• 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.
  • The Politics forum has been nuked. Please do not bring political discussion to the rest of the site, or you will be removed. Thanks.

News [DF] PS5 Firmware 2.0 Tested

Jaysen

Banned
Mar 19, 2021
1,587
2,958
430
Installed a 1TB SN850 with a cheap heatsink from Amazon and the thing is awesome so far. No noticeable heat gain, or noise increase. I moved everything from the main drive onto it, and it’s handled everything as good or better. Returnal actually seems to be less juddery now, but that could be my imagination. Just the space increase alone was worth it.
 
Mar 27, 2020
14,670
37,642
725
Uncharted bunghole
Installed a 1TB SN850 with a cheap heatsink from Amazon and the thing is awesome so far. No noticeable heat gain, or noise increase. I moved everything from the main drive onto it, and it’s handled everything as good or better. Returnal actually seems to be less juddery now, but that could be my imagination. Just the space increase alone was worth it.

Nice to have options right?

Did you use a heatsink with those rubber bands?
 

Imtjnotu

Member
Nov 4, 2017
3,933
13,399
740
san diego
Nope, I used a double sided heatsink that secured with six screws.
It fit in the PS5 perfectly. Like it was designed specifically for this usage.
i think this is still too tall. i thought the tallest ssd we could use was 11mm? that thing alone is 11mm.

the one i use is 9mm with out the SSD.

 

Jaysen

Banned
Mar 19, 2021
1,587
2,958
430
i think this is still too tall. i thought the tallest ssd we could use was 11mm? that thing alone is 11mm.

the one i use is 9mm with out the SSD.

Nope, works perfectly. The cover fits right over it.
 
Dec 14, 2008
34,060
2,723
1,490
It appears that soldering the SSD modules onto the board was a bad idea, that will only lead to PS5 bricks in the far future. As the interface they’re using works just fine using the expansion slot. This makes their original SSD module solution wacky tech that was in the end probably not worth the time/effort/R&D. I bet PS5 revisions will have an approach like the Series consoles, replaceable, but not without a headache, but that’s ok because it’s a fairly extreme repair.

Someone needs to test if the PS5 has a “fail to expanded SSD mode” for when that eventually happens.
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.
 

Paulxo87

Member
Oct 24, 2017
707
1,312
490
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.

right. I never understood this concern.
 

Md Ray

Member
Nov 12, 2016
4,221
13,559
785
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.
👏👏 good post.

It appears that soldering the SSD modules onto the board was a bad idea, that will only lead to PS5 bricks in the far future. As the interface they’re using works just fine using the expansion slot. This makes their original SSD module solution wacky tech that was in the end probably not worth the time/effort/R&D. I bet PS5 revisions will have an approach like the Series consoles, replaceable, but not without a headache, but that’s ok because it’s a fairly extreme repair.

Someone needs to test if the PS5 has a “fail to expanded SSD mode” for when that eventually happens.
Dj Khaled Congratulations GIF
 
Mar 27, 2020
14,670
37,642
725
Uncharted bunghole
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.

Yeah the battery would die before the storage does. Plus the PS5 SSD had DRAM so that should extend its life. People are seriously getting the wrong idea that SSDs last only a couple of years.
 

ReBurn

Member
Dec 6, 2008
14,120
6,080
1,405
SC USA
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.
Apple is huge on planned obsolescence so I don't know if they're the best example to use here. By soldering components to their boards and locking key components to the device's serial number you have no choice but to purchase a new device if some components fail.

More than half of iPhone owners upgrade their phones as soon as they pay off them off. About 40% of Android users do the same. Phones often aren't used in conditions where 30-100 gigabyte payloads are regularly downloaded, installed and deleted. There also aren't a lot of metrics on smartphone reliability after two years or from the second hand market. Phone manufacturers don't really care if their stuff breaks after a year or why. If you're under warranty they just give you a new one. Plus they are fragile and portable so they get broken at a pretty high rate from drops or other physical damage. They typically don't survive long enough for storage to the the thing that fails.

