Someone on Era says that this isn't because of the SSD speed differences, but because of the Cpu speed difference (3.8 for BC games on the XSX).
Damn you guys are at it again celebrating too early.
This is bc games with no code changes.
When a code change is done ps5 can load games like spiderman and ratchet in 2 seconds.
The xbox has shown games using the velocity architecture like dirt and taking up 15 seconds depending on whats loading and how heavy it is.
Their software solution hasn't made up the gap, since its not used here in bc. Neither is all that hardware in ps5, designed to do so for next gen games or re-coded games.
Ps5 will beat, by multiples, series x load times on games actually meant for the system or re coded.
Uh, wasn't the common talking point for the SSD I/O in PS5 that it would be easier to leverage and "just worked"? It'd seem like MS's XvA is the one that has to be more explicitly programmed against. Also DiRT 5 as was being tested by people earlier wasn't yet finished, it's also not a multiplat natively designed for either next-gen platform, so....?
I did. Before that I said if you actually believe that XSX has faster I/O than the PS5 say it. Don't toe the line, actually say it so you can be held accountable.
You're being awfully defensive. People wanted to see proof in results, now we're getting them. We'll get more as time goes on. I don't think it's impossible to acknowledge that there are clearly BC titles Series X (and S) beat PS5 in when it comes to load times, but there might be a few BC titles PS5 beats those systems in WRT load times, that haven't been tested yet.
Personally I still stand by the idea that MS's solution punches above its weight in this regard (surprisingly moreso than even I thought would be the case initially), but I still expect Sony's solution to be the faster overall. Just, not by anywhere near the margin the paper specs would suggest. Real-time use-case performance, I wouldn't expect any blow-outs.
People don't understand this and people need to hold off on celebrating. You can run SSDs on XB1 and PS4 and you're not getting true SSD speeds.
That doesn't matter too much; look at the chart Longdi posted above and apparently PS4 Pro has faster cold boot time than both PS5 and Series X, but PS4 Pro wasn't designed with SSDs in mind.
BC games (and all games in general), even if they aren't programmed explicitly to the full capabilities of a system's I/O design, they still have the raw hardware to leverage. People need to keep in mind the BC games being tested here by and large aren't even leveraging most of MS's XvA features; it's unknown if they are leveraging Sony's raw or not. I'd assume they aren't programmed against most of Sony's featureset.
At the very least we can already get a glimpse of how these solutions will perform in practice and it turns out it's not going to be the clear-out blow-out in favor of PS5 a lot of people were repeatedly saying, because there's still other parts of the system architecture that have to work with the data coming off the SSDs. I think once both are getting pushed to their limits, Sony's solution will maintain a lead in terms of I/O, but the real-world margin between them and MS on this note is going to be a
lot smaller than people have been conditioned to think it'd be going simply off some of the (few) specs on the I/O we got on paper early in the year.
Guys, this is not a raw measure of I/O. It's a measure of how the two machines handle backwards compatibility.
I'm amazed some of the more technical posters are using this moment to talk about "equality of solutions" or something. It seriously isn't.
I'm especially surprised at thicc.
The writing was always on the wall that the XSX would have a better BC solution than what Sony was doing, at least in my opinion. That's what these results are saying and perfectly conform to my expectations. Has dick all to do with the SSD solution in either console beyond a minimal point.
Surprised how? I've followed info on both systems for a long time now, and looked into various research papers and patents with regards to them. There are a lot of things regarding how data read from the SSDs is actually used that a lot of people haven't been considering.
Yes it's true BC optimizations count for a good deal here, but in the same breath you have DF being very impressed with PS5's BC, so would that not mean Sony have done a lot of optimizations with their own BC? You can't have your cake and eat it too; if BC is at the root then we'd be seeing that manifest in actual BC software performance on PS5 being behind, too, but that doesn't look to be the case.
I still think PS5's solution will maintain an edge but there's nothing outside of some paper specs (which did not list a lot of details, like random read speeds on the NAND modules...I could try getting documentation on that if the specific part models of the NAND for both systems was more clearly listed (and the documentation even available in a PDF)) suggesting it'll be a blowout. This isn't a slight at Sony's solution, just a tip of the hat to MS's for being more elaborately planned out than some people were willing to consider.
The solutions are actually more apples-to-oranges in the first place which is why I always thought going simply off paper specs wasn't the best idea particularly surrounding the SSD I/O because that seems to be the area the systems diverge from the most. The way some people were treating MS's solution however they would think it was an afterthought, which didn't make sense even months back, and we're starting to see why that wasn't a good notion since we're getting actual data showing how these systems run some of these games, something we've been waiting to see since the end of last year.