• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

New PS5 model lighter, uses less W, and possibly more efficient in cooling

Maybe there is just better chip selection used for the newer models. So they can reduce voltages and therefore reduce the cooling solution. I wouldn't thing they use 6nm as it would be more expensive, it would have also reduced the die size.
Maybe it is smaller. The slanted plate (that is applied directly on the back of the APU) does look smaller on the pics. But maybe the new shape is not indicative of the APU behind, we'll see.
Wsg9qXI.png


I wouldn't thing they use 6nm as it would be more expensive
The price increase happened exactly at the same time of 1200 release.
 
Last edited:

DonJimbo

Member
It's not going to get smaller any time soon because Sony is making all these nifty faceplates that they want people to pay them 50 bucks for.
It makes me sad have to wait a bit longer to get the slim revision
 
Last edited:
I'm still wondering why Austin didn't removed the heatsink and tried to compare and see if the chip has gotten smaller.

I mean who the hell gets this far in a teardown,
R2j5oW5.jpg

saw it had a new motherboard and doesn't continue? SMH.
Someone who has no clue what he is doing and is afraid to tear down more.
 

Allandor

Member
Wouldn't the smaller CPU make the console cheaper as it would use less silicon? I thought that was one of the main reasons games consoles drop in price
Not really. As the production methods get more expensive. A smaller Die wouldn't compensate that at these levels. But a new process step might compensate it if other parts can also get smaller. E.g. the cooling solution can get smaller (and therefore most of the times cheaper), the power supply might get smaller (and output less power) (-> cheaper parts), the mainboard might even get smaller as less intensive voltage regulation might be needed, ...
But process steps these days got really, really expensive.
 

M1chl

Currently Gif and Meme Champion
I'm still wondering why Austin didn't removed the heatsink and tried to compare and see if the chip has gotten smaller.

I mean who the hell gets this far in a teardown,
R2j5oW5.jpg

saw it had a new motherboard and doesn't continue? SMH.
That will be in next episode to keep the clout going
 
No. From my fuzzy memory it was about 30 to 50% heavier than PS5 one (1000 model). It's probably like more than twice heavier than the 1200 model now!
Yes the Series X heatsink is heavier, but that does not equal more expensive
Whatch the GN video on cooler master heatsink production.
The more complex the design the more expensive. Series X heatsink is extremely simple to produce.
 
I'm still wondering why Austin didn't removed the heatsink and tried to compare and see if the chip has gotten smaller.

I mean who the hell gets this far in a teardown,
R2j5oW5.jpg

saw it had a new motherboard and doesn't continue? SMH.

Someone who has no clue what he is doing and is afraid to tear down more.
Because he wants to do another video to milk the story and make more money. Of course he has already done it.
Just not uploaded yet.
 

Loxus

Member
Because he wants to do another video to milk the story and make more money. Of course he has already done it.
Just not uploaded yet.
If that's the case, he'll also run the risk of losing out on some views if someone else gets a video out on the motherboard before him.
 

jaysius

Banned
Now that the internal components are trimming down the external bulky fugly shape is making less and less sense.

Hopefully the PS5 slim will look like something that makes sense in a home theater setup rather than a 5 year old failed art project.
 

John Wick

Member
Maybe there is just better chip selection used for the newer models. So they can reduce voltages and therefore reduce the cooling solution. I wouldn't thing they use 6nm as it would be more expensive, it would have also reduced the die size.
So what your saying is they are wasting chips as they are using a better selection? So what are they doing with the worse selection of chips?
It's more likely they are using 6nm as their was an article where Sony booked more 6nm wafers with TSMC that's why PS5 has been in better supply.
 
Yes the Series X heatsink is heavier, but that does not equal more expensive
Whatch the GN video on cooler master heatsink production.
The more complex the design the more expensive. Series X heatsink is extremely simple to produce.
More copper means more expensive. But that's not the end of the problem on Xbox.

