Last month, Sony quietly released a new revision of the PS5 that weighs about 300 grams less than the original model. Why, you ask? Because the new refreshed model has a smaller heatsink, that gives less cooling performance, but there's more to it than just that.

The new PS5 model was tested first by YouTuber Austin Evans in a YouTube video, the temperature test showed that the new PS5 was ~3-5°C hotter than the old model, and concluded that the new heatsink is a sort of a downgrade. The video generated some controversy mainly because he used his phone to measure out the temperatures, and the lack of proper testing.

Now more testing has been done by two more channels, HarwareBusters and GamerNexus. The combined result from both shows that the new PS5 is actually an improvement; the CPU, APU, VRM & memory all run cooler than the old PS5, with 6 to 20°C difference between the two.

Much more detailed explanations are on the YT videos below.

Source

 

[TR_mahmutpek]

Gamer nexus said that launch ps5 has bad cooling on vrm. Glad to see this fixed.

 

[AkiraKurusu]

Now to wait until this actually becomes available, what with the chip shortage and scalpers...

 

[tech3475]

Now to wait until this actually becomes available, what with the chip shortage and scalpers...

Right now, working retail has been the easiest way I’ve found. I could buy one today if I wanted to (no joke), don’t know about the revision though.

I honestly wont be surprised if most of the Christmas temps this year join just to try and get one.

 

[Foxi4]

Hasty conclusion on the part of Hardware Busters. In order to measure the new design fairly you have to use the exact same fan and PCB, which GN has done in their testing. To my eye it looks like the SOC temperature has increased, which is an expected consequence of reducing thermal mass of the heat sink, *but* the temperature has also decreased in other areas of interest such as the VRM and the memory, which shows that overall placement of fin arrays and heat pipes is more optimal. Whether that’s good or not remains to be seen, however I will point out that both memory and VRM’s are quite happy running at very high temps whereas the SOC is not. It’s also not particularly representative in console since Sony favours noise level over temperature in software, so both sinks will run more or less within margin of error regardless.

 

[Spider_Man]

Lack of proper testing and grammar.... and a smartphone lol

No one says -xx "hotter" to try, toud report it runs xx coller compared to the launch model.

And just because something is smaller, doesnt mean it is less sufficient.

Cant see sony adding inferior cooling hardware when all these units will just over heat and be sent back.

Its not like nintendont pushing a crappy tablet clone with shit fans when its been pushed to run watered down games and out of all the revisions thus far, they've not coughed up any dollar to provide an improvement cooling system........ but are happy to hike the price up for a cheap shitty tiny OLED screen that isnt going to provide any benefits to the gamer, when your running inferior games at lower resolution and frames on a tiny OLED, when your likely to have at least 40"+ OLED 4k+ HDR tv at home.

Sony aren't like nintendont who find any means to keep crapping out new revisions, knowing its loyal fansheep will blindly rush out to buy each irritations of nintendont consoles.

 

[AkiraKurusu]

WAIT - I just remembered there's both a discless PS5 and one with the disc drive.
Is that still a thing with this revision?

 

[Frankbel]

I wait for the midgen revision.

 

[linuxares]

TLDR; Just get which ever.

 

[r5xscn]

Hasty conclusion on the part of Hardware Busters. In order to measure the new design fairly you have to use the exact same fan and PCB, which GN has done in their testing.

Afaik Aris from HB said their test units come with the same exact fan model (check at 4 minutes mark). The PCB (or the fan controller with the configuration to be more precise) might be different tho.

 

[smf]

Cant see sony adding inferior cooling hardware when all these units will just over heat and be sent back.

They might, the fat ps3 had a lot of issues. Hopefully they've learned from their mistakes.

Certainly you can't just look at the temperature of the air coming out of the back of a ps5 and determine whether it's better or not.

If the same amount of heat is being generated inside the ps5 then the one with the hottest air is actually the better working one.

 

[Deleted member 568892]

It uses a bit more power said the Digital Foundry video/.

