Sony Takes a Step Backwards with PlayStation 5, Cuts on Cooling Capacity in the new Revision

[mtcn77]

If the memory is now more hot than previous, that's really bad news, since PS5's memory ran already toasty. The APU was never a big issue.

Do you know what part of chip gets hot when stress testing because heat cycles kill faster than absolute temperature.

 

[Chomiq]

Test results from GN and DF are out:

Long story short - performance wise they run exactly the same. VRM and MOSFETs thermals - improved, memory - within 2-4C difference due to better coverage on new chasis.

So yeah, so much for some folks saying boost algo is the same for PS5 SoC as for Ryzen CPU on a PC.

GN even pulled a frankenconsole with a combo of chassis A and B.

 

[mtcn77]

Good for them. They really need to up their game and not fall to the same mistake again conforming to cancel culture, if they are any serious about their journalistic integrity.

PS: I find enough is enough with the naysayer pundits. We shouldn't run hasty judgement based on their manipulations and actively silence all their flamebait in all platforms - even when they use their tech channels to push internet hate.

 

[lexluthermiester]

Test results from GN and DF are out:

Long story short - performance wise they run exactly the same. VRM and MOSFETs thermals - improved, memory - within 2-4C difference due to better coverage on new chasis.

So yeah, so much for some folks saying boost algo is the same for PS5 SoC as for Ryzen CPU on a PC.

GN even pulled a frankenconsole with a combo of chassis A and B.

That was predictable... I mean really, who would have thought that engineers would know their craft. And wouldn't you know it, the newer, redesigned heatsink(which costs MUCH less money and resources to make) works very well. What a fricken shocker..

 

[Tardian]

That was predictable... I mean really, who would have thought that engineers would know their craft. And wouldn't you know it, the newer, redesigned heatsink(which costs MUCH less money and resources to make) works very well. What a fricken shocker..

I did.

 

[R0H1T]

Good for them. They really need to up their game and not fall to the same mistake again conforming to cancel culture, if they are any serious about their journalistic integrity.

PS: I find enough is enough with the naysayer pundits. We shouldn't run hasty judgement based on their manipulations and actively silence all their flamebait in all platforms - even when they use their tech channels to push internet hate.

Or view hungry YTers ~ they're worst IMO, always chasing clicks & then claiming we made a (honest?) mistake later on!

 

[Tardian]

Or view hungry YTers ~ they're worst IMO, always chasing clicks & then claiming we made a (honest?) mistake later on!

Was mich nicht umbringt, macht mich stärker.

“The world breaks everyone and afterward many are strong at the broken places. But those that will not break it kills. It kills the very good and the very gentle and the very brave impartially. If you are none of these you can be sure it will kill you too but there will be no special hurry.”

Ernest Hemingway, A Farewell to Arms

Mr Reaper is in no special hurry for me.

 

[mtcn77]

Ernest Hemingway, A Farewell to Arms

Mr Reaper is in no special hurry for me.

He was an alcoholic. They never overcome splitting.

 

[Tardian]

He was an alcoholic. They never overcome splitting.

Greatest is a heavy burden. Living up to his public persona would have required lubrication. The mask becomes permanently affixed.

High flying, adored!
I hope you come
To terms with boredom
So famous so easily
So soon
It's not the wisest thing to be, you won't care if they love you
It's been done before
You'll despair if they hate you
You'll be drained of all energy
All the young who've
Made it
Would agree

Songwriters: Lloyd-Webber Andrew / Rice Timothy Miles Bindon
High Flying, Adored lyrics © Evita Music Ltd.

 

[mtcn77]

Greatest

It is a black and white splitting, too. Don't be too superfluous.

 

[Tardian]

It is a black and white splitting, too. Don't be too superfluous.

Happy to discuss before we get accused of going off-topic. My aunt just turned 90. She wasn't alive when A Farewell to Arms was published. The use of the N-word was a literary tool. Leave the PC stuff to TPU.

My PS5 (disc version) arrives on October 4, 2021.

 

[AleXXX666]

You should know why Sony would cut the cooling down, their engineers probably figured out the cooling was just a bit too good and most likely would offer a long life and in todays world, a company just cant have people owning things for a long time without having to keep buying new ones. After all, how else will all the over priced execs pay for all their bullshit.

it's people's OWN decision what to buy and what not to buy, my friend... i bought 3060Ti for $900 just because I wanted one, and not crying and yelling "how big prices are". but if I haven't sold my old 2070S once, I kept using it till these, lol.