For larger data loads I've personally seen quite a few heavy use developer MacBook Pro laptops where I work have their SSD fail as soon as 2 years after I purchased them. The constant P/E cycles from builds, installations, local database destruction and restoration, etc. can take a toll on an SSD. Since you can't really get modern Apple SSD computers repaired I've taken to just replacing them at the 3 year mark or if they break after the extended warranty expires. But Apple doesn't really intend for you to keep that MacBook for more than 2 years, anyway.

As far as consoles go the on-board storage chips are probably rated for a 10 year average lifespan like most SSD's. But like other SSD's the actual life will be determined by environmental conditions and how the device is used. They're probably going to be more durable than mechanical drives but I wouldn't be surprised to see some of the PS5 and Xbox systems that get the heaviest use start to see storage break down around the time the inevitable mid-gen refreshes start to hit. Most will probably last the generation, though.
 
Last edited:

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Wouldn't that write speed give you an idea of how fast an update to the game would be applied?

The key point is that the PS5's I/O complex exists to facilitate and expedite transfers from drive to system ram. Its all about efficient pipelining where data is pulled in in such a manner that the stuff that needs to be worked on by the processor arrives first allowing it to get its workload started while the rest silently loads in parallel into other regions of ram. Expand that across multiple task threads all working concurrently and you can see how useful having multiple tiers of fetch priority is.

Drive-to-drive copying is comparatively simplistic, and most importantly is never likely to be a performance critical operation. As in while faster is obviously better, any differential isn't going to be a deal-breaker.

The only thing that matters is for both drives to hit their targets in terms of being able to feed the system at the rates demanded.
 

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
The key point is that the PS5's I/O complex exists to facilitate and expedite transfers from drive to system ram. Its all about efficient pipelining where data is pulled in in such a manner that the stuff that needs to be worked on by the processor arrives first allowing it to get its workload started while the rest silently loads in parallel into other regions of ram. Expand that across multiple task threads all working concurrently and you can see how useful having multiple tiers of fetch priority is.

Drive-to-drive copying is comparatively simplistic, and most importantly is never likely to be a performance critical operation. As in while faster is obviously better, any differential isn't going to be a deal-breaker.

The only thing that matters is for both drives to hit their targets in terms of being able to feed the system at the rates demanded.

How does your reply address my comment? I'm not talking about the read speed of the drive. To quote your earlier comment, talk about missing the point, whoosh!

I'm not talking about drive to drive copying. I'm not talking about read speeds.

If the write speed on the NVMe is 7x the write speed of the internal drive, this should have an impact large write operations, such as the "copying" phase where the update is patched to an installed game, which was, and still is, an issue with games on the system.

So yeah, I found the results from that test to be the most interesting part of the video. The various results from game loading tests showed expected performance.
 

Panajev2001a

GAF's Pleasant Genius
Jun 7, 2004
19,774
14,507
2,110
How does your reply address my comment? I'm not talking about the read speed of the drive. To quote your earlier comment, talk about missing the point, whoosh!

I'm not talking about drive to drive copying. I'm not talking about read speeds.

If the write speed on the NVMe is 7x the write speed of the internal drive, this should have an impact large write operations, such as the "copying" phase where the update is patched to an installed game, which was, and still is, an issue with games on the system.

So yeah, I found the results from that test to be the most interesting part of the video. The various results from game loading tests showed expected performance.
I think the copy data from internal SSD to additional bay SSD showed much higher transfer than the reverse operation for example. It does have an impact in those operations.

The “copying/game update patching” phase I suspect is not purely I/O bound.
 
Last edited:
  • Like
Reactions: PharaoTutAnchAmun

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
I think the copy data from internal SSD to additional bay SSD showed much higher transfer than the reverse operation for example. It does have an impact in those operations.

The “copying/game update patching” phase I suspect is not purely I/O bound.

Would make an interesting test to confirm it.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
How does your reply address my comment? I'm not talking about the read speed of the drive. To quote your earlier comment, talk about missing the point, whoosh!

I'm not talking about drive to drive copying. I'm not talking about read speeds.