Have you even looked at the whole cooling solution on Xbox ? It's a very complicated, heavy and expensive stuff compared to the very simple and classic system on PS5 (minus liquid metal stuff). What was expensive on PS5 was the RD for the innovative liquid metal solution (2 years of tests). But the rest is cheap (2 or 3 times less copper PS5 1200 vs XSX) and classic stuff. Also the paste on XSX means the system will inevitably make more noise along the years...
JFPinmx.png
 
Last edited:

jm89

Member
Then why it's still so fucking huge, they should shrink the enclosure.
Gotta sell those covers, or else gotta have to go through the ball ache of creating different sized ones.

I'd think they are probably waiting for an even further reduction to really create a slim model with a reduced enclosure.
 
More copper means more expensive. But that's not the end of the problem on Xbox.

Have you even looked at the whole cooling solution on Xbox ? It's a very complicated, heavy and expensive stuff compared to the very simple and classic system on PS5 (minus liquid metal stuff). What was expensive on PS5 was the RD for the innovative liquid metal solution (2 years of tests). But the rest is cheap (2 or 3 times less copper PS5 1200 vs XSX) and classic stuff. Also the paste on XSX means the system will inevitably make more noise along the years...
JFPinmx.png

Watch this:


And watch this:


The latter part is how the Series X heatsink is made.
The former is how the 4 different heatsinks for the PS5 are made.
The original 1000 PS5 had even more copper than the Series X. Just now with the 2000 Model, the amount of material is vastly reduced.


Also this is the heatsink of the Series X:
gsmarena_010.jpg

This is the original PS5 heatsink:
the-ps5s-heatsink-size-is-absolutely-massive-1024x597.jpg




The 1000 PS5 heatsink was vastly more expensive than the Series X heatsink.
The 2000 PS5 heatsink might be cheaper then the Series X heasink. But only on a single unit cost basis.
They had 3 revisions already they needed expensive tooling and R&D for.
A metric ton copper is $7.5k currently. Was as high as 10k during peek commodity inflation tho.
If you talk about the alu mainframe of the Series X, yes that is quite expensive.
 
Last edited:

PaintTinJr

Member
The less-power - on the new PS5 -when playing games seems like it might break Sony's variable clock (roughly) constant power design paradigm rule, unless that the "roughly constant power" is actual just an upper limit of power draw for a workload at the determined variable frequency, and draws less if it needs less, or they are saving the power from improvements in the external hardware components (outside the APU) or are using a new revision of the logic system to determine variable clock from workload presented at set power. The change to the cooling system suggests to me the APU is the source of the saved power .

My assumption of how they could have saved 20-30watts of power for the APU - without a node change or without an APU redesign - was that the original APUs were effectively working prototypes and that the binning of the first console chips - at the performance level we got - resulted in chips that were largely just making the grade, and as fab-ing has perfected over millions of chips made, all the chips now are able to bin at much higher clocks than needed, so instead of using them at higher clock they've been able to under-volt a decent amount, and save a decent amount of power.
 

01011001

Banned
The less-power - on the new PS5 -when playing games seems like it might break Sony's variable clock (roughly) constant power design paradigm rule, unless that the "roughly constant power" is actual just an upper limit of power draw for a workload at the determined variable frequency, and draws less if it needs less, or they are saving the power from improvements in the external hardware components (outside the APU) or are using a new revision of the logic system to determine variable clock from workload presented at set power. The change to the cooling system suggests to me the APU is the source of the saved power .

My assumption of how they could have saved 20-30watts of power for the APU - without a node change or without an APU redesign - was that the original APUs were effectively working prototypes and that the binning of the first console chips - at the performance level we got - resulted in chips that were largely just making the grade, and as fab-ing has perfected over millions of chips made, all the chips now are able to bin at much higher clocks than needed, so instead of using them at higher clock they've been able to under-volt a decent amount, and save a decent amount of power.

my first thought was maybe different memory chips and a more efficient PSU
 

Loxus

Member
The less-power - on the new PS5 -when playing games seems like it might break Sony's variable clock (roughly) constant power design paradigm rule, unless that the "roughly constant power" is actual just an upper limit of power draw for a workload at the determined variable frequency, and draws less if it needs less, or they are saving the power from improvements in the external hardware components (outside the APU) or are using a new revision of the logic system to determine variable clock from workload presented at set power. The change to the cooling system suggests to me the APU is the source of the saved power .

