Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

Status
Not open for further replies.
Fair enough, but it doesn't support theories of some wild hyper customized SSD HW solution. I'm sure they'll use some kind of PCIe 4.0 NVMe drive anyway and customizations will be on software stack size instead.

There are 2 scenarios that I see could be going on right now.

1. PS5 SSD is clearly superior to XSX SSD. All stars have aligned for Sony and they have concocted a superior SSD solution that MS (and partners) cannot possibly match.
1a. It's just that MS cannot possibly match it in time because it was too late when they learned of what Sony was doing.
1b. MS is unwilling to match Sony's speed and would settle for an off-the-shelf solution.

2. MS's solution is also highly customized and there is no significant difference between MS's and Sony's SSD solution.

3. .....:runaway:
 
Do Sony consoles ever really launch with blockbusters? Not that I can remember anyway.

PS5 with say.. Horizon 2 could be the first.
 
Wasn't just that figure, was also saying Sony would have zero.

Thought you was one of the people pushing back on my view of the importance of gen exclusives, but could just be misremembering.

But, yea agree with what you've said.

Yeah, I'd be surprised if there wasn't a single PS5 exclusive of any kind at launch. Especially as some Japanese publishers didn't fully embrace XBO. Persona 5, for example, showing that even though Atlus was releasing some games on XBO, they weren't releasing all of their published games on XBO.

Then again, maybe all those trips Phil Spencer has made to Japan that he keeps mentioning are going to pay off next gen? After all, I was quite shocked to see the Yakuza series coming to XBO and Game Pass.

One of the reasons that Atlus stated for Persona 5 not coming to other platforms was that they couldn't afford it. Perhaps, Phil Spencer is going to again try what they tried with X360. Throw money at Japanese developers to get their games on the Xbox platform. Even if it isn't to make them exclusive but just to make them appear on Xbox in addition to PlayStation, that would be a win for MS.

Also this would mean that Godfall wouldn't come in time for the launch window. And the Gearbox representatives at the Game Awards stated it was coming Holiday 2020...

Regards,
SB
 
0.5s to 1s not a big deal.
30s to 1 minute is a big deal. But both companies promised to virtually eliminate load times
A promise means nothing. Spencer once promised to bring VR to XBX.

Unfortunately for them, MS already stated their SSD speed: about 2GB/s max (in the best case scenario as usual because it's always the best case scenario when they state numbers) and we know a few things from the Sony patent: in most cases I am expecting 2-3 times faster loadings on PS5 in real games.

You have chosen odd examples (1s or 1mn loading on a regular SSD are very rare right now). I think it's more going to be like: 1.5 s instead of 4s, 4s instead of 10s, 10s instead of 25s. #SSDGate
 
A promise means nothing. Spencer once promised to bring VR to XBX.

Unfortunately for them, MS already stated their SSD speed: about 2GB/s max (in the best case scenario as usual because it's always the best case scenario when they state numbers) and we know a few things from the Sony patent: in most cases I am expecting 2-3 times faster loadings on PS5 in real games.

You have chosen odd examples (1s or 1mn loading on a regular SSD are very rare right now). I think it's more going to be like: 1.5 s instead of 4s, 4s instead of 10s, 10s instead of 25s. #SSDGate
You said 100% improvement; I was just using your numbers. If you take at face value that Sony is barely on the scale for loading. Then 100% of that is still virtually nothing. I assume you chose 100% because it's 2.5GB/s vs 5GB/s.

But aside from looking at raw bandwidth numbers, you need to see how that translates to overall performance. Sony is saying near 0, loading times are non-existent. So double near nothing is meaningless metric to discuss.

so either Sony has to have a 1000% better solution; 1s to 10s or this discussion isn’t about anything really that substantial.

If you think Sony has made something that is 1000% better than what's on the current market; that's quite an accomplishment considering no other manufacturer has been able to.

and the only way Sony is going to get a 1000% improvement is going to be ReRAM if I recall the numbers correctly.

And now we are into some silly cost/subsidy discussion. Because then you'd have to believe in short time period:
Added BC
Made it run cool
Added ReRAM
A large drive to hold games
and continue to push VR
All the while making something equivalent to as powerful as XSX.
For the same price points.

....
So you're banking basically that MS has been smashing eight balls the whole time since X1X.
 
Last edited:
Launch line up compared,

PlayStation 4 launch exclusives (11):

Blacklight: Retribution
DC Universe Online
DriveClub
Killzone Shadow Fall
Knack
Minecraft
The PlayRoom
Resogun
Tiny Brains
Warframe
War Thunder

Xbox One launch exclusives (11):

Crimson Dragon
Dead Rising 3
Fighter Within
Forza Motorsport 5
Killer Instinct
LocoCycle
Peggle 2
Powerstar Golf
Ryse: Son of Rome
Zoo Tycon
Zumba Fitness: World Party

I'm a Sony guy in large measure because of the exclusives but I think it's entirely plausible mS launches with more first party exclusive titles. They did just buy a number of studios.

As the generation settles I would expect Sony to surpass MS first party support but at launch who knows?
 
Not the thread to revisit list wars so let's not delve too deeply on it. But to be pedantic Minecraft launched on both platforms the same time and not during launch at all but in September 2014.
https://en.m.wikipedia.org/wiki/Minecraft , https://www.ign.com/articles/2014/09/04/minecraft-ps4-edition-release-date-confirmed , https://www.cnet.com/news/minecraft-for-xbox-one-to-launch-on-friday/


Especially when the rumor of launch titles in 2020 is such garbage.