If the write speed on the NVMe is 7x the write speed of the internal drive, this should have an impact large write operations, such as the "copying" phase where the update is patched to an installed game, which was, and still is, an issue with games on the system.

So yeah, I found the results from that test to be the most interesting part of the video. The various results from game loading tests showed expected performance.

Patching is still more reading than writing. 2 data objects (read sources) are combined to create a final (write) result that matches a prescribed hash value (checksum read) for validation.

Most relevantly its not a time-critical task, there's no JIT requirement and so the OS could mandate it as a low-priority process. Especially so as for security/stability concerns accuracy, not speed, is the paramount concern.
 
  • Like
Reactions: Whitecrow

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
Patching is still more reading than writing. 2 data objects (read sources) are combined to create a final (write) result that matches a prescribed hash value (checksum read) for validation.

Most relevantly its not a time-critical task, there's no JIT requirement and so the OS could mandate it as a low-priority process. Especially so as for security/stability concerns accuracy, not speed, is the paramount concern.
Per the way this was implemented on the PS4, it's more writing than reading. The entire game file is written back to the drive, be it the HDD or the SSD/NVMe. Despite what Cerney said, I still see a similar behaviour on the PS5, albeit faster.

Your original question was what was the point of DF even doing this test. I gave you one possible scenario where the performance of the system would be impacted by the write speed, when updating the games already on the SSD. You asked the question, I gave you an answer. If you can explain to me that the write speed being 7x faster on the NVMe drive won't improve time to patch the game, I'm all ears. But please spare me the replies explaning how it doesn't matter, low priority, etc. This wasn't my point.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Per the way this was implemented on the PS4, it's more writing than reading. The entire game file is written back to the drive, be it the HDD or the SSD/NVMe. Despite what Cerney said, I still see a similar behaviour on the PS5, albeit faster.

Your original question was what was the point of DF even doing this test. I gave you one possible scenario where the performance of the system would be impacted by the write speed, when updating the games already on the SSD. You asked the question, I gave you an answer. If you can explain to me that the write speed being 7x faster on the NVMe drive won't improve time to patch the game, I'm all ears. But please spare me the replies explaning how it doesn't matter, low priority, etc. This wasn't my point.

You can't write something without reading it first unless the data is generated entirely programmatically. Which patches/diff files aren't. Obviously.

For streaming efficiency you want contiguous data on a platter-based hdd, meaning fewer bigger files rather than many smaller ones. When patching security and redundancy is important, you don't want to risk corrupting the file in place during the process so you make a work copy. So essentially what you have is a scenario where potentially changing 100 bytes of a 1Gb file, requires an additional 1Gb of disk space to make the working copy. This is why patch sizes on download often weren't reflective of the amount needed to apply them.

This is not a concern on a SSD because there's no longer seek-times to worry about and more importantly data does not to be laid out contiguously like you do on a linear access storage medium like 5400rpm HDD. The whole copy process where the heavy sequential writing is done no longer applies.

And I mean this in a physical sense as much as in a file-system way. Writes are always scattered across NAND cells to balance usage and preserve device longevity, its an entirely different paradigm.
 
Last edited:
  • Like
Reactions: Chris_Rivera

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
You can't write something without reading it first unless the data is generated entirely programmatically. Which patches/diff files aren't. Obviously.

For streaming efficiency you want contiguous data on a platter-based hdd, meaning fewer bigger files rather than many smaller ones. When patching security and redundancy is important, you don't want to risk corrupting the file in place during the process so you make a work copy. So essentially what you have is a scenario where potentially changing 100 bytes of a 1Gb file, requires an additional 1Gb of disk space to make the working copy. This is why patch sizes on download often weren't reflective of the amount needed to apply them.

This is not a concern on a SSD because there's no longer seek-times to worry about and more importantly data does not to be laid out contiguously like you do on a linear access storage medium like 5400rpm HDD. The whole copy process where the heavy sequential writing is done no longer applies.

