Yeah has absolutely nothing to do with ssds or their compression techniques. In fact if they used the same hard drives as before we would get exactly the same results.
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
PS5 / XSX Backwards Compatible Load Time Comparisons
- Thread starter Raigor
- Start date
DF said they have a BC load time video coming I think, so hopefully we'll see some thorough testing of it soon.
Oops, you're right, DECREASING! 😅 my brain has melted from the elections.
🤦♀️ PS5 is running in BC mode with everything clocked down to simulate a PS4.... These results means nothing when it comes to next gen or optimized BC games.
Except it's actually been the opposite.
BC enhancements have actually greatly exceeded general board expectations here.
But games run with way higher performance in terms of FPS and dynamic res. So that's not entirely true. Sounds like boost mode may be the standard.
Pretty surprising that Xbox has the lead here, specially if you think about that Xbox basically has the troll move of "Well, you think this is fast? Now where the game is loaded once, next time you come back after playing 5 other games, it's 10 seconds and you're in again at the exact point you left it 2 weeks ago, quick resume baby".
But in general, fantastic job on both consoles, both trying to get the absolute best out of it.
But in general, fantastic job on both consoles, both trying to get the absolute best out of it.
That explains why sony didn't release load time comparison videos with BC titles unlike MS. 1~2 weeks until we finally get to see the whole picture with next gen titles.
This Video by IGN clearly show the difference between BC and optimised for PS5:
The loading difference between PS4 Spiderman and PS5 Spiderman is 0:43 to 0:05 and 1:07 to 0:11 on Miles Morales. That's about what we can expect when games are optimised for the PS5 SSD.
If you compare that to the other load times it becomes clear, that that isn't a "Hardware problem".
The loading difference between PS4 Spiderman and PS5 Spiderman is 0:43 to 0:05 and 1:07 to 0:11 on Miles Morales. That's about what we can expect when games are optimised for the PS5 SSD.
If you compare that to the other load times it becomes clear, that that isn't a "Hardware problem".
Weird but at the same time MS has had BC experience for a few years so no wonder it takes better advantage of the hardware, just wonder if it's something that can be fixed
There are a group of people who expected Sony to show at least some love for their awesome catalog of games pre-PS4. I know I'm in that group, and so while I think this news is great I'm still soured that PSX and PS2 have been forgotten again.
That's a comparison between systems, no? They should do vanilla Spider-man BC on PS5 then Upgraded Sider-Man via Miles Special Edition and look at the differences.
It's Vanilla Spiderman. It would have stated Spiderman Remastered otherwise.
Regardless BC software optimization for Sony is probably a fraction of what MS invest into it and I don't think they put any work in optimizing around the 6 lanes available on the PS5 ssd (what makes it so fast) in regular BC games so these results really tell us nothing about next gen.
【PS5】話題の爆速SSDの実力は!? PS5 vs PS4 Proロード時間比較
PS5注目情報まとめhttps://www.famitsu.com/news/202011/05208871.htmlカスタムSSD 高速化の秘密https://www.famitsu.com/news/202011/09209074.html2020年11月12日発売予定のプレイステーション5は高速SSD（So...
www.youtube.com
I see, hmmm? Very surprised about the results. Wonder if there's some weird drive speed cap going on for non-optimized games.
Likely this. The XB BC team went in harder on implementing BC, while we knew from Cerny's Road to PS5 talk (backed up by DF's PS5 BC vide) that the PS5 was just using PS4 and PS4 Pro legacy modes - original coding that relies on mostly original clock speeds (CPU/GPU/Loading).
I suspect cross-gen games and BC titles may indeed be faster on Xbox than PS5, while games built for next gen specifically from the ground up will account for the PS5's SSD. That goes true for Xbox's GPU features as well.
Launch period and cross-gen is always a nebulous period before new hardware features kick in for true.
Last edited:
How is that a good way to compare the consoles? Microsoft isn't going back and patching games to change their file structure to be optimised for the new ssds either.
A fair comparison would be dirt 5. Do journalists have the game for PS5?
Sure.
An all in one PS5 that can do it all would actually be a great selling point, I think.
I'm guessing the issue is that the PS5 titles are able to use the dedicated hardware for getting data and the PS4 titles are being slowed down by the CPU having to do that.
XSX probably has done some work to optimize that latter process.
XSX probably has done some work to optimize that latter process.
Do you want this?【PS5】話題の爆速SSDの実力は!? PS5 vs PS4 Proロード時間比較
PS5注目情報まとめhttps://www.famitsu.com/news/202011/05208871.htmlカスタムSSD 高速化の秘密https://www.famitsu.com/news/202011/09209074.html2020年11月12日発売予定のプレイステーション5は高速SSD（So...
oh my fucking god the ps4 pro version is so bad
jesus
it's almost like they made that version specifically to showcase how much better ps5 is and everyone needs to buy one lmao
This is a complete lie, you are looking at full boot times.
The actual load times are way faster.
I still have my fingers crossed that they do a software update down the road, heck I would pay for the feature to be honest.
I'm hoping DF looks into this, as PS5 games seemingly make all use of that speed. BC games throttling?
