• 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.

Ex-Sony Exec Says PC Ports of Old PS5 Games Are Almost Like Printing Money

Durin

Member
PS5 console can still run games decently, no stuttering
Eh, the stutter struggle is definitely a realer thing on PC, but it is does appear in some console titles too (Jedi Survivor had them on PS5, and performance mode didn't keep 60fps). It's also going to be a solved problem in future as engine makers (Epic is directly speaking to Unreal stutter issues on their official blog now) , dev knowledge sharing, and diminished returns in graphics will push for fixing the issue so more people buy and enjoy the big-budget games vs. leaving bad reviews.

It is funny to watch fanboys get mad more games are available to more people to play, and spin that as a bad thing. If your box has no other value besides artificially walling off games, it's not a particularly good device. I've heard console-centric users give other reasons, so Sony will either be fine, or eventually all the cross-pollination in gaming will make future "consoles" be a form factor computer that gives the same experience.
 

ProtoByte

Weeb Underling
Not just that, there are hidden costs with doing PC versions that don't get talked about, such as:

1: Longer development times, meaning console players have to wait longer to get games at all

2: Less games, when you consider there are various 1P titles which were developed due to GAAS (which was pushed due to PC) that have since been cancelled. Regardless of the reason, those are cancelled games that PS5 owners can't play now, but were also games designed for PC with Day 1 in mind.

Arguably speaking, if those were traditional/non-GAAS titles, they would've been designed specifically with the PS console audience in mind and had a better shot at actually releasing, rather than getting cancelled.

3: Increased costs. The amounts listed for Nixxes or whoever to finish a port are not entirely accurate to determine the cost for bringing out a PC version. Some people do not consider that if a console game gets a PC version a year or two years after release, if it's a major AAA game...then part of the testing, scope etc. has to be designed with PC in mind from the get-go. The development pipeline absolutely has to be scoped out for handling parallel PC development, even studios like ND stated they retooled their pipeline to implement expediated development for PC versions of games.

Those are all costs that factor into the initial budgets which most assume are only for the PS5 version of the game; truth is those costs are attributed to PS5 because that's where the game releases first. It is not possible to build a native PC version of a modern AAA single-player epic in 1-2 years, and yes while some aspects of game development are done on the PC platform, that does not necessarily mean they're being done on Windows, it sure as heck doesn't mean they (1P studios in this case) are running Windows x86 binaries with full DX12U support from the jump, and it doesn't mean they aren't using console-based SDKs for running compiled builds or debugging.

I put that last part in because yurinka yurinka yurinka yurinka in particular loves to harp on the "games are made on PC anyway" thing but never seems to understand that it doesn't outright mean console-exclusive games don't have Windows binaries/x86-executables and DX12U compliant versions running optimized Windows code with Steam support years ahead of any PC commercial release of the game (if ever).
Yep. Don't know what's so hard to understand. It's been said in the Insomniac leaks, and ND said it publicly. This shit is having a knock on effect, and PS gamers are beta-testing and subsidizing for Steam.

And even then, with just a handful of port projects and over 18 months, they can't put out a good enough SM2 port.

On top of that, and it really, really bears repeating: Xbox ALREADY DID THIS. It will not work for them as they put stuff on PS or Switch either. It will only accelerate the decline.

Btw, I was reading the IconEra thread on this. Entertaining and good work being done over there. Lol
 
Last edited:

ProtoByte

Weeb Underling
Oh, and of course Moriarty drew this out of him. That guy is smart, but has ridiculous blind spots. The notion that this isn't harming PS is pure cope.

To echo something said elsewhere
"If they won't come to you, go get them" is a fucking stupid idea. It's just another way of saying This (see: everything) is an Xbox
 

Gp1

Member
Gran Turismo 7, take my money Sony.

I don't want something from 2017 with perfectly good graphics remastered.

Ps. Let's say that Nintendo Switch exclusivity means nothing for PC gamers. 5 minutes of Googlefu and adios exclusivity.
 
Last edited:

Zathalus

Member
Of course porting costs are low, all the development was done on the game already for the PS5. God of War Ragnarok was ported by a grand total of 4 developers. Sony 1st party developers don’t really adjust anything due to this. PS5 1st party games are still heavily optimised for that console, as you can see from all the technical quirks the PC ports have, such as the reliance on PCIE bandwidth for good performance.