 

[GrumpyOtaku]

Err, except that the customers are not so stupid and will not buy the new ones You know what customers satisfaction is?

You have too high an opinion of the sheeple in the general public. If these companies hadn't already been able to get away with this sort of nonsense in the past, they wouldn't be emboldened to pull even more of it now and in the future. The general public will just nod with blank stares in their eyes, and then bend over and take it again. After all these are the same sorts that voted for Creepy Joe and to keep Emperor Newsom.

 

[lexluthermiester]

You have too high an opinion of the sheeple in the general public.

You both assume most people care. They do not. Most people just want their system to run so they can have fun. It's only us geeks that care about the nitty-gritty tech specs.

 

[Tardian]

I just hope to live long enough to play GT8 on my PS6 and then GT9 on my PS7.

 

[chrcoluk]

Good news on the cooling, I expect better airflow as well as better contact has helped, a bigger heatsink is not necessarily good when space is at a premium for air movement.

My concern is however DF seems to have insider info which suggests Sony potentially had input on the video. The bit that got my attention is when he mentioned how the boost algorithm works, this information as far as I can tell is not released to the public by Sony. Apparently the performance isn't variable from unit to unit based on temps and power limits but rather the clocks are forced and if for some reason the console cannot handle it then it will crash/shutdown instead. So looks like a conversation happened between DF and Sony, whether it was just for technical clarification or more was said will never know.

It was good they used the same PS5 board on both cases, as the voltage needed for each clock will vary from unit to unit dependent on silicon quality, so that was the best way to test.

 

[TheoneandonlyMrK]

Good news on the cooling, I expect better airflow as well as better contact has helped, a bigger heatsink is not necessarily good when space is at a premium for air movement.

My concern is however DF seems to have insider info which suggests Sony potentially had input on the video. The bit that got my attention is when he mentioned how the boost algorithm works, this information as far as I can tell is not released to the public by Sony. Apparently the performance isn't variable from unit to unit based on temps and power limits but rather the clocks are forced and if for some reason the console cannot handle it then it will crash/shutdown instead. So looks like a conversation happened between DF and Sony, whether it was just for technical clarification or more was said will never know.

It was good they used the same PS5 board on both cases, as the voltage needed for each clock will vary from unit to unit dependent on silicon quality, so that was the best way to test.

Not sure how you get from it being one profile for all like a typical AMD product to individual voltage curve's personally.

They typically use one power voltage and fan curve per SKU and set it up so that ANY quality of chip above the minimum defect point works as intended and largely the same as any other.

Variations in quality and build quality then only get reflected in operating temperature alone and not performance.

 

[Shrek]

That was predictable... I mean really, who would have thought that engineers would know their craft. And wouldn't you know it, the newer, redesigned heatsink(which costs MUCH less money and resources to make) works very well. What a fricken shocker..

Indeed:

“An engineer can do for a dollar what any fool can do for two.”
― Arthur Mellen Wellington

 

[Tardian]

Indeed:

“An engineer can do for a dollar what any fool can do for two.”
― Arthur Mellen Wellington

However, an accountant may be needed to scale back the gold-plating.

 

[Valantar]

Good news on the cooling, I expect better airflow as well as better contact has helped, a bigger heatsink is not necessarily good when space is at a premium for air movement.

My concern is however DF seems to have insider info which suggests Sony potentially had input on the video. The bit that got my attention is when he mentioned how the boost algorithm works, this information as far as I can tell is not released to the public by Sony. Apparently the performance isn't variable from unit to unit based on temps and power limits but rather the clocks are forced and if for some reason the console cannot handle it then it will crash/shutdown instead. So looks like a conversation happened between DF and Sony, whether it was just for technical clarification or more was said will never know.

It was good they used the same PS5 board on both cases, as the voltage needed for each clock will vary from unit to unit dependent on silicon quality, so that was the best way to test.

This has been widely reported. Here's Eurogamer/DF's initial article reporting on how PS5 boosting works. There is zero indication that Sony had input on this video - this was reported 18 months ago. It is by no means new information. It really, really, really shouldn't be worrying to you that press has better access to information than the general public. That is literally their job.

