Sony PlayStation cross-platform game strategy

[Deleted member 13524]

IMO RC should theoretically have bigger audience than gow. It’s appealing to both genders it’s a Pixar style shoter-plat former/adventure game with sense of humor which can be enjoyed by younger and older players. And it has deep history with ps it started as ps2 game. So it’s definitely one of the most iconic games on PlayStation.

Isn't that a bit like saying Disney's Coco should have had higher ticket sales than Avengers Endgame because the former caters to a wider audience?

The biggest part of the non-mobile gaming population is now at 25-45 IIRC, and those tend to like more adult-oriented games than Ratchet&Clank. It's also a bracket of the population where many of them started to have kids, so the father+son theme really resonated with them (just like TLoU back in the day).
Not that there's anything wrong with R&C, but from my pool of gamer friends I'm probably the only one really looking forward to play Rift Apart, whereas all of them would buy GoW Ragnarok as soon as they could.

Don't worry people R&C will sell a few millions more copies when it releases on PC in a few months.

I don't think Rift Apart will be on PCs before DirectStorage is available, widespread and stable.

 

[mr magoo]

Isn't that a bit like saying Disney's Coco should have had higher ticket sales than Avengers Endgame because the former caters to a wider audience?

.

Yeah but shouldnt Fortnite have bigger audience than doom eternal I think its valid assumption. Also Avengers have bigger/wider audience than Coco ( i dont even know what Coco is) the comic books are known all over the world i hear about "Coco" first time.

 

[zed]

Don't worry people R&C will sell a few millions more copies when it releases on PC in a few months.

wasnt R&C selling better than any of the previous rachet games?
But yes I assume it will eventually come to the PC, though not for a couple of years (and we know how good ppl are at delaying their instant gratification, hence noone brought cyberppunk until they had released a working version)

 

[Deleted member 11852]

If someone has absolutely no intention of ever buying a PC for games (because the hardware is expensive, especially now or they just aren't interested in the hassle of owning a gaming PC), then how else are they going to play the exclusives that are available on the Xbox?

Microsoft are committed to releasing their games on PC and Xbox. If there are two options (Xbox and PC), and consumers discount one (PC), that doesn't mean there is only one option - and that it's in some way exclusive. That's choice. That's life.

Microsoft have made accessing some of their games really easy with xCloud. So you don't even need an Xbox even if you do not intend to buy a PC.

It's far more likely for a console owner to get a competitor's console than it is for a console owner to get a PC. Not only are consoles similar in cost (thus will fit in their budget), but they also more easily slot into the same gaming environment (the living room).

I don't think there is any published data on which you can confidently make this statement, and it does not reconcile with Sony's statement on the rationale for developing PS4 Pro (and they do have data) but if want to play all of Microsoft and Sony first party exclusive games and also want to play PC games then a PS5 + PC is your best bet because all of your bases are covered. Similarly Switch and PC if you're need to chose your console. Xbox and PC is mostly redundant. If PC games are not a factor, and you don't need a PC for other things, then consoles may be the more appealing option.

I own the suite - PS5, Switch, XSX and PC. And only dilemma this brings is which platform to buy games on.

 

[Karamazov]

I don't think Rift Apart will be on PCs before DirectStorage is available, widespread and stable.

I was only making a little joke, but i don't think they'd need to compensate any slower i/o or ssd on pc, they would only have to target a few more gigs of ram.

 

[Deleted member 11852]

I was only making a little joke, but i don't think they'd need to compensate any slower i/o or ssd on pc, they would only have to target a few more gigs of ram.

Maybe a few, maybe a lot.

R&C:RA takes up 33.6Gb on my PS5. Until Insomniac do one of their Spider-Man on PS4-like GDC tech breakdowns on how game assets are stored, it could be that a wide variety of textures, geometry, shaders and audio effects would need to be kept in RAM "just in case" - depending on the world diversity, and the diversity of the parallel worlds you may imminently leap into through through portals. In addition to the base memory needed to run the immediate game world, do you need an extra 8Gb? 16Gb? 24Gb or most of that 30+ Gb? ¯\_(ツ)_/¯

Not all assets are going to be required at all times but if I/O is your bottleneck, you still need to be filling extra RAM with all the possible assets that might soon be needed (above and beyond whatever they're doing on PS5) which is an extra layer of asset-management.

Just shoving everything into RAM, like in days gone of installing games to a temporary RAM disk is the ultimate brute-force approach to improving load-times but the game will need to run on more modest PC hardware.

