I hope the new PS5 cooler still works well... awaiting more testing

[R-T-B]

cause I am NOT comfortable with my console hitting 90-95 Celsius on the VRM's

VRMs are designed for temps well above that.

 

[terroralpha]

i'm going to play devil's advocate here and say that it's possible that they did it because of the rising cost of raw materials. speaking from my work experience, our cost for raw aluminum has nearly doubled in the last year. just in Q1 of this year alone it went up 30%. i think sony is just doing what they have to to make as many consoles as possible without raising prices.

 

[Vya Domus]

Update from DF:

I don't understand what they are talking about, this doesn't make any sense, the algorithm takes into account temperature and power readings. It's not governed by the cooler but it's governed by things affected by the effectiveness of the cooling.

Plus, how are they even able to tell that the frequencies will not change, do they have debug consoles that display this information ?

 

[Chomiq]

I don't understand what they are talking about, this doesn't make any sense, the algorithm takes into account temperature and power readings. It's not governed by the cooler but it's governed by things affected by the effectiveness of the cooling.

Plus, how are they even able to tell that the frequencies will not change, do they have debug consoles that display this information ?

You're talking about boost algo on modern GPU. They're talking about performance on a PS5 APU. More details:

On PCs, boost is controlled by temperature and current readings off the SoC, but these are, in turn, impacted by factors like part binning and ambient temperature. Sony’s boost mode doesn’t take these factors into consideration. Instead, CPU and GPU clock rates are determined based on monitored system activity. Instead of monitoring the performance of your specific SoC, the boost model Sony deploys is based on a model SoC. Every console uses the same model for how and when to boost frequencies, keeping performance identical even when air temperature or power draw might allow some chips to push higher. This keeps performance identical, regardless of ambient temperature.

So if it's PS5 it's going to boost like a PS5 should and not as high as cooling allows it for your specific PS5.

 

[Vya Domus]

You're talking about boost algo on modern GPU. They're talking about performance on a PS5 APU. More details:

So if it's PS5 it's going to boost like a PS5 should and not as high as cooling allows it for your specific PS5.

So basically SoC temperature has no impact on clocks ? I find this very hard to believe.

 

[Chomiq]

So basically SoC temperature has no impact on clocks ? I find this very hard to believe.

The PlayStation 5's method for SoC boost allows for dramatic increase in GPU clocks much higher than PCs, no matter the ambient temperature

Sony offered details into how the boost clocks of the Zen 2 CPU and RDNA 2 GPU in the PS5's SoC are regulated. Sony's approach of using a constant power but varying frequency should allow all PS5 units to process data in the same manner irrespective of ambient temperature. It also allows the GPU...

www.notebookcheck.net

 

[Vya Domus]

In case of the PS5, however, things are a bit different. The console is given a certain power limit based on the cooling system.

So cooling does matter ? Worse cooling -> lower power limit -> lower clocks.

This is all extremely contradictory. I have a feeling Sony made these claims just so that nobody would question how the performance would change according to the cooling and ambient temperatures assuming that nobody could detect any difference if there was one. It's very bizarre that you would have an SoC with dynamic clocks that simply ignores temperatures and only takes into account power readings or workloads.

 

[FireFox]

also it took many many years for the slim/updated models to release not less than a year.

Sadly these days it has become standard to release a product and a few months later release an update of it or a new version

death to consoles long live the PC Master race.

I wouldn't like to see the death of the Consoles, some people cant afford to build a PC and a Console even they are expensive but still more affordable.

Sony won't be getting a penny from me. PC for life!!!

Sony will still get some money from me because there are a few Games i play but are just for the PlayStation brand

 

[Valantar]

Raw material cost increases are no doubt the main driver behind this. But I'd also expect Sony to have tons of thermal metrics collected from PS5 users around the world in all kinds of usage scenarios, which has likely told them all they need to in order to cut the heatsink down significantly while still ensuring that it has sufficient cooling capacity.

It's entirely possible that the original design was over-engineered (though we've seen evidence of piss-poor RAM cooling, on the verge of being an issue, this change doesn't really affect that), so this might just be bringing things into a more sensible balance.

I'd be wary of linking airflow temps too closely to SoC temps though - that's a very complex interplay between airflow, thermal mass, surface area, ambient temperatures, etc. Hopefully someone can stick some thermocouples in there and actually take some readings.

This new PS5 model comes with the new custom 6nm chip right ?

You mean 7nm, right? Or has there been a die shrink that literally nobody has written about? AFAIK there is no change in the underlying hardware of this.