My assumption of how they could have saved 20-30watts of power for the APU - without a node change or without an APU redesign - was that the original APUs were effectively working prototypes and that the binning of the first console chips - at the performance level we got - resulted in chips that were largely just making the grade, and as fab-ing has perfected over millions of chips made, all the chips now are able to bin at much higher clocks than needed, so instead of using them at higher clock they've been able to under-volt a decent amount, and save a decent amount of power.
I think your digging to deep into this.

These are the results from Austin's video.
1000: 218W
1100: 229W
1200: 202W

There is no difference between 1000 and 1100 chips, yet the 1000 consumes around 11W less then the 1100.

The only difference between the 1000 and the 1100 is the heatsink. I'm not 100% sure but the PS5 fan can consume as much as 29W depending on the model.
DC12V / 1.9A = 22.8W and 12V / 2.4A = 28.8W

That plus the fan speed required to work with the heatsink maybe the difference in power consumption between the revisions.

Imo, we need someone more professional like Gamers Nexus to do these kind of test.
 
Last edited:

Loxus

Member
Imagine coming into a thread DISCUSSING THE PS5 HARDWARE and talking about Sony ponies. SMH.
Reaction Smh GIF by Lil Yachty
 
Last edited:

Allandor

Member
So what your saying is they are wasting chips as they are using a better selection? So what are they doing with the worse selection of chips?
It's more likely they are using 6nm as their was an article where Sony booked more 6nm wafers with TSMC that's why PS5 has been in better supply.
E.g. use them with different power-profiles or stuff them in systems with the "old" design. Over time normally production gets better so they might just be at a point where they can better test the APUs or they now have just stricter tests if a APU can work with the new power targets. I guess especially at the start they used higher voltages just to be sure everything works with (almost) every chip. But currently, not the chip-production itself might be the bottleneck for the production, it might just be other parts. So they can more strictly select and therefore safe some money on the other parts (that might be the bottleneck for the production of the console).

Yes, those are just my guesses and no facts at all.
 

PaintTinJr

Member
my first thought was maybe different memory chips and a more efficient PSU
That might have happened too IMO, because from my time tinkering with Raspberry Pis and pushing their memory to its capacity limits - while overclocking and running heavy multi-core program - heat dissipation on memory is superior using bigger chips (nearly full) than twice as many smaller chips (nearly full) - as the spare capacity on a bigger chip ends up as one chunk giving slightly more thermal headroom proportionally IMO (going by my tests), and the larger surface area proportionally dissipates heat better too.
 

M1chl

Currently Gif and Meme Champion
Watch this:


And watch this:


The latter part is how the Series X heatsink is made.
The former is how the 4 different heatsinks for the PS5 are made.
The original 1000 PS5 had even more copper than the Series X. Just now with the 2000 Model, the amount of material is vastly reduced.


Also this is the heatsink of the Series X:
gsmarena_010.jpg

This is the original PS5 heatsink:
the-ps5s-heatsink-size-is-absolutely-massive-1024x597.jpg




The 1000 PS5 heatsink was vastly more expensive than the Series X heatsink.
The 2000 PS5 heatsink might be cheaper then the Series X heasink. But only on a single unit cost basis.
They had 3 revisions already they needed expensive tooling and R&D for.
A metric ton copper is $7.5k currently. Was as high as 10k during peek commodity inflation tho.
If you talk about the alu mainframe of the Series X, yes that is quite expensive.

You should also get into consideration that aluminium chunk which holds those two boards together, that cannot be cheap. The whole thing (XSX) was heavier than PS5 OG.
 
You should also get into consideration that aluminium chunk which holds those two boards together, that cannot be cheap. The whole thing (XSX) was heavier than PS5 OG.
I did? And yes, it's not cheap. One of the most expensive non electronic parts of the console. More expensive than the heat sink for sure.
Not only because of the weight, but that is super custom and hard to manufacture.
If you talk about the alu mainframe of the Series X, yes that is quite expensive.
But it is a dual purpose functioning part of the console for structural integrity, too.
I was talking about heatsink costs.

