cycle in the HWinfo64 software to record the temperature of the M.2 SSD itself in a more real-time manner by adjusting the following settings. Why don't we use CrystalDiskInfo software to record the temperature? Because when there are multiple temperature sensors on an M.2 SSD, CrystalDiskInfo only displays the first sensor in order; sometimes the drive manufacturer will adjust the sensor temperature order so that the master chip is not always first, resulting in viewing temperatures from other locations. Furthermore, CrystalDiskInfo is very slow in updating the temperature sensors, and the difference between the displayed temperature and the actual temperature can be as much as 5~10 °C due to the lack of real-time updating of the temperature sensors.

• Global:20 ms
• Disk SMART every 1 cycle
• Embedded Controller every 1 cycle

∆ Last time someone said they wanted to see the test environment, so the motherboard was set up on a bare test platform .... There's nothing to photograph.

The height of the ∆Z540 PCB is shown in the diagram. The E26 controller chip is the tallest, and if it is a hard flat heatsink, it may not be able to touch the Flash particles and DRAM Cache.

 

In the actual test, it is very surprising that the original cooling armor on the GIGABYTE B650I AORUS ULTRA motherboard is actually the highest temperature, the author originally thought that it would have a small fan at least stronger than the Jonsbo M.2 SSD cooler, but the result is actually a surprisingly small loss of 3°C. The author also thought that the cooling armor on the B650I AORUS ULTRA motherboard is the highest temperature, but the author originally thought that it would have a small fan at least stronger than the Jonsbo M.2 SSD cooler, but the result is actually a surprising small loss of 3°C.

Considering that the cooling effect inside the chassis will be even worse, the motherboard's original AORTS M.2 Thermal Guard III and the passive cooling Jonsbo M.2 SSD cooler can at most be used inside the chassis to maintain the "overheating speed", and the lifespan of the drive will be a bit more worrying if it is written to for a long period of time under a high load (the author personally hopes that the maximum temperature of the SSD will not exceed 65 °C as a base point). (Personally, I would like to see the maximum temperature of SSDs not exceeding 65 °C as a benchmark).

The active air-cooled small tower T-FORCE DARK AirFlow I SSD Coole performs quite well, with a maximum temperature of only 53 °C. Compared to the previous two players, the AirFlow I fan running at full speed has 8000 RPM, and the noise is unacceptable, just like the AORTS M.2 Thermal Guard III. The noise is unacceptable, so we suggest you to choose the "Quiet Mode" to use the fan speed curve.

The T-FORCE SIREN GD120S AIO SSD Cooler with all-in-one water cooling has a maximum temperature of 48°C. I never expected to see a maximum temperature within 50°C at high loads on a PCIE Gen5 SSD, so it is true that the cooling performance of water cooling is still unquestionable, but considering the cost and installation hassle, it is only 5°C lower than AirFlow I. For "practical" gamers, it may not be as attractive as it is. However, considering the cost and installation hassle, it is only 5°C lower than the AirFlow I. It may not be so attractive to "practical" players, but the author's principle is: I want all the cool things! Anyway, it will be on the water eventually, so why don't we just start with ....?.

Test chart of ∆ Gen5 x4 SSD with different coolers.

The AORTS M.2 Thermal Guard III on the ∆GIGABYTE B650I AORUS ULTRA motherboard was tested using a thermal imaging camera to measure the surface temperature, but part of the lid is made of plastic (detection point #1), and the maximum temperature on the surface of the heatsink was around 43.9 °C. The temperature of the heatsink was around 10.5 °C, and the maximum temperature of the heatsink was around 10.5 °C.

The maximum temperature on the surface of the ∆ Jonsbo M.2 SSD hard disk cooler is about 50.3 °C.

The surface temperature of ∆ AirFlow I is basically only about 33.2 °C. The surface temperature of ∆ AirFlow I is only about 33.2 °C.

 

Conclusion

In the face of PCIE Gen5 SSDs, both motherboard brands and gamers are paying more attention to the heat dissipation of M.2 SSDs, but as for the future, do we need to go for a four-heat-pipe tower or an all-in-one water-cooler like the one pictured here? I think only a few extreme gamers will play around with this, but most importantly, it's still too troublesome and impractical to use.

Most SSD brands have decided to leave it to the motherboard manufacturers to handle. Micron has launched a version of the T700/T705 with its own heatsink, but I have actually played with the T700 version with a heatsink, and the maximum temperature performance could not reach my personal expectation. If I really want to buy a PCIE Gen5 SSD, I will choose the version without a heatsink and use it through the motherboard or other brands of active coolers to extend its lifespan and maintain its high speed read/write performance.

 

In this test, the best performer is indeed the all-in-one water-cooled GD120S, followed by the active air-cooled AirFlow I, plus the cheapest passive cooling Jonsbo M.2 SSD cooler with a maximum temperature lower than the motherboard's default solution, but there are a few issues to note with these "add-on" coolers.

That is, motherboards now have a convenient M.2 SSD removal of DIY convenience clips and other devices, the height of these clips are almost always designed for M.2 SSD bare bar directly fixed, with the motherboard itself designed for the installation of the heatsink, the use of these additional mounted heatsinks are very likely due to the thickness or height of the problem, and must be changed back to using the M.2 SSD screws to lock fixed the entire set of SSDs.

Nowadays, motherboards are also equipped with bottom thermal pads and thermal armor for the first M.2 SSD expansion slot to help dissipate the heat from the double-sided particles, so if you want to use an additional cooler, you have to remove the original thermal pads, or even remove the motherboard's complimentary M.2 SSD bottom armor, to get a better compatibility for installation.

The motherboard itself provides the worst cooling solution, but it meets the principle of not heating up and speeding down and does not require so much disassembly, which makes it the most convenient choice for lazy people.

 

In conclusion, if you want to get the best cooling effect from PCIE Gen5 M.2 SSDs, you should use an active cooler. However, these active M.2 SSD coolers have problems such as large size, which will affect the installation space of CPU coolers or graphics cards, causing compatibility problems with each other, so if users can solve these problems perfectly or just ignore them, it is highly recommended to use an active cooler for this stage. It is highly recommended to use an active cooler, and many motherboards come with accessories for this purpose.

Finally, I would like to remind you that the temperature results are for reference only, most people actually use it installed in the chassis and use it with a graphics card, when the waste heat from the graphics card operation is discharged upwards, the neighboring M.2 SSD cooler 100% on the floor of the graphics card will be affected, and the M.2 SSD overheating and slowing down due to the waste heat from the graphics card will result in a poor performance, the above information is for your reference. The above information is for your reference.