Test results from GN and DF are out:

Long story short - performance wise they run exactly the same. VRM and MOSFETs thermals - improved, memory - within 2-4C difference due to better coverage on new chasis.

So yeah, so much for some folks saying boost algo is the same for PS5 SoC as for Ryzen CPU on a PC.

GN even pulled a frankenconsole with a combo of chassis A and B.

The SoC is also marginally warmer, but yeah, overall the new cooler looks good. Mainly speaks to the original design being (surprisingly) bad, which might indicate either a rushed design or someone on the cooler design team being prone to over-engineering rather than efficient design. Either way, a more efficient cooler is good for pretty much everyone. Less materials used, lighter weight=less shipping emissions, lower cost has the potential to enable sales/rebates/price cuts in the future (though I'm not betting on that last one). And I don't think the savings are sufficient to really impact Sony's bottom line, so it's not like they're laughing all the way to the bank either.

 

[BorisDG]

... Mainly speaks to the original design being (surprisingly) bad ....

Why it's bad? Actually it's still better since the SoC is much cooler. That the memory on the other hand is now better it's not exactly due to the cooler itself, but more to the backplate and the contact points. Fan seems also a factor. The "hated" Nidec looks better than NMB.

 

[chrcoluk]

Well I read that article, and part of it states the opposite of what is been said.

Here is the exact quote.

According to Sony, all PS5 consoles process the same workloads with the same performance level in any environment, no matter what the ambient temperature may be.

So how does boost work in this case? Put simply, the PlayStation 5 is given a set power budget tied to the thermal limits of the cooling assembly. "It's a completely different paradigm," says Cerny. "Rather than running at constant frequency and letting the power vary based on the workload, we run at essentially constant power and let the frequency vary based on the workload."

This indicates there is nothing unusual about the boost algorithm and instead seems to indicate that Sony have made sure the voltages they run at are the same for every single chip that goes in each console. (binned voltages same as worst chips for the better chips). Effectively removing silicon lottery on performance.

It really, really, really shouldn't be worrying to you that press has better access to information than the general public. That is literally their job.

Each to their own opinion, I personally think reviewers have a too friendly relationship with the companies that make the products they review. I have no issue if they have gone out and researched to find the info in a public document somewhere, my issue is if the information is provided exclusively to the reviewer(s) so they can publish it. I wont comment on this again in this thread though as its off topic.

 

[Valantar]

Why it's bad? Actually it's still better since the SoC is much cooler. That the memory on the other hand is now better it's not exactly due to the cooler itself, but more to the backplate and the contact points. Fan seems also a factor. The "hated" Nidec looks better than NMB.

It's bad (=inefficient) because they manage functionally the same (slightly worse SoC, slightly better everything else) cooling with a significantly smaller and simpler cooler design. Even if some of that change is due to better cold plate/midframe design, the results still demonstrate that the original design was poorly optimized and laid out, making far worse use of its materials and construction. When you can reduce the cooler's mass by 16% (1639g v. 1368g) without meaningfully affecting performance (+~4°C, or +5.6%), this speaks to the original design being significantly ineffective. While thermal dissipation per cooler mass isn't the be-all-end-all of cooler efficiency, this is a big enough discrepancy to pretty conclusively indicate that the initial cooler design had a design with significant issues. Of course, there are also other fixes to rather mind-boggling issues, like the previous tiny dimples in the midframe contacting the RAM - something that borders on outright incompetent design. But overall, the new design seems more sensible, better designed to the task (rather than overbuild to the degree of becoming problematic), and more than sufficient for what it's made for. When that can be done with what is visibly a simpler (and thus cheaper) design using less materials, again, that speaks to the original design being a poor design.

Well I read that article, and part of it states the opposite of what is been said.

Here is the exact quote.

This indicates there is nothing unusual about the boost algorithm and instead seems to indicate that Sony have made sure the voltages they run at are the same for every single chip that goes in each console. (binned voltages same as worst chips for the better chips). Effectively removing silicon lottery on performance.