The games are full booting faster on Series X. So clearly they are not taking advantage of PS5 SSD. Some games are close but others are not even close.
That is entire different from next gen games which are optimized for the hardware.
Not yet. It's been announced, it'll be free for existing owners, but it won't be out until after Cyberpunk ships.
Yikes
Can you please explain to me how XSX is equivalent to a ps4 now ?
Your statement was completely wrong. The ps5 loading a bit slower than XSX because it has a unique architecture does not mean it's not significantly boosting the ps4 load times.
Where did i compare the console? I'm Just stating that BC loading times are in no way representative for how this will Play out on Next Gen, as the Spiderman games have shown.
Games with long boot times are CPU limited.
Every single step of the asset loading process needs to be optimized for next gen, otherwise you get the same slow loading that appear on PC games running on a $1000 SSD. A single bottleneck on one of the five steps in the process makes all the other optimization useless. So it requires some serious work from devs, no BC method can magically rewrite the entire storage code.
Every single step of the asset loading process needs to be optimized for next gen, otherwise you get the same slow loading that appear on PC games running on a $1000 SSD. A single bottleneck on one of the five steps in the process makes all the other optimization useless. So it requires some serious work from devs, no BC method can magically rewrite the entire storage code.
No but you are wrong, you can pretend it's terrible when it's not.
Aren't we talking about the video in the OP showing same game booting on PS5 vs Series X where Series X handily beat PS5 despite having 2x the SSD speed?
That is my point. In some games they are close and in others not close. To which i said, PS4 games are not really taking full advantage of the PS5 SSD as much as i thought they would compared to Xbox Series S/X which has 2x less bandwidth.
That is not an indictment on how next gen games will perform.
Sorta close but not really
Not even close
I'd love to know the exact bottleneck, if it is indeed the BC games being CPU limited. As clearly PS5 developed titles have zero problems with the new bandwidth.Games with long boot times are CPU limited.
Every single step of the asset loading process needs to be optimized for next gen, otherwise you get the same slow loading that appear on PC games running on a $1000 SSD. A single bottleneck on one of the five steps in the process makes all the other optimization useless. So it requires some serious work from devs, no BC method can magically rewrite the entire storage code.
Yeah, I think people should hold off making any conclusions from these comparisons because next gen games will show real advantages of PS5's SSD.
This better work much faster when optimized, this is pathetic.
Weird... looks like the PS5 is intentionally limiting the storage access speed of certain games. I can only assume this was done for compatibility reasons. Games may crash when they load much faster than they ever anticipated. I can only assume that they will improve as time goes and developers release individual patches.
It seems like Sony is virtualizing legacy HDD in some ways.
Miles Mores PS5 cold-boots in 4 seconds and loads saves in 1.5 seconds.
Miles Mores PS5 cold-boots in 4 seconds and loads saves in 1.5 seconds.
Games with long boot times are CPU limited.
Every single step of the asset loading process needs to be optimized for next gen, otherwise you get the same slow loading that appear on PC games running on a $1000 SSD. A single bottleneck on one of the five steps in the process makes all the other optimization useless. So it requires some serious work from devs, no BC method can magically rewrite the entire storage code.
Makes sense. I would also imagine that the Xbox would be at a slight advantage here due to the 3.8 Ghz frequency without SMT enabled, no?
even if it's mostly CPU-bound, PS5 and XSX have way faster CPUs than their predecessors, and they both load more demanding next-gen games way faster
there has to be some other bottleneck in the software architecture in last-gen games
there has to be some other bottleneck in the software architecture in last-gen games
The software stacks and how they map to and use hardware are a big deal in getting the best out of the SSDs - whether it's 'velocity architecture' or whatever you want to call Sony's.
Last gen games don't use these stacks. They each use i/o stacks that are both different from each other, and different from the next-gen stacks.
All we can really say for sure right now is that MS on average has seemingly mapped their prior gen file i/o stacks to their new system better than Sony did, from a speed point of view at least (assuming, in these cases, the games' load times are bottlenecked on I/O and not something else).
With next-gen stuff, we'll have to wait and see.
I think the difference will be negligible in-game. The comparison shows the game being booted up. Waiting an additional 20 seconds for a game you're likely to play for a couple of hours at a time isn't that bad. The difference between a regular load screen when using quick travel from one area to another shouldn't be that big. Probably just a few seconds at most. Those are the load screens you'll actually be seeing when playing.
Yeah, this is the most likely reason in my opinion, and it wouldn't really surprise me either. It could also be that they haven't done a lot of optimizations on the software side.
The hardware is obviously fast enough to do much better, so that can be ruled out.
It does. Check IGN video where they compare the Spiderman games ( both optimised die PS5 SSD ) loading time difference between PS4 and PS5.
I think yes, while it should be the norm to use SMT with modern code, it's assumed old gen code cannot take advantage of it. It looks like Sony didn't want to sacrifice yield just for BC.
Also some games might be running is a the slow clock fallback mode, or with capped storage speed, if they discovered a nasty race conditions when testing those games. Those would need a patch.