 

[P4-630]

Sony recommends that both the PlayStation 5 and PlayStation 5 Digital Edition to be operated in a temperature range between 5°C to 35°C (which is 41°F to 95°F)

 

[Valantar]

Sony recommends that both the PlayStation 5 and PlayStation 5 Digital Edition to be operated in a temperature range between 5°C to 35°C (which is 41°F to 95°F)

That's pretty standard, no? Most electronics tend to have specified operating temperatures.

 

[Bones]

Steve at GamersNexus already showed the ram or VRM's were hitting 90+ celsius on the original PS5 cooling design. Longevity is important to me in consoles, plus as the video I linked in post 1 showed... 5 celsius may not sound like much, but if it effects the frequency to much it could have negative impact on gaming experience, and no because Sony would have had to test every PS5 game from start to finish to see if there are certain areas of certain games that end up hitching due to a lower boost clock (if there is even a lower boost close remains to be seen... so we simply need more research and reviews at this point)

Also, the Gamers Nexuys thing, can't remember if it was ram or VRM's, but how does the 5 celsius on the APU increase effect those other temps? That's another set of research done we need... cause if we knocking on 100 celsius now on the VRM's... longevity...

also, just to clarfiy a bit: I was actually waiting on a PS5 revision that would IMPROVE cooling not add 5 celsius more... based on steve from gamersnexus testing of the PS5 claiming it ran a little hot for comfort. so that's why I made this thread... it's not that 5 celsius is a massive deal (testing pending and more reviews etc) its mostly that I legit was hoping a future revision would improve the cooling, this is a step backwards imo, which ironically is the title of the article I linked, but it is true imo.

Longevity has already been affected by the choice of TIM used (LM) so it's most likely not going to last 5 years without an issue.
EDIT:
If the coolers are using nickel plating, then LM won't be a factor in terms of longevity but if not..... That means it won't end well in a few years - We'll see.

Sony still uses the lead free solder (Silver Solder) that was used in the older PS3's that has massive problems with it cracking due to thermal expansion and contraction over time. The only real solution is to reball the chips with standard 60/40 Tin/Lead solder so hopefully it never happens again.
People that's has it done so far say no issues once the reballing has been done that way.

Also the thing about how high thermals are being allowed to get is only contributing to that problem, the hotter it gets the more thermal expansion you'll have and that just "Works" the solder even more per heating and cooling cycle.
I'd not be as worried about the VRM's as much but that isn't an issue you just sweep under the rug either, it's actually smart to pay it a bit of mind and improve that too if possible.

 

[Naito]

Would like to see a comparison between the original and the revision in terms of die temperatures and boost clocks. It seems counterintuitive to reduce the cooler's effectiveness in a system that relies on boost clocks. Surely it'd be even less likely to hit the maximum specified clocks now?

 

[Chomiq]

Would like to see a comparison between the original and the revision in terms of die temperatures and boost clocks. It seems counterintuitive to reduce the cooler's effectiveness in a system that relies on boost clocks. Surely it'd be even less likely to hit the maximum specified clocks now?

As far as I'm aware you have no way to measure boost clocks on PS5. The only thing you can do is wire it up with thermocouples to measure temperature at the component level and then do a pure performance comparison against the older SKU.

 

[ppn]

You shouldn't be getting PS5 nonPro regardless. Just wait for the Pro on 5nm and play on higher FPS, don't ruin the experience with this early model capped performance. That cooler is suboptimal, the heatsink should run around the fan, now 5cm away from it. Whether 1200 grams or 1500 grams, it's worse than regular videocards. A lot of lost efficiency there. Just make it look 1:1 like the RTX 3090 or 3070, it is 250 watts after all, so similar cooling would do. And can't you just make your own water loops and get rid of the stock cooling anyway.

 

[Aquinus]

also, just to clarfiy a bit: I was actually waiting on a PS5 revision that would IMPROVE cooling not add 5 celsius more... based on steve from gamersnexus testing of the PS5 claiming it ran a little hot for comfort. so that's why I made this thread... it's not that 5 celsius is a massive deal (testing pending and more reviews etc) its mostly that I legit was hoping a future revision would improve the cooling, this is a step backwards imo, which ironically is the title of the article I linked, but it is true imo.

I'm pretty sure that we all understand that. The real question is if that difference in temperature will actually result in a tangible performance degradation. Without knowing how much thermal headroom there is or how much compute headroom there is, then there really is no way to know for certain without having both devices and running the same scene on them multiple times while using some HDMI capture thing to do some analysis on the outputted frames.