And I mean this in a physical sense as much as in a file-system way. Writes are always scattered across NAND cells to balance usage and preserve device longevity, its an entirely different paradigm.
Have you patched a PS4 game on a PS5? Is there a copying process? Why? Given it's on the SSD. And I still see "copying" with a PS5 game being patched despite what Cerny said (and you copy/pasted from the Road to PS5)
 
Last edited:

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Have you patched a PS4 game on a PS5? Is there a copying process? Why? Given it's on the SSD. And I still see "copying" with a PS5 game being patched despite what Cerny said (and you copy/pasted from the Road to PS5)

I've had a PS5 for all of 2 days at this point, so couldn't tell you. And given that the default behaviour is to install all PS4 content to an external drive if available (I have my old PS4 drive attached) its extremely probable that it works as before because it has to function correctly when working with a 5400rpm platter via USB bus.

And no, I didn't copy/paste any part of my post. What I wrote was the technical facts as I understand them - claiming its a different paradigm between SSD and HDD is not empty marketing hype. The underlying mechanics of these devices are radically different even if from the end-user perspective they do the same job.
 

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
I've had a PS5 for all of 2 days at this point, so couldn't tell you. And given that the default behaviour is to install all PS4 content to an external drive if available (I have my old PS4 drive attached) its extremely probable that it works as before because it has to function correctly when working with a 5400rpm platter via USB bus.

And no, I didn't copy/paste any part of my post. What I wrote was the technical facts as I understand them - claiming its a different paradigm between SSD and HDD is not empty marketing hype. The underlying mechanics of these devices are radically different even if from the end-user perspective they do the same job.
Well as mentioned I still see the copying/patching behaviour on the PS5. While obviously faster compared to the HDD on the PS4, it isn't instant. And this is on the internal SSD. So again, going back to my original point, the test DF did comparing the write speeds of the drives has a valid use case, regardless of what you think.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Well as mentioned I still see the copying/patching behaviour on the PS5. While obviously faster compared to the HDD on the PS4, it isn't instant. And this is on the internal SSD. So again, going back to my original point, the test DF did comparing the write speeds of the drives has a valid use case, regardless of what you think.

I'm going to disagree on the basis that this needs more detailed investigation to get an accurate picture of what's actually going on.

Consider how the PS5 has observably some sort of i/o throttling mechanism (part of the priority staging?) which you clearly see manifested in how PS4 titles load times are basically unchanged regardless of whether they are run from the onboard SSD or any sort of external drive.

This being the case, who's to say that the disparity in speed copying to/from the SSD's is not a matter of interface limitations, but somehow a result of how the system operates under specific conditions? Could it be some sort of O/S bug where transfer from the secondary SSD is throttled the same way that transfers from external SSD's are?

For all we know there may be some measure in place that aggressively limits incoming transfer write speeds on the primary SSD, maybe to reserve i/o bandwidth on that device for other system or game operations. I can see that being plausible as its not going to be a deal-breaker in terms of day-to-day use to slow down bulk writes given the intended range of utility of a console.

At the end of the day its a curiosity, not anything anyone should be concerned about. Taking a few extra seconds to copy stuff across isn't a deal-breaker when its going to be such an infrequent action. I mean realistically, outside of these tests why should anyone want to be shunting gigs of data between the internal drives?
 
Last edited:
  • Like
Reactions: Chris_Rivera

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
At the end of the day its a curiosity, not anything anyone should be concerned about. Taking a few extra seconds to copy stuff across isn't a deal-breaker when its going to be such an infrequent action. I mean realistically, outside of these tests why should anyone want to be shunting gigs of data between the internal drives?
Just stop. I mentioned a scenario where the write speed would have an impact, and you got all defensive. You seem to admit there could be a difference in the speed but because it's something infrequent isn't not a dealbreaker. The frequency wasn't my point, but that there could be an impact on patching speed. A write speed difference of 7x is nothing to be sneezed at.
 
Last edited:
  • LOL
  • Thoughtful
Reactions: Clear and Shmunter

DarkestHour

Member
Oct 27, 2015
1,750
1,219
580
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.