Even if one side has more NextGen launch titles, they both will have their entire current catalog of games available (most likely) with extremely compelling experiences.
 
A promise means nothing. Spencer once promised to bring VR to XBX.

Unfortunately for them, MS already stated their SSD speed: about 2GB/s max (in the best case scenario as usual because it's always the best case scenario when they state numbers) and we know a few things from the Sony patent: in most cases I am expecting 2-3 times faster loadings on PS5 in real games.

You have chosen odd examples (1s or 1mn loading on a regular SSD are very rare right now). I think it's more going to be like: 1.5 s instead of 4s, 4s instead of 10s, 10s instead of 25s. #SSDGate
I missed that, where did they say it was 2GB/s?

There's a lot of software work to get the most out of fast main storage, but I wonder if the crypto and LZ compression and JPG codecs might the biggest bottleneck on the hardware side. If they have a 5GB/s drive, no matter how they optimize the game's API, they need decompression and crypto asic that can operate at that speed. If they didn't plan for this very early in the SoC design, they can't just change that now.

We do lossless playback at work from fibrechannel block devices, and the frames are compressed in zip. At 1 GB/s it takes almost all of of an intel processor to decompress into a frame buffer with the rest of the cpu dealing with file system calls. Forget 5GB/s as a backround loading process without a big chunk of dedicated asic.
 
Spencer stated 40x I/O vs last gen.
Works out to be 2.5 GB/S if math is done using Xbox one.
 
If they have a 5GB/s drive, no matter how they optimize the game's API, they need decompression and crypto asic that can operate at that speed. If they didn't plan for this very early in the SoC design, they can't just change that now.

Does it have to be in the SoC design? A secondary decompressor chip wouldn't work?

Forget 5GB/s as a backround loading process without a big chunk of dedicated asic.

How about if they allot a portion of the SSD to serve as a scratchpad to uncompress large data to it. At 5gb/s it could fill a 16gb ram in 4 seconds without involving the cpu or without having to spend for a powerful decompressor.
 
That was said in the same jest as Sony saying Ps5 was 10 times faster.

So far we have 0 direct specifications.

https://www.straitstimes.com/tech/new-faster-xbox-coming-out-next-year

In a teaser shown at the briefing, Microsoft said the solid-state drive will be 40 times faster than the drive in the current Xbox.

Sony had said that its upcoming PlayStation will be able to load 10 times faster than the current generation of the console.​
 
In a teaser shown at the briefing, Microsoft said the solid-state drive will be 40 times faster than the drive in the current Xbox.

Sony had said that its upcoming PlayStation will be able to load 10 times faster than the current generation of the console.​

Using typical fanboy logic that would mean: "XSX ssd four times faster than PS5 ssd confirmed! LOL at Sony!"
 
I can't imagine either going with anything other than a pcie 4.0 nvme at the end of 2020.

Regardless of what type of controller it's using, and regardless of whether it's soldered.

Cerny's presentation was before nvme pcie 4 were available, and he said the current implemetation in that secret hardware was slower than the final product. That fits well with using a pcie 3 on a standard PC at this point in time (the complete console not being finished). And 4.0 for later devkit revisions.
 
Does it have to be in the SoC design? A secondary decompressor chip wouldn't work?



How about if they allot a portion of the SSD to serve as a scratchpad to uncompress large data to it. At 5gb/s it could fill a 16gb ram in 4 seconds without involving the cpu or without having to spend for a powerful decompressor.
The SoC would have the pcie lanes anyway, so it's the best place to put it. If it's on the south bridge it multiplies by two the number of high speed external channels (nvme->south bridge->SoC) or even worse with a dedicated external chip(nvme->SoC->Codec Chip->SoC). And then one of those external paths is uncompressed bandwidth.
 
Spencer stated 40x I/O vs last gen.
Works out to be 2.5 GB/S if math is done using Xbox one.
Current HDD on base consoles max out at about 50MB/s (theoretical, never reached in-game). You are using XBX numbers (62.5 MB/s here). And 50MB/s is already pushing on base consoles and generous.
 
Current HDD on base consoles max out at about 50MB/s (theoretical, never reached in-game). You are using XBX numbers (62.5 MB/s here). And 50MB/s is already pushing on base consoles and generous.

Sony only claimed 10 times faster. So why can't everyone remain reasonable and stop reading into PR statements?
 
I'm a Sony guy in large measure because of the exclusives but I think it's entirely plausible mS launches with more first party exclusive titles. They did just buy a number of studios.
That's counter-logical. Assuming two years development of a good launch title, Sony having first party studios for years places them more likely to have more 1st party titles for PS5 than MS who's studios are too new to have much ready for XBsX.

We're just talking timelines and intentions here, not studios. Sony having less as rumoured would be because they want to maximise profits from their current 80+ million install base, rather than drive adoption to the PS5 (although they did state they want the fastest gen switch ever). MS having more 1p titles would mean smaller titles leveraging the new studios MS has had for 18 months prior to launch and giving them a AA game remit. Whatever they decide, MS are less able to have more 1P titles for XBsX than Sony could have for PS5 if Sony wanted to get all its studios working from 2017 for a 2020 new console. Which would be a dumb plan - too many AAA games would watered down sales for a small userbase!
 
Status
Not open for further replies.
Back
Top