All in all, I bet the difference will be minimal. I seriously doubt that Sony would dump all of this money into the PS5 only to gimp it. Sony as a company isn't dumb and understands what bad PR can do.

 

[P4-630]

That's pretty standard, no? Most electronics tend to have specified operating temperatures.

That's what I mean.

 

[AsRock]

90c is crazy now they lessened the cooling more, and i bet these people who are doing these tests are in a well air conditioned room too. There is even less chance of me getting one now even if these changes make no performance difference.

 

[Solid State Soul ( SSS )]

You mean 7nm, right? Or has there been a die shrink that literally nobody has written about? AFAIK there is no change in the underlying hardware of this.

No, Sony was i discussion with TSMC to book custom 6nm node for future PS5 models

Sony Reportedly Prepping PS5 With 6nm AMD CPU

Sony's partners are gearing up for a redesigned PS5.

www.tomshardware.com

Anyway, folks from Digital Foundry are already planning a video on this. That will most definitely cover the performance aspect. GN will cover thermals and we're set.

Consoles are fixed performance, developers do extensive optimization to extract as much fidelity with what performance a console has, variable console performance would make matters more complicated.

New console revisions are just temperatures, acoustics, power consumption , weight, ect...

90c is crazy now they lessened the cooling more, and i bet these people who are doing these tests are in a well air conditioned room too. There is even less chance of me getting one now even if these changes make no performance difference.

these temps are not that uncommon among consoles, even pre built PCs often and reference GPUs run at those temps

 

[AsRock]

No, Sony was i discussion with TSMC to book custom 6nm node for future PS5 models

Sony Reportedly Prepping PS5 With 6nm AMD CPU

Sony's partners are gearing up for a redesigned PS5.

www.tomshardware.com

Consoles are fixed performance, developers do extensive optimization to extract as much fidelity with what performance a console has, variable console performance would make matters more complicated.

New console revisions are just temperatures, acoustics, power consumption , weight, ect...


these temps are not that uncommon among consoles, even pre built PCs often and reference GPUs run at those temps

Sure, and i do what i can to stay well away from those temps too, how ever with the terrible design which clearly they cared more for comes first.

 

[Rahnak]

Not too surprising if you consider how much the price of raw materials has been rising. I don't think Sony would do this if they didn't think they had the headroom. The other option was to raise the price and that wasn't gonna happen.

For anyone concerned wait until the inevitable slim re-design. Those are usually more efficient.

 

[Ruru]

Well, I haven't had any interest at all in any point so nothing to lose here.

But it's simply so damn common for manufacturers to make changes to hardware without any notice.

 

[dorsetknob]

Sony as a company isn't dumb and understands what bad PR can do.

Seem to Remember that a few years ago that the above Comment is misleading

anyone else remember Sony's Audio CD Rootkit Debacle
For those young'in's Read this

Sony BMG copy protection rootkit scandal - Wikipedia

en.wikipedia.org

 

[dragontamer5788]

I'm on PS4-Pro model. So unlike a lot of PS4 fans, I only bought my PS4 a few years ago, and still have a decent backlog. I think I'll be able to wait for the PS5-refresh, bit it a hypothetical slim model or "pro" edition. Either way, I'd expect the hardware to be a bit better by then.

This heatsink is definitely cheaper and worse. There's a couple of crap articles who clearly don't understand thermodynamics going around the internet, so I feel like putting my $0.02 into the discussion. But a "worse heatsink" isn't necessarily a worse design. We'll need to investigate a bit more to see if the few degrees-C changes anything. I doubt it however, I expect the original heatsink was probably just over-built.

 

[Valantar]

No, Sony was i discussion with TSMC to book custom 6nm node for future PS5 models

Sony Reportedly Prepping PS5 With 6nm AMD CPU

Sony's partners are gearing up for a redesigned PS5.

www.tomshardware.com

Interesting! IMO quite unlikely to be used for such a low-key refresh though - if the power savings from a node shrink are so minor as to barely warrant a heatsink update, and not a wholesale console shrink (especially given the gargantuan size of the PS5), they likely wouldn't be willing to pay the (potentially tens of) millions of dollars needed to make it happen. It would also be the first large die mass produced on the N6 node, which ... I guess they could be, but it seems early considering that N6 volumes are only just picking up compared to N7. I wouldn't be surprised if we saw an update/slim/whatever at CES or similar though, given that TSMC is promising N6 output to match N7 output by the end of the year. I could always be wrong, but this doesn't seem like change on the scale of a die shrink.