I'm not disagreeing with you, but iPhones and iPads rarely have as many writes as a console will. Also, I'd wager that Apple is using higher quality flash in their phones to justify the prices versus a console being made with the cheapest parts available to stay within that price range.

Regardless, flash storage is so much more resilient now that any failure would take an incredible amount of writes to wear it out and it's a non-issue.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Just stop. I mentioned a scenario where the write speed would have an impact, and you got all defensive. You seem to admit there could be a difference in the speed but because it's something infrequent isn't not a dealbreaker. The frequency wasn't my point, but that there could be an impact on patching speed. A write speed difference of 7x is nothing to be sneezed at.

The only one being defensive here is you. You have no comeback to my reasonable counter-points and you simply cannot bear to admit that I'm right.

My entry point was that I thought the test was basically meaningless and I've stuck to my guns on that opinion. If we were talking about performance on a file-server it'd be relevant and noteworthy, but it isn't.

Its a game console where read speed (into system ram) is all that actually matters for day-to-day operation.
 
Last edited:
  • Like
Reactions: Chris_Rivera

ethomaz

is mad because DF didn't do a video on a video of a video of a video on PS5
Mar 19, 2013
42,943
46,379
1,310
39
Brazil
I'm not disagreeing with you, but iPhones and iPads rarely have as many writes as a console will. Also, I'd wager that Apple is using higher quality flash in their phones to justify the prices versus a console being made with the cheapest parts available to stay within that price range.

Regardless, flash storage is so much more resilient now that any failure would take an incredible amount of writes to wear it out and it's a non-issue.
Console writes is not very low? I mean you just have writes when install things.

It is not like PC that have constant writes all the time due the OS.

BTW I believe iPhone have a lot os writes too… people loves to take pics and videos… imo I take way more pics with my Phone per month than I install games in my console… size wise.

Unless of course you install dozen of games on your console per month… I usually install one when it releases… that means each 2-3 months?

In any case a mechanical HDD will fail way before you reach the number of times a SSD cell can be written.
 
Last edited:
  • Like
Reactions: Chris_Rivera

ethomaz

is mad because DF didn't do a video on a video of a video of a video on PS5
Mar 19, 2013
42,943
46,379
1,310
39
Brazil
Gamesaves (joke), Though you have the always recording the gameplay feature.
I forgot the always recording gameplay :D
It can be disabled but it is on by default.

That is probably the main writer in the console.
 
Last edited:

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
The only one being defensive here is you. You have no comeback to my reasonable counter-points and you simply cannot bear to admit that I'm right.

My entry point was that I thought the test was basically meaningless and I've stuck to my guns on that opinion. If we were talking about performance on a file-server it'd be relevant and noteworthy, but it isn't.

Its a game console where read speed (into system ram) is all that actually matters for day-to-day operation.
No comeback to what exactly? Your ‘bugs in the OS’ hypostasise? 😂

you’ve already admitted it’s possible the difference in write speed would impact the patch time which was my point. 👍
 
Last edited:

iHaunter

Member
Sep 6, 2015
3,560
4,274
625
No need to buy a 2 TB SSD right now. Very over priced. I agree this will be a very long gen. But thats actually a better reason to just go for a cheap 500 GB M.2 for now. Wait til like 2025+ as Gen5 SSDs come out and prices of these Gen4 drives will fall a lot. We will even be able to use Gen5 M.2 SSDs in PS5 too.

For example,

667 GB usable space on internal SSD
+ 500 GB usable space on M.2 SSD
+ 1840 GB usable space on external USB HDD ("2 TB")
= 3.0 TB of usable space. 60/40 split external / internal

Much cheaper and allows for plenty of space, especially since so many games are still PS4 games playable from an external HDD.
IDK what you people doing with 3TB of games.

That's like 60+ games.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
you’ve already admitted it’s possible the difference in write speed would impact the patch time which was my point. 👍

Possible but actually unlikely, and either way completely unproven by the anecdotal evidence provided by this "test".

All we have to go on is a discrepancy when data is copied from the onboard SSD to the expansion SSD based on which unit is the source and which the destination. Given that patching involves communication between system ram and the storage medium and does not require shifting data between the SSD's its relevance is dubious by any standard.