 

[Karamazov]

but you can't have more than 12-13GB at once on PS5 for games, and you feed that at, let's say the max theorcial 22GB/S even if it's not realistic i guess, ram speed is faster than this in general, so maybe with 4-8GB more you'd use that as a cache to load what's needed.

 

[PSman1700]

Maybe a few, maybe a lot.

R&C:RA takes up 33.6Gb on my PS5. Until Insomniac do one of their Spider-Man on PS4-like GDC tech breakdowns on how game assets are stored, it could be that a wide variety of textures, geometry, shaders and audio effects would need to be kept in RAM "just in case" - depending on the world diversity, and the diversity of the parallel worlds you may imminently leap into through through portals. In addition to the base memory needed to run the immediate game world, do you need an extra 8Gb? 16Gb? 24Gb or most of that 30+ Gb? ¯\_(ツ)_/¯

Not all assets are going to be required at all times but if I/O is your bottleneck, you still need to be filling extra RAM with all the possible assets that might soon be needed (above and beyond whatever they're doing on PS5) which is an extra layer of asset-management.

Just shoving everything into RAM, like in days gone of installing games to a temporary RAM disk is the ultimate brute-force approach to improving load-times but the game will need to run on more modest PC hardware.

All thats assuming gamers arent using direct storage/nvme combos for their gaming systems. DS is a W10/W11 feature and i doubt there will be many pc gamers looking to play next generation games lacking nvme hardware.

 

[Deleted member 11852]

but you can't have more than 12-13GB at once on PS5 for games, and you feed that at, let's say the max theorcial 22GB/S even if it's not realistic i guess, ram speed is faster than this in general, so maybe with 4-8GB more you'd use that as a cache to load what's needed.

The game install size is 33Gb but that is probably compromised mostly of compressed data. You would expect most of the textures are compressed with kraken so decompressed in realtime, along with any other assets that are compressed and can be decompressed on the fly as well. So the data being in a format that is usable by a PC's hardware may be biggered because not all PCs can decompress all types of data on the fly.

Without knowing how much asset diversity is sprinkled around the game worlds little, perhaps with little thought about PS5's RAM availability because PS5 can just load the data in the blink or an eye, how can you calculate how much data you would need to keep in a PC's RAM to work around slower I/O?

 

[Karamazov]

But can you decompress faster than the max I/O output of the ps5, which should still be slower than extra ram ?

 

[Deleted member 11852]

But can you decompress faster than the max I/O output of the ps5, which should still be slower than extra ram ?

I'm confused, I thought we were discussing how much extra RAM a PC would require if it was SSD or I/O constrained? If PS5 can load basically hundreds/thousands of assets a second whereas the PC can't and R&C has masses of variety of assets, or could need them in any instant, how much extra RAM would be required? I.e. what the sum memory size of all those assets being in RAM rather than being loaded from disc.

 

[Karamazov]

let's say R&C uses 10GB of ram for asset, that can be filled at 5.5GB/S, with 20GB of ram on PC at half the I/O speed it would be more than enough to keep up i guess,
Of course devs would have to optimize to load more stuff in advance to compensate the slower speed, but in the end the result would be similar, Then if you're dealing with an old HDD like in the PS4 you'd need a lot more. $$$$

 

[Deleted member 11852]

let's say R&C uses 10GB of ram for asset, that can be filled at 5.5GB/S, with 20GB of ram on PC at half the I/O speed it would be more than enough to keep up i guess,

If only 10Gb of that compressed 33Gb PS5 is uncompressed assets, which seems like a low-ball. It will have to wait for the inevitable Insomniac GDC presentation but I'd be surprised if there wasn't at least 40Gb of compressed assets in PS5s install. If the game is using them really liberally, you kind of need almost all of it present if you know your drive and I/O can't deliver it.

If you're aware of this when you design the game, it's not a problem. But if you design a game for PS5 then find you have to go and review your asset usage, and micro-manage it then that could be a lot of work. I can't remember his exact words, but Mark Cerny's messaging was that devs really don't have to worry about this stuff any more. Which is great. Until you do.

 

[iroboto]

let's say R&C uses 10GB of ram for asset

seems extremely high imo. If the developers are abusing the SSD, and it uses a fairly efficient svt system, the streaming pool is probably not all that large. I expect most of them to have less than 2GB for pool size. UE5 PS5 demo only required 768MB for instance.

 

[Silent_Buddha]

Microsoft are committed to releasing their games on PC and Xbox. If there are two options (Xbox and PC), and consumers discount one (PC), that doesn't mean there is only one option - and that it's in some way exclusive. That's choice. That's life.

Microsoft have made accessing some of their games really easy with xCloud. So you don't even need an Xbox even if you do not intend to buy a PC.