There is also no apparent sales impact on the console itself, despite it being almost 5 years of ports already. PS5 sales are tracking almost equal to the PS4 despite zero prices drops and the massive shortage that occurred due to the chip shortage a few years back. MAU is at an all time high, and revenue has blown past the PS4 generation. If this is what the supposed destruction of their ecosystem looks like, then Sony is probably doing something right.
 

Mibu no ookami

Demoted Member® Pro™
If SIE's AAA model was so much safer by incorporating PC sales, then why have numerous AAA GAAS that were in larger part developed due to wanting Day 1 PC...cancelled? We're not talking about no-name games, either: TLOU2 Factions, Spiderman GAAS, GOW GAAS etc. This isn't even touching on the reasons why those games were cancelled; I know the reasons are different for each one, I know in some cases the studios decided they wanted to do more non-GAAS and couldn't do both. But it doesn't change the fact that multiple AAA games from them have in fact been cancelled, and they were all games pushing for Day 1 PC.

You're conflating AAA and GaaS now? You're talking about completely different models. How far into production was the Spider-Man GaaS? You have no idea. You only know it existed because it was part of a leak, but even the rest of the leak suggests that it wasn't even in production. The fact that you realize the studios decided they didn't want to make these games should allow you realize that PC wasn't a factor in them not making them.

We can even touch on some of the XDEV stuff if we want; prioritizing PC Day 1 didn't do much for the Until Dawn remake, now did it? While it's not a AAA, Day 1 for LEGO Horizon didn't help that game perform any better at retail. And I don't even need to touch on Concord considering it's failure has been discussed to death.

You've got yourself a major comprehension issue. Would Until Dawn have sold better had it not prioritized PC Day 1? Make that make sense...

You don't seem to understand that the cost to port to PC versus the returns is what makes porting to PC make sense. It doesn't mean a game that won't sell will all of sudden be a success.

I would also touch on your misconception with port costs but I already did that when replying to another person.

I'm not even going to bother looking it up...


MAU is increasing because they're selling more PS5s, they still have a very active Helldivers 2 community on Steam, and they still have PS4s out there. Keep in mind, MAU is a measurement of Monthly Active Users; the way in which it's measured doesn't necessarily mean it's concurrent. It could be cumulative, it could be tracking for a minimum period of 10 hours or 10 minutes. No way to know without clarification, and that goes for ALL companies using metrics like MAU by the way.

MAUs can only increase if they're selling PS5s to new users if the PS4 users are still there. Whether it is 10 hours or 10 minutes means that someone is actively using their PS4/5. That they use it at least once monthly suggests that they're active users. You're hopeless.

You're also conflating MAU with units sold; MAU increasing doesn't mean the rate of systems being sold is 1:1. You can have multiple user profiles on a single PS5 and each of those could count as a MAU for all we know (the same is very likely the case on platforms like Xbox; even with Steam a single person could have multiple profiles and if they log into each one in a given month each would get counted as 1 MAU).

You can have multiple user profiles. So tell me what percentage of owners have multiple profiles. You're talking about probably less than 1% but you'll walk off the ledge to make your bullshit argument.

Also worth mentioning, is there is 0% chance PS5 sells 66 million systems between now and 2028. FY 2025 (ending March 30) and FY 2026 (ending March 30, 2027) will probably be the peak years for PS5 sales, and yes a lot of that is going be due to GTA6. But it also assumes that GTA6's PC/Steam port is at least a year after the console versions. If it comes earlier, say mid-2026, and also the next GTA Online is out around then, well that is probably going to deflate a decent portion of console sales, which IMO negatively impacts the chances of PS's FY 2026 matching FY 2025 in volume of unit sales.

How many times do you have to be wrong about things before you shut up? You already moved the goalpost that MAU won't fall until PS6. It's clownish at this point.