What's being stressed is data bus handling between the two SSD's, same as if you attached the same NVMe drive via usb you'd see a similar result and obviously that wouldn't be reflective of the internal SSD's capabilities either.

Your whole argument is demonstrably bogus anyway given the size of the difference is way greater than could logically be a product of the NAND's speed! You aren't going to see a 7x differential between a gen 3 and gen 4 NVMe drive despite the former being rated too slow to function on the PS5!
 

NickFire

Member
Mar 12, 2014
9,478
10,602
980
Installed a 1TB SN850 with a cheap heatsink from Amazon and the thing is awesome so far. No noticeable heat gain, or noise increase. I moved everything from the main drive onto it, and it’s handled everything as good or better. Returnal actually seems to be less juddery now, but that could be my imagination. Just the space increase alone was worth it.
I bought the same one, but different heat sink today. Then got home and found out my beta code is the PS4 one. Hoping Sony sends me a code sooner than later now.
 

Shmunter

Member
Aug 25, 2018
11,034
24,964
835
I forgot the always recording gameplay :D
It can be disabled but it is on by default.

That is probably the main writer in the console.
I personally think it uses ram and doesn’t write till user saves or until it’s really long. This is why the o/s ram reserve seems so large.
 
Last edited:

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
Possible but actually unlikely, and either way completely unproven by the anecdotal evidence provided by this "test".

…Your whole argument is demonstrably bogus anyway given the size of the difference is way greater than could logically be a product of the NAND's speed! You aren't going to see a 7x differential between a gen 3 and gen 4 NVMe drive despite the former being rated too slow to function on the PS5!
Actually the only thing unproven is your excuse of ‘bugs’ in the OS.

If you have any data showing the write speed of the NVMe drive isn’t significantly faster than the internal, to the extent of 7x, please provide it here.
 

ethomaz

is mad because DF didn't do a video on a video of a video of a video on PS5
Mar 19, 2013
42,943
46,379
1,310
39
Brazil
I personally think it uses ram and doesn’t write till user saves or until it’s really long. This is why the o/s ram reserve seems so large.
Makes sense.
But I believe it is saved on SSD and start to be discarded after 15 minutes… you can increase this time to hours.
 
Last edited:

Shmunter

Member
Aug 25, 2018
11,034
24,964
835
Makes sense.
But I believe it is saved on SSD and start to be discarded after 15 minutes… you can increase this time to hours.
I’d be curious on the file sizes coming out of the PS5 using the new codecs.
 

Clear

Member
Feb 2, 2009
13,769
10,001
1,365
Actually the only thing unproven is your excuse of ‘bugs’ in the OS.

If you have any data showing the write speed of the NVMe drive isn’t significantly faster than the internal, to the extent of 7x, please provide it here.

I don't need to. You made an assertion and I debunked it thoroughly with facts and logic. Both in terms of flaws in the testing methodology and by offering alternate hypotheses just as plausible as yours.

More to the point I've maintained that it really doesn't matter one way or the other, because regardless of the cause its not something that anyone needs to be concerned about!

If you want to debate the issue further then advance your argument.
 
  • Like
Reactions: Panajev2001a

jasondefaoite

Member
Jan 18, 2020
409
1,648
390
I don't need to. You made an assertion and I debunked it thoroughly with facts and logic. Both in terms of flaws in the testing methodology and by offering alternate hypotheses just as plausible as yours.

More to the point I've maintained that it really doesn't matter one way or the other, because regardless of the cause its not something that anyone needs to be concerned about!

If you want to debate the issue further then advance your argument.
I'd agree with you, but then we'd both be wrong.
 
  • Empathy
Reactions: Panajev2001a
Mar 16, 2014
1,534
1,526
780
I’m not getting speeds into the 6000mb/sec range on my 980 pro 🙁 it’s just barely over 5500mb/sec read. It’s the 2TB model.
 