That is a rather selecive quote that does not tell the whole story. That quote essentially describes any CPU or SoC with a reasonably modern opportunistic boosting system with thermal and power limits. The key clarification comes from this quote, two paragraphs further down:

While it's true that every piece of silicon has slightly different temperature and power characteristics, the monitor bases its determinations on the behaviour of what Cerny calls a 'model SoC' (system on chip) - a standard reference point for every PlayStation 5 that will be produced.

The PlayStation 5 has variable frequencies for CPU and GPU, with an internal monitor adjusting clocks to keep the system within its power budget.
"Rather than look at the actual temperature of the silicon die, we look at the activities that the GPU and CPU are performing and set the frequencies on that basis - which makes everything deterministic and repeatable," Cerny explains in his presentation. "While we're at it, we also use AMD's SmartShift technology and send any unused power from the CPU to the GPU so it can squeeze out a few more pixels."

This is how the boosting system differs from PC and mobile CPUs, APUs and SoCs - through boosting to predetermined levels of performance based on the workload as tested on a prototypical 'model SoC' rather than real-world, in-use thermal and power metrics. This means that there will be variance in real-world PS5s depending on the silicon lottery - some will consume a bit more power, some will run a tad hotter - but they will all deliver consistent performance. (They are clearly binned to ensure that all SoCs used meet baseline performance and power metrics to be able to reasonably match model SoC performance and not overheat).

Each to their own opinion, I personally think reviewers have a too friendly relationship with the companies that make the products they review. I have no issue if they have gone out and researched to find the info in a public document somewhere, my issue is if the information is provided exclusively to the reviewer(s) so they can publish it. I wont comment on this again in this thread though as its off topic.

I know this is a bit OT, but this really rubs me the wrong way - probably due to this hitting too close to home (I work in and have my education in media studies). What you are presenting here seems like sensible, basic media literacy - "be critical of your sources", "question the closeness of reporters to their sources", etc. However, your application of it as demonstrated above is simplistic and black-and-white to a level of near parody. What you are arguing for here is essentially the abolition of journalism. You say they should "find the info in a public document somewhere" - where? Which public documents? They don't exist. This information is not made public, it is proprietary information that companies do not part with voluntarily. You might find something useful in something like a patent application or grant, but those are slow to be published and don't actually show if the patent was ever implemented or if they just patented an idea. There would be no confirmation of this actually being present. This boost system would have been a complete black box to the public if not for journalists asking questions.

Now, yes, the tech press is generally far too close to the industry, and most "reporting" (especially, but not limited to YouTube) toes the line between reporting and advertising very closely. Of course a lot of it is advertising, or "sponsored content" as they like to call it. DF is by no means flawless in this regard - there was a previous situation (their 3080 preview, I think) where their video omitted a key paragraph from the written article that highlighted how Nvidia had set the limitations for what they were allowed to benchmark and report on. This was clearly problematic, but it was also a preview, explicitly worded as such, and something that by its very definition was not a full review. They should have been equally clear on this in the video as in writing, but the issue is not their getting access in the first place.

But - and this is crucial - balancing closeness to sources with the responsibility to deliver information to the public in a way that is as neutral and unbiased as possible? That is their job. That is the very core of journalism. Doing the work of gaining access in order to gain insight and provide information that the public would otherwise not have had access to. Discussing this balance is extremely valuable. But you show no interest in discussing it, instead dismissing it outright, implicitly stating that any access beyond publicly available documents is problematic to an unacceptable degree. This is an extreme judgement, and not one that stands up to any type of scrutiny, as it makes reporting news essentially impossible. Should crime reporters only base their reporting on public police statements, and not seek out detectives or witnesses? Should political reporters refuse to talk to politicians, instead relying solely on press releases? Remember, actually talking to someone is the only way you can possibly hope to ask them questions. Basing your work on already published documentation gives the sources more power, both through giving them the power to preemptively edit their statements as well as removing the power of others to ask follow-up questions. And without access - without contacts, some degree of closeness - you're never going to get to ask a follow-up question. This is obviously not the same as saying "we should blindly trust everythign everyone claiming to be a journalist says". Not even close. But some degree of trust is necessary. The world cannot work without trust. And without some trust in journalistic integrity, we lose access to massive amounts of valuable information, instead letting companies and those in power define the truth.

 

[Shrek]

Really need to change the title given that it seems not to be correct.

 

[TheoneandonlyMrK]

Really need to change the title given that it seems not to be correct.

I'd say provide a update to the OP too.