• 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.

[DF] PS5 Firmware 2.0 Tested

Excited Lets Go GIF by Travis
 

ethomaz

Banned
except for battlefield 5 , every game they tested has at least 2 seconds loading advantage in favour of the samsung SSD
Just 2 out of 10 game tests showed 980Pro with at least 2 seconds advantage.
Same for internal advantage 2 tests of 10.

The others 6 tests are near tie or below 2 seconds difference.

The 980Pro had an advantage? Yes but it was very close to tie… that probably means faster SSDs won’t give better times because the others parts like the I/O complex is limited to that internal.l speeds.

We are talking here in 1000MB/s (18%) more speeds that didn’t translate in better times on PS5. The difference between 5500, 6000, 6500, 7000 MB/s will be very small.

The big question is how the sub-5500MB/s SSDs will results in stressful gameplay parts.
 
Last edited:
He was saying external like it was a USB drive. The M.2 expansion isn't really "external" since you install it in the system.

I've seen people call the NVME an external drive. You can't benchmark PS5 off USB externals since PS5 games can't he played off them.

For the most part Digital Foundry was comparing the soldered drive to the NVME expansion.
 

RoadHazard

Gold Member
Very uninformative video. No comparison of loading into a proper PS5 game (Plague Tale is a patched PS4 game), and no test of how under-speced drives perform (of course a 7GB/s drive is gonna fare well).

All the PS4 games loading a little faster with the Samsung drive is not surprising at all, since those games will not be using the new storage APIs and will therefore simply be helped by higher raw throughput.
 
Last edited:

Kenpachii

Member
They hint at it being artificially limited due to the fact that write speed isn't a factor during gameplay... I wonder if they're trying to increase then longevity of the hardware?

Probably not a priority. Cheap SSD's always had this, high read performance, shit write performance. The reason for this is, u read more out then u write towards it.

Not much of a issue tho, read is what matters the most.
 

longdi

Banned
Rich said the speed vary even when you turn ON/OFF the internet.

No he wrote this, "However, in copying the data back to the internal SSD, Xbox Series X was always significantly faster."

They hint at it being artificially limited due to the fact that write speed isn't a factor during gameplay... I wonder if they're trying to increase then longevity of the hardware?

I would think a faster write speed helps to maintain longevity, write it down fast and turn down the power, rather than prolonged slower writes

Probably not a priority. Cheap SSD's always had this, high read performance, shit write performance. The reason for this is, u read more out then u write towards it.

Not much of a issue tho, read is what matters the most.

Possible..Epic China engineer did leaked that Sony choice of their more parallelised ssd was to lower for costs rather than performance reasons..
 
Last edited:

Fake

Member
except for battlefield 5 , every game they tested has at least 2 seconds loading advantage in favour of the samsung SSD

The Witcher 3: White Orchard was the same from the SSD without heatsink.

PS4 App Loading Times (Seconds) Stock Internal PS5 825GB Samsung 980 Pro 500GB Samsung 980 Pro 500GB/Heatsink
The Witcher 3: Novigrad Centre - 48.01 - 45.07 - 44.67
The Witcher 3: White Orchard 22.88 22.88 22.70
Cyberpunk 2077: Ripperdoc 41.12 43.93 38.48
Cyberpunk 2077: Maelstromers HQ 33.08 31.08 31.12
Fallout 4: Commonwealth 16.50 14.87 15.13
Fallout 4: Diamond City 15.27 14.01 14.17
Final Fantasy 15: Lestallum 29.18 27.43 27.60
Final Fantasy 15: Hammerhead 24.57 24.08 23.87
Battlefield 5: Nordlys 31.50 34.57 34.57
Battlefield 5: Tirailleur 31.58 34.63 34.53

No he wrote this, "However, in copying the data back to the internal SSD, Xbox Series X was always significantly faster."

Is on the video, not the article.
 
Last edited:
They test Ratchet & Clank in the first few minutes, did you even watch the video?

If the video indicates anything that slower drives should see even bigger dips at those points.

Maybe this is what Mike was making a reference to with his testing?
 

MastaKiiLA

Member
Looks like expanding storage will be a good idea. I won't both next year's drives land, but looks good. Shouldn't have to worry about losing any performance with expanded storage. It was a good idea going with the storage choice they made. PC tech was already up to speed at launch. Probably way ahead on write speed, based on results people are seeing.
 

YCoCg

Member
Do you understand what "loading into" means? Going through a portal is not the same thing as loading a save from the main menu.
They mentioned it's roughly the same, did you really need to be shown them loading a game when the portal shift sequence stresses the game more?
 

ethomaz

Banned
Very uninformative video. No comparison of loading into a proper PS5 game (Plague Tale is a patched PS4 game), and no test of how under-speced drives perform (of course a 7GB/s drive is gonna fare well).

All the PS4 games loading a little faster with the Samsung drive is not surprising at all, since those games will not be using the new storage APIs and will therefore simply be helped by higher raw throughput.
This.

They forget to test gameplay too.
 

RoadHazard

Gold Member
They mentioned it's roughly the same, did you really need to be shown them loading a game when the portal shift sequence stresses the game more?

How do you know portals stress the game more in terms of loading? Loading into a save from the main menu is generally faster than going through a portal, but that's partly of course because the portal transition takes longer than it actually needs to in order to look good.

I just don't think the portals are a very good test of raw loading, since that transition is doing a bunch of other stuff in real-time that might be slowing it down more than loading the data does. We don't know if the SSD or the CPU or something else is the limiting factor there. Loading a save is a more pure test of how quickly loading happens, because you just get a 1-2s black screen and then you're in the game.
 
Last edited:
Top Bottom