 

[Foxi4]

Afaik Aris from HB said their test units come with the same exact fan model (check at 4 minutes mark). The PCB (or the fan controller with the configuration to be more precise) might be different tho.

Testing two separate units is not a like for like comparison due to silicon lottery and manufacturing tolerances, you test heat sinks on the exact same test bench in a controlled environment. If you’re using two separate machines then the test is void by default, even if you have “the same model fans and PCB’s”. You should use the same fan and the same test machine since there may be differences in the fan windings and so on. You have to eliminate as many variables as possible - the same PCB and fan are a must. From what I gathered, HB is testing two different systems at a different ambient temperature without a fan/PCB swap, so their results only apply to their specific machines and are not a measurement of performance difference between the old and new design, there are too many variables that were not controlled for. The GN video I attached is more representative.

 

[spotanjo3]

I am glad that I have an old version of PS5. A lower is better!

 

[Foxi4]

Lack of proper testing and grammar.... and a smartphone lol

No one says -xx "hotter" to try, toud report it runs xx coller compared to the launch model.

And just because something is smaller, doesnt mean it is less sufficient.

Cant see sony adding inferior cooling hardware when all these units will just over heat and be sent back.

Its not like nintendont pushing a crappy tablet clone with shit fans when its been pushed to run watered down games and out of all the revisions thus far, they've not coughed up any dollar to provide an improvement cooling system........ but are happy to hike the price up for a cheap shitty tiny OLED screen that isnt going to provide any benefits to the gamer, when your running inferior games at lower resolution and frames on a tiny OLED, when your likely to have at least 40"+ OLED 4k+ HDR tv at home.

Sony aren't like nintendont who find any means to keep crapping out new revisions, knowing its loyal fansheep will blindly rush out to buy each irritations of nintendont consoles.

Sony wanted to avoid another YLOD disaster, so the original heat sink is overkill. It was redesigned for the purposes of cost savings and weight reduction. It’s doing a good enough job and although the SOC runs hotter, memory and VRM seem to be cooler. With that being said, the differences are minute.

 

[AlexMCS]

still waiting for that PS5 slim, preferably water cooled.

 

[Arck]

will wait for the PS6, a bit disappointed for what you get for 500$ with the PS5, am still fine with my PC right now

 

[smf]

still waiting for that PS5 slim, preferably water cooled.

You can water cool them now, I'm not sure of your chances getting sony to supply you one though.

I think we're too close to limits to see an easy die shrink, like ps3/ps4

 

[linuxares]

still waiting for that PS5 slim, preferably water cooled.

oh god no.... you think that company would be able to do a water cooling?

 

[r5xscn]

Testing two separate units is not a like for like comparison due to silicon lottery and manufacturing tolerances, you test heat sinks on the exact same test bench in a controlled environment. If you’re using two separate machines then the test is void by default, even if you have “the same model fans and PCB’s”. You should use the same fan and the same test machine since there may be differences in the fan windings and so on. You have to eliminate as many variables as possible - the same PCB and fan are a must. From what I gathered, HB is testing two different systems at a different ambient temperature without a fan/PCB swap, so their results only apply to their specific machines and are not a measurement of performance difference between the old and new design, there are too many variables that were not controlled for. The GN video I attached is more representative.

I just noted that the fans are the exact same model. I agree with your comment to some degree. But the point here is that they are testing the whlle system and not just the heatsink. If the heatsink is smaller but the pcb is better then there should be no issue for users. Of course silicon lottery still applies. But I am not going to nitpick on small details like that because those are still within the manufacturing variances. As Aris said the sample size is too small to draw any useful conclusion.

In my opinion, both tests are flawed because of the sample size. Additionally, GN said their original model might not be representative because they reapplied the liquid metal.

I am not really sure regarding testing on different ambient temperatures, but if Aris can't nail the exact ambient temperatures for both systems, then I doubt GN can. Aris is a well known PSU reviewer and testing PSU in hot and cold condition is a basic thing for a PSU test.

If you want more data, you might also want to add Igor's review.

In summary, they are testing the systems and not just the heatsinks. Sample size needs to be bigger.