If we want to take everything into account it becomes to complicated and impossible to estimate on a single unit costs basis.
PS5 has 2 metal frames used for passive cooling, too. Even tho they're just thin sheets and not a big block. Liquid cooling and all the barriers needed to not make it leak, are additional added costs.
 
Last edited:

PaintTinJr

Member
I think your digging to deep into this.

These are the results from Austin's video.
1000: 218W
1100: 229W
1200: 202W

There is no difference between 1000 and 1100 chips, yet the 1000 consumes around 11W less then the 1100.

The only difference between the 1000 and the 1100 is the heatsink. I'm not 100% sure but the PS5 fan can consume as much as 29W depending on the model.
DC12V / 1.9A = 22.8W and 12V / 2.4A = 28.8W

That plus the fan speed required to work with the heatsink maybe the difference in power consumption between the revisions.

Imo, we need someone more professional like Gamers Nexus to do these kind of test.
I don't think I am mainly because, I personally don't think there is 16watts to save between the 1000 and 1200 models- without APU under-volting.

If we consider the PSU first, IIRC the PS5 PSU is already achieve 80 Plus Gold certification standards, and if I further remember, the difference between Gold and Platinum in the 80% load range is only 2-3% gain in efficiency, so at best upping the costs for the PSU to get the 3% of 218watts, is just 6.5watts. And the bigger issue here is improving your PSU doesn't alter the thermals of the APU - to provide any headroom to reduce the heatsink or fan size - it is purely an improvement of efficiency of taking AC power and converting it to DC, so it will still present the same 218watts DC to the PS5, just from(iirc) ~235watts AC, rather than ~240watts AC (working on ~90% and 93% efficiencies at 80% load or less).

Then you've got the change in heatsinks. The difference between the 1000 and 1100 power use could well be the change in fan and heatsink if the original was over-engineered and just scaled back because with 15-20million made and sold they have enough user data running software to say the parts can be reduced, however in that scenario where the APU is using the same power and dissipating the same wasted energy as heat, the 1100 consoles would need to either produce more noise to cool the APU with a less effective cooling system, let the system ambient exhaust heat increase, or need to have improved the transfer rate of the liquid metal cooling - which I think Sony might have talked about. So no scenarios I believe have happened.

As I understand it, a fabbed processor at early yields is designed with redundant pathways to help yields, and as the process produces chips overtime that don't need the redundant features because they all bin perfectly at the original spec, chip resistant to power lowers and thermals improve - allowing for a redesign that can remove those redundant elements and further improve the chip even on the same node.

We know within the PS5 APU there was a certain amount of CU redundancy and on any of the chips with a faulty CU they probably get hotter (increasing chip resistant to power) needing the higher power level to be sure they work as required, whereas the newer chips should run cooler and provide an option to under-volt - further generating less heat - which has then given headroom for a smaller heatsink, a smaller fan and cooler and quieter running.
 

Loxus

Member
The whole PS5 is built around an unknown budget. I don't think that power budget will change unless it gets a die shrink.

Besides, this is from PlayStation.
Active Power Consumption
PlayStation®5 model number: CFI-1016A
HD
  • Active gaming (PS5 games: three game average): 196.9 W
  • Active gaming (PS4 game: Marvel’s Spider-Man): 107.1 W
  • Active gaming (PS4 game: Battlefield 4): 90.3 W
  • DVD playback: 54.1 W
  • Blu-ray playback: 53.0 W
  • Streaming media: 55.6 W
  • Home menu user interface: 43.1 W
UHD
  • Active gaming (PS5 games: three game average): 197.7 W
  • Active gaming (PS4 game: Marvel’s Spider-Man): 107.9 W
  • 4K Blu-ray playback: 75.7 W
  • Home menu user interface: 44.4 W
Low power
  • Rest*: 0.3 W
  • Rest (Supply Power to USB Ports): 3.7 W
  • Rest (Stay Connected to the Internet): 1.0 W
  • Rest (All): 3.7 W
Tested: 15/12/2020