Hence why there is a distinction between exclusive and console exclusive. Something MS started using years ago and Sony are now starting to use as well.

I don't think there is any published data on which you can confidently make this statement, and it does not reconcile with Sony's statement on the rationale for developing PS4 Pro (and they do have data) but if want to play all of Microsoft and Sony first party exclusive games and also want to play PC games then a PS5 + PC is your best bet because all of your bases are covered.

Does it not? According to sales data PS4-P sold what?: approximately, 20% as many units as PS4? Someone will surely correct me if I got the number wrong.

Of those, how many were people that got the PS4-P just because it was a better PS4 but had Zero intentions of buying a PC? Basically what percentage of that 20% actually seriously considered getting a PC? On top of that how many people that bought a PS4-P got a PC anyway and still bought all their games, minus PS exclusives, on PC?

Obviously you and I do not know. And I doubt Sony truly knows either. It's a similar question that PC developers have wrangled with for years. If piracy didn't exist, how many more copies of games would they sell? Or to put it another since piracy does exist, how many game sales are lost to piracy? In Sony's case, the question is, since PC exists, how many sales are lost because PC exists?

One hint to the possible answers for those questions is that their experience and data with PS4-P might be what led directly to them deciding to start releasing their AAA exclusives on PC. Basically, if the data before and after PS4-P's released show zero impact on PS game sales (exclusive or otherwise) before and after PS4-P's release relative to past sales data for mid-gen PlayStation sales data, then there is no reason not to release exclusives on PC. Basically whether PC exists or doesn't exist the same split for 3rd party sales (a barometer of whether you truly picked up or lost a console buyer) would imply that you neither picked up nor lost any significant number of console buyers. Console buyers (the ones who make Sony money) being those who buy all their games, especially 3rd party games, on their console of choice versus PC buyers (the ones who don't make Sony much money) who buy all their 3rd party games on PC and only buy exclusives on console. That 3rd party licening revenue is why console's exist. So, if there was no change in 3rd party game sales post-PS4-P launch then that would indicate that it had almost zero impact WRT PC buyers.

Thus, if (hypothetical since we don't know) there is no change in 3rd party sales, you might as well maximize profits of 1st party exclusives by releasing them on PC as well.

Another way to interpret that is that due to PS4-P having little to no impact WRT the stated goal of limiting player loss to PC (similar to the argument of whether or not DRM has an impact on sales being lost to piracy), instead they'll see if releasing exclusives on PC will get people to buy a PlayStation console. But I'm skeptical of this due to one key sticking point. Those players, with some exceptions, are still likely to buy all their 3rd party games on PC, thus defeating the purpose of getting them to buy a PlayStation console (the licensing revenue from 3rd party games on console).

In either case, you want to start off slowly by testing the waters and at each step seeing the effects on your core business model, console sales and more importantly 3rd party game sales on your console. This is where things can get tricky when looking at it from the outside. It's likely, IMO, that as long as 3rd party game sales are not impacted, a reduction in console hardware sales would be seen as either acceptable or good. Why good? Because then the cost of sales for exclusive games becomes better. You then need to invest less money in order to attain the same or greater level of revenue and profits.

Currently, you have 2 types of console game buyers.

• Console gamers who buy all their 3rd party games on console.
• Exclusives gamers who only buy exclusives on console but do not buy 3rd party games on console.

In both cases the cost of sales has at the base, the cost of the console hardware (there's more to the cost of sales than just that, but at its base the console hardware is required). The question as to whether reduced console hardware sales, assuming that 3rd party console games sales were unaffected, would be good is whether that reduction in hardware console sales is worth the loss of 30% licensing on X number of 1st party games in order to sell the games on PC storefronts?

Further complicating that value assessment is the fact that you'll also sell more of your exclusive games if you also sell them on PC. So the question then becomes does reducing your investment in hardware, again assuming little to no impact on 3rd party game sales, and increasing the amount of 1st party game sales lead to revenue and profits that far exceed any loss from X percentage of 1st party game sales losing 30% revenue due to being sold through a 3rd party PC storefront?

That last bit is likely the key, IMO, to determining how Sony proceed WRT PC releases of their 1st party games.

Microsoft also had to go through this process, and their conslusion is obvious. After a few years of staggered releases of a limited number of Xbox exclusives being released on PC (similar to what Sony is currently doing), they've committed to day and date releases on PC. Basically, PC sales likely have little to no impact on 3rd party game sales on Xbox or the increase in revenue and profits far outweighed any loss in 3rd party game sales revenue on console, so there's no reason not to release 1st party games day and date on PC.