MAU can of course keep increasing as you think it will, because MAU is a separate metric not necessarily tied to consoles being sold. If more PS4 owners buy PS5s, or more Xbox owners end up buying PS5s, or even if more PS4 owners use their systems in more active ways, all of that increases MAU. If Helldivers 2 saw an explosion in new players, that'd count towards MAU. The player communities for MLB The Show and GT7 will help increase MAU, etc.

It literally is tied to console sales, you'll get there one day.
 
I wholeheartedly disagree with this "GaaS is made for PC" notion.

Sony had their very own Fortnite (3rd person MP shooter with wacky skins/microtransactions) 5+ years before Fortnite was released and it was exclusive on PS3:

Now I'm not saying if they remaster UC3 MP it should be exclusive on PS4/5... it makes sense to capture the PC audience too. Why not? MTX revenue was already pretty good on PS3 (a "failed" console).

But this "Sony only knows how to make SP games" notion needs to stop. :) It's historical revisionism at best (and it's usually coming from people who never, ever played PS3 MP games).

ps: Imagine an alternative universe where UC3 MP became Sony's "Fortnite" and Epic Games never managed to capture the MP TPS market with Fortnite. Even the Epic Games Store wouldn't exist today.

Sony could have had literally billions of dollars to fund AAA SP games today. Too bad they left money on the table. They wouldn't need shitty, woke GaaS attempts like Concord.


I agree with you, but I think you're misinterpreting what I meant with that part. I wasn't trying to insinuate SIE only make single-player games or that's the only thing they know how to make. After all, stuff like GT exist and have existed since the PS1. They very clearly can make non-GAAS and GAAS titles successfully, we see the latter right now with GT7 and Helldivers 2.

My actual point with that part was to suggest that current SIE only greenlit or decided to push GAAS investment this gen primarily to justify expanding on PC and, to have leeway in games they could push Day 1 on PC (specifically, Steam). Basically, many of the GAAS they had cooking aside from GT7, Helldivers 2 etc. were funded so they could expand into PC support with Day 1 titles without looking like they were doing the Xbox thing of bringing non-GAAS to PC Day 1 instead.

I'd say SIE were very aware of the controversy that strategy was causing between MS and the diehard Xbox fanbase, and they didn't want to repeat that with the diehard PlayStation fanbase, which is quite bigger. So GAAS was probably seen as their way to have their cake and eat it, too. Plus, it was Western investors/shareholders gassing them up to pursue GAAS, and make GAAS-related acquisitions to rebuttal MS buying Zenimax & ABK.

Again, I wasn't trying to imply SIE only know how to make single-player games; I've been disputing that narrative for a good while now, plus I know from their own legacy that such claims are nowhere near true.
 
We have leaked internal budgets of the Sony ports, both Nixxes ones and from other porting teams. So they are totally accurate: oficially the whole project of porting these games costed that.

No it didn't. The total costs also include amounts the leaks wouldn't reflect because they would've been baked into the initial development of the game, i.e engine overhauls to support other SDKs like DX12U.

Again, you don't make a port for Spiderman 2 to Windows & Steam in 15 months unless you already set up the engine and pipeline for that port ahead of time...which would be baked into the costs of the original budget. This is just common sense.

We also saw that these SP games had a separate budget for the separate project that was the previous, original console version made by Insomniac.

No, you saw the costs for what was allotted to studios like Nixxes to bring the game to PC, but that would've included various assets and code already set with some degree of compatibility on that platform, baked into the costs amortized to the PS5 version.

They only cost around a handful millions, and produce dozens of millions, so they are very cheap and profitable. We know that even Sackboy adventures port got profitable.

It only costs teams like Nixxes a handful of million to finish up work on the port and get it ready for Windows & Steam; other actual costs (i.e developing various textures and assets to future-proof for eventual PC release with various scaling options) would have been done while working on the PS5 version.

There is absolutely nothing leading to think this is the case. They said that the porters had to retool their engine for their TLOUP1 engine to adapt it for a PC release. That's all.

ND themselves said they retooled their engine pipeline to better accommodate PC. They've said that. It's on record.

Console games, even the console exclusive ones, always have been developed, tested and debugged in computers since the 8 bit days.