RafterXL

Member
Oct 24, 2017
351
746
395
My XPG Gammix S70 reads at 6200mb/sec. Not bad for $139. Heatsink is too big but I just screwed the one side on the cover without the clip, and I've been running games on it all night, heat isn't an issue.
 
  • Like
Reactions: Tripolygon

Kupfer

Member
Nov 20, 2018
956
1,517
450
I don't see any reason to add new drive.
There is plenty enough space for me
Same here, I have like more than a handfull of games on my PS5 and still room for more.
I don't need to have my whole library installed nor games, which I dont play regularly.
Otherwise - if somneone is a huge CoD fan, I get it why you need that 4TB drive.
 

M1chl

Currently Gif and Meme Champion
Dec 25, 2019
12,130
23,331
1,045
Prague, Czech Republic
The storage on your phone or iPad is soldered on. How many bricked phones and iPads do you know of due to storage failure?

Apple was confident enough in the longevity of modern SSD's to solder them onto Macbooks too now. I think Apple probably knows better than most GAF posters how durable SSD's are these days.

The useful lifetime of PS5 will end long before storage failure due to flash memory will be a thing.
I mean how many data's you are moving on iPad vs PS5. I admit I wrote this post at the beginning of the gen:

Why both manufactuers sporting soldered SSD on the board? | NeoGAF

And since the SSD on PS5 could act as a RAM extension for the games (as the presentation goes), I think it's legitimate concern. Besides that 2019 MacBook, which I wrote from back then shit the bed, because of SSD. Yeah I did a lot of compilation and that sort of thing (moving large amount of data/day), it was warranty repair, so it's I guess okay (also only 256GBs, so less reserved cells). But data unrecoverable (granted not big of deal for console). So it depends the usage of the drive, it depends how much cycles each memory cells have to withstand. Doing server tech a sort of side job also and I had a quite a few NVMe drives starting to wear out (they well, they are pretty much write something all the time, it's in science institute and since it has HDD backup in each day, it simply does not matter that they last year and half, it's more of a buffer).

So for me it depends how the SSD will be used, but based on the presentation it seems like you can use it as virtual ramdisk. Also this obviously comes towards both consoles, but back then it wasn't exactly known how S|X SSD looks like.

So yes you can say that I am concerned it and my favorite thing would be just to have some cheaper console with NVMe slot, so I could sleep well in the night. In my life I have been unfortunate enough to lost quite a bit of SSD drives. Not sure if curse.

Let's wish our drives to stay healthy for a loong time my dudes and dudettes.
 

Shmunter

Member
Aug 25, 2018
11,034
24,964
835
I mean how many data's you are moving on iPad vs PS5. I admit I wrote this post at the beginning of the gen:

Why both manufactuers sporting soldered SSD on the board? | NeoGAF

And since the SSD on PS5 could act as a RAM extension for the games (as the presentation goes), I think it's legitimate concern. Besides that 2019 MacBook, which I wrote from back then shit the bed, because of SSD. Yeah I did a lot of compilation and that sort of thing (moving large amount of data/day), it was warranty repair, so it's I guess okay (also only 256GBs, so less reserved cells). But data unrecoverable (granted not big of deal for console). So it depends the usage of the drive, it depends how much cycles each memory cells have to withstand. Doing server tech a sort of side job also and I had a quite a few NVMe drives starting to wear out (they well, they are pretty much write something all the time, it's in science institute and since it has HDD backup in each day, it simply does not matter that they last year and half, it's more of a buffer).

So for me it depends how the SSD will be used, but based on the presentation it seems like you can use it as virtual ramdisk. Also this obviously comes towards both consoles, but back then it wasn't exactly known how S|X SSD looks like.

So yes you can say that I am concerned it and my favorite thing would be just to have some cheaper console with NVMe slot, so I could sleep well in the night. In my life I have been unfortunate enough to lost quite a bit of SSD drives. Not sure if curse.

Let's wish our drives to stay healthy for a loong time my dudes and dudettes.
Not RAM expansion. There are no swap files or virtual memory. The ssd’s purpose is solely for putting stuff into ram quickly. Loading , streaming, etc.
 
  • Like
Reactions: Allandor