PlayStation®5 model number: CFI-1116A
HD
  • Active gaming (PS5 games: three game average): 199.0 W
  • Active gaming (PS4 game: Marvel’s Spider-Man): 104.7 W
  • Active gaming (PS4 game: Battlefield 4): 90.5 W
  • DVD playback: 54.1 W
  • Blu-ray playback: 53.3 W
  • Streaming media: 54.1 W
  • Home menu user interface: 44.0 W
UHD
  • Active gaming (PS5 games: three game average): 201.1 W
  • Active gaming (PS4 game: Marvel’s Spider-Man): 108.5 W
  • 4K Blu-ray playback: 75.5 W
  • Home menu user interface: 45.5 W
Low power
  • Rest*: 0.36 W
  • Rest (Supply Power to USB Ports): 3.1 W
  • Rest (Stay Connected to the Internet): 1.2 W
  • Rest (All): 3.2 W
Tested: 16/12/2021

These are proper tests done that show there's not much of a difference between the 1000 and 1100 and power consumption varies.

Imo, Austin's test should not be always taken seriously.

Edit:
Digital Foundry test shows the PS5 doesn't go over 203W.



This other video also shows Astro's Playroom power consumption under 200W on the OG PS5.



I don't know what to say about those Austin Evans tests.
 
Last edited:

bboris77

Neo Member
I am VERY happy to report that my brand new Digital Edition CFI-1215B model is extremely quiet, with very minimal coil whine. It really is a night and day compared to my launch CFI-1000B model. You can still hear a very minimal amount of coil whine when you put your ear close to the unit, but it is absolutely inaudible from a few feet away. The original model was clearly audible from across the room and would only be drowned out by cranking up the sound on the TV/Amplifier.
 

Melchiah

Member
It can be fixed. Xbox series x puts a dampening goo on the components to prevent coil whine and it works great. Series X is the quietest console I've ever owned. By contrast, the coil whine in the PS5 makes it by far the loudest console I've owned.

Talk about hyperbole. It's nowhere near the Hoover level of PS4 Pro, especially when accessing menus/maps and the fan goes overdrive, even with a coil whine. All I can I hear is a very low hum, when I put a game on pause. It can't be heard when playing, and I sit only three meters away from the system.

It's good thing they set a power draw threshold for the PS5 to avoid the Pro issue. It's beyond me why the issue (of unlocked framerates in menus/maps) was never fixed for Pro. I still play Let It Die on the Pro, as I've heard it has issues on the PS5, which didn't put the system on a Hoover state until I bought a 4K TV. Now it's very loud in some of the menus, closed areas with multiple enemies, and when you're idling in the base. It's odd, since the game featured supersampling on 1080p displays, and was pretty silent, at least on Pro standards.

EDIT: Forgot to mention, that forcing the system to output at 1080p didn't fix the issue either.
 
Last edited:

Tams

Member
All this at no extra cost...

... oh wait, no, that's not right. The PS5 now costs £30.00 more here in the UK and is more expensive in many other territories as well bar the U.S.

Oh, cry me a river. Prices have gone up across the board, and while I know times are hard, an extra 30 quid is not going to make any difference. If that's an issue for you, you should probably consider if getting any console is a priority.
 

Rykan

Member
Talk about hyperbole. It's nowhere near the Hoover level of PS4 Pro, especially when accessing menus/maps and the fan goes overdrive, even with a coil whine. All I can I hear is a very low hum, when I put a game on pause. It can't be heard when playing, and I sit only three meters away from the system.
You are aware that the amount of coil whine can vary greatly from console to console, right?
 

Rykan

Member
Yes, and I find it hard to believe it could even approach the Hoover level of PS4 Pro.
I've owned two different PS4 pro's and neither of them came even anywhere close to the coil whine that my PS5 was producing when running a game at 60 fps or higher.
 

Melchiah

Member
I've owned two different PS4 pro's and neither of them came even anywhere close to the coil whine that my PS5 was producing when running a game at 60 fps or higher.
You are aware that the amount of PS4 Pro's fan noise can vary greatly from console to console, right? I can hear mine three rooms away, even when I'm just idling in Let It Die's base. Only PS3 fatty comes close. Can you say the same about the coil whine?

As for "Series X is the quietest console I've ever owned", I find that also a hyperbolic statement, since the 32-bit systems like PS1 were pretty much dead silent. Let alone the earlier gens.
 