So now you're saying 'computers' instead of PC? Because you do realize a PC is an IBM-compatible, right? A computer can be anything. The Amiga was a computer. The 68K was a computer. SHARC systems are computers. SPARC systems are computers. The Atari Falcon is a computer. NONE of those are PCs as in being IBM-compatibles with x86/x86-64 or Windows at their core, though.

And even then, you are grossly exaggerating the role computers played in development of retro games especially. Most of the standout games for older systems, the ones that really pushed the hardware, and even most of the mid games, used devkits based on that hardware and linkers with various types of systems (some PCs, some microcomputers, all using various processors i.e x86, 68K, 6501, ARM, PowerPC, MIPS etc.).

So yes in a sense a computer would be utilized, but you are using "computer" as a stand-in for IBM-compatible x86/x86-64 Windows PC, which especially if we're going back to 8-bit/16-bit etc. eras, was NOT the main computer type studios used for testing and debugging parts of their games, or emulating hardware revisions in SDKs.

Like, c'mon now.

Even for Game Boy or 8 bit home consoles, devs had emulators on their computers to don't need to test everything on real console hardware because to do it in the dev's PC (specially back then) is faster and cheaper.

PCs, or computers? Why are you using them interchangeably, when PCs historically meant IBM-compatibles running DOS/Windows on x86 processors?

In current days, modern engines obviously allow to test the games in the PCs devs have. Even VR only games can be tested on the PC without using your PC monitor instead of the headset.

Computers are used in parts of the development pipeline. I never said otherwise. I'm debunking your idea that ALL development is done on computers, especially PCs, especially since you're probably insinuating Windows-based systems.

That part is false. If such were the case, we wouldn't need specific hardware SDKs like what the PS5 itself has.

Nowadays you don't even need to export a build, engines allow you to test the game inside the same engine. And to export a PC build is as easy in all engines as to select in the menu something like File>Export>and select PC/Windows as platform out of the platform lists configured.

As an example, this is how it looks the list of platforms to export in public engines like Unity, UE or Godot:

export-android.png

9a751ffede6e9e6d11cf8678a04cf24a7f9b13b9.jpeg

how-do-i-compile-a-release-build-v0-ncp48q28ab4c1.png

I think you're letting your anecdotal experience working on PCs conflate your understanding of the actual dev scene, especially for AAA games, and especially when going further back in older generations. It also kind of has a recency bias on your part.

Ports only need a few porters during a handful months when the engine already was used to commercially release PC games. In case of the first (commercial) PC port in that engine, they may need up to one or two years more to adapt the engine to export full featured PC builds with commercial level quality. A work that pretty likely most of it won't be needed to be done for upcoming PC ports of future games that use that engine. Next ones should only require only a few months.

Right. The engine already used commercially on PC = part of the initial version's development was spend in adjusting the engine to support a future PC port. That's a cost attributable to PC which is not going to be reflected in the cost of the port coming after the initial game's release!

In the later stages of development, console specific games will need to be tested and debugged with devkits/testkits to see there how console specific things (performance, trophies/friends/other platform specific things, certain console specific visual/tech specific things etc) are workeing and if they need to be fixed and tested. And well, even in engines very prepared for multiplatform development like UE or Unity there are always bugs that only appear when running the game on a specific hardware even in theory shouldn't.

Why are you bringing up UE or Unreal? When I said that not all game dev takes place on a PC, I obviously would've by association suggested that console-specific engines would've been included in that statement.

I'm not talking about engines like Unreal or Unity because I know chances are that's going to be a console/PC multiplat from the jump. But even in those cases, you will have a lot of games that still heavily utilize the console-based devkit hardware, and not just at the "tail-end" of the development like you're suggesting. The reason why is because various parts of the game logic need to be tested early on to make sure there is performance on the target hardware (the console) to accommodate those engine features.

You're treating it like a waterfall process; I've done database design in the past and even from that limited time I did, you learn that the waterfall method is inefficient. So with game development, a lot of studios working on console are CONSTANTLY going back-and-forth between working on PC (or computer) hardware, and working on the game via console devkits, which in most cases might be hooked up to a computer (not necessarily Windows-based or even x86-based i.e some might be running ARM) via ethernet or some PCIe linker cable, running SDK packages provided by the platform holder.

That's what a lot of the modern game development scene looks like, especially for non-Microsoft