This is not to say that Sony will arrive at the same conclusion. It's entirely possible that they'll conclude that the best way forward is to stick with staggered releases. They may even conclude that not all 1st party games should be released on PC. But the possibility also exists that they'll follow in Microsoft's footsteps.

Basically the path that MS went down isn't necessarily the path that Sony will go down. But the market dynamics that prompted MS to go down this path are similar to what Sony are now experiencing. MS were first to investigate this due to a lower console install base, but the dynamics and potential benefits are similar between the two. Similar, but not the same, so they may not end up with the same conclusion.

Regards,
SB

 

[BRiT]

We know R&C performs the same even with lesser speed drives such as the 3200 mb/s DRAM-less Western Digital SN750 SE. So that's likely a more reasonable performance mark to use.

https://www.eurogamer.net/articles/...st-and-best-nvme-ssds-tested-on-playstation-5

Of course, the first order of business was to transfer Insomniac's storage-heavy Ratchet and Clank: Rift Apart to the drive and to play it, where I found that the in-game experience was essentially identical to playing the game on both the internal SSD and the much faster SN850. The game has split-second performance dips when moving between the game's signature dimensional portals, and these stutters seem to vary ever-so-slightly in severity from one run to the next, but the bottom line is that across a number of portals, the SN750 SE, SN850 or the internal drive could a touch faster or a touch slower.

 

[Shifty Geezer]

With 33 GB total data, there can't be that much changing from scene to scene. How many biomes/levels are there? Seems there's 9 planets. Ergo ~4 GB of data per biome, discounting cutscene and game data; realistically more like 3 GB per planet. If you can arbitrarily leap from one to another, you can't precache, but if you know what levels are available coming next, 3 GB should be enough to cache the data for the next planet in RAM ahead of time. It also mean 3 GB/s would be enough to load the whole planet in just one second, which fits in comfortably with BRiT's point above.

I don't think R&C is really pushing the storage solution and is beyond a tasty PC port with a minimum SSD target spec. There's probably not value in trying to engineer the game down to slower solutions as users are far more likely to have an SSD than 32+ GBs of RAM! Storage shouldn't be a bottleneck until you get massive datasets being read JIT- 200-300 GB games of either incredible detail or in-game teleporting. Creating content like that seems infeasible, but how about a game like NMS where the content is generated and baked in game, building the universe as you play it?

 

[iroboto]

but how about a game like NMS where the content is generated and baked in game, building the universe as you play it?

I think if the goal is to generate content and baked in game, you'd have to write back to the hard drive then as much as you are reading from it?

 

[Deleted member 11852]

Does it not? According to sales data PS4-P sold what?: approximately, 20% as many units as PS4? Someone will surely correct me if I got the number wrong.

No that's right, but what does that have to do with your claim:

It's far more likely for a console owner to get a competitor's console than it is for a console owner to get a PC. Not only are consoles similar in cost (thus will fit in their budget), but they also more easily slot into the same gaming environment (the living room).​

Sony's statement only says Sony are aware that mid-gen, some PlayStation owners are migrating to PC. How does Sony's statement verify that it's more likely for a console owner to get another brand of console? What am I missing?

Currently, you have 2 types of console game buyers.

• Console gamers who buy all their 3rd party games on console.
• Exclusives gamers who only buy exclusives on console but do not buy 3rd party games on console.

I don't fit into these two categories. I own a gaming PC and all three consoles. I typically make my game buying decision based on the game. I like some games to be portable and I'm happy to give up framerate and resolution if Switch is an option. Most non-strategy games I would prefer to buy on a console, but not always. For some games I would prefer PC because of mods, particularly Bethesda games, which do have mods on consoles but they are rather limited compared to what you can do on PC. And these are the types of games I will literally spend 100s of hours in.

We know R&C performs the same even with lesser speed drives such as the 3200 mb/s DRAM-less Western Digital SN750 SE. So that's likely a more reasonable performance mark to use.

Sure, but I assumed the discussion was centred on - whatever R&C needs on PS5 - there is a deficient on the PC.

Slower drives are not causing problems with the PS5 for R&C but PS5 is still benefiting from being able to load potentially very highly-compressed assets into a single RAM-pool, accessible immediately by both CPU and GPU.

 

[Deleted member 11852]

With 33 GB total data, there can't be that much changing from scene to scene. How many biomes/levels are there? Seems there's 9 planets. Ergo ~4 GB of data per biome, discounting cutscene and game data; realistically more like 3 GB per planet.

That would be the conventional game world and asset approach.
Is R&C doing that or is it doing something different? ¯\_(ツ)_/¯