Last edited:

Rykan

Member
You guys don't know what coil whine is.
Yes I do.
You are aware that the amount of PS4 Pro's fan noise can vary greatly from console to console, right? I can hear mine three rooms away Can you say the same about the coil whine?

As for "Series X is the quietest console I've ever owned", I find that also a hyperbolic statement, since the 32-bit systems like PS1 were pretty much dead silent. Let alone the earlier gens.
I've owned two different PS4 pro's and neither of them came even anywhere close to the coil whine that my PS5 was producing when running a game at 60 fps or higher.
As for the Series X part, I'm obviously talking about systems that actually have cooling fans in them or mechanics that are capable of producing extensive noise. The only noise that a PS1 could create is the disk drive.
 

Putonahappyface

Gold Member
I am VERY happy to report that my brand new Digital Edition CFI-1215B model is extremely quiet, with very minimal coil whine. It really is a night and day compared to my launch CFI-1000B model. You can still hear a very minimal amount of coil whine when you put your ear close to the unit, but it is absolutely inaudible from a few feet away. The original model was clearly audible from across the room and would only be drowned out by cranking up the sound on the TV/Amplifier.
tom hardy bane GIF


I have the original disc version from release and I can't hear anything except the subtle sound of the fan at maybe two feet away and it's whisper quiet!
 
Last edited:

Melchiah

Member
Yes I do.


As for the Series X part, I'm obviously talking about systems that actually have cooling fans in them or mechanics that are capable of producing extensive noise. The only noise that a PS1 could create is the disk drive.
You said before, that the coil whine is widely reported. The very same applies to Pro's fan noise. It's odd to dismiss the same. Especially when people thought the PS5 would be as bad before the launch. The coil whine might be annoying, but I doubt you could hear it three rooms away like the Pro's fan noise.
 

M1chl

Currently Gif and Meme Champion
I am VERY happy to report that my brand new Digital Edition CFI-1215B model is extremely quiet, with very minimal coil whine. It really is a night and day compared to my launch CFI-1000B model. You can still hear a very minimal amount of coil whine when you put your ear close to the unit, but it is absolutely inaudible from a few feet away. The original model was clearly audible from across the room and would only be drowned out by cranking up the sound on the TV/Amplifier.
This type of noise?



Timestamped
 

Tams

Member
This type of noise?



Timestamped


Yep. Many electronics have it.

It's inductors vibrating at an audible frequency. It can be affected by the amount of current going through the inductor. If the electromagnet is poorly made (e.g. badly coiled), then it is more likely to cause coil whine.

Unfortunately, there's not much you can do about it if the product isn't customisable (like a graphic card is - you can change the frequency), and even then, you may have to make some sacrifices to prevent coil whine (like not drawing a lot of power and therefore not getting the highest framerates).

The best option is almost always to return the product if it annoys you.

 

M1chl

Currently Gif and Meme Champion
Yep. Many electronics have it.

It's inductors vibrating at an audible frequency. It can be affected by the amount of current going through the inductor. If the electromagnet is poorly made (e.g. badly coiled), then it is more likely to cause coil whine.

Unfortunately, there's not much you can do about it if the product isn't customisable (like a graphic card is - you can change the frequency), and even then, you may have to make some sacrifices to prevent coil whine (like not drawing a lot of power and therefore not getting the highest framerates).

The best option is almost always to return the product if it annoys you.

Just checking if this is what poster wanted to say : ) But it's always nice to have some knowledgeable poster : )
 

Rykan

Member
You said before, that the coil whine is widely reported. The very same applies to Pro's fan noise. It's odd to dismiss the same. Especially when people thought the PS5 would be as bad before the launch. The coil whine might be annoying, but I doubt you could hear it three rooms away like the Pro's fan noise.
I'm gonna be honest with you mate, I have no idea what you're trying to argue here. The coil whine is widely reported. What point are you trying to make here? That the PS4 pro is louder? It certainly isn't from my experience.

You can doubt all you want. The feeling is mutual, because I don't believe for even a second that you can hear your PS4 pro from three rooms away. The fans are nowhere near large enough for that to even be possible.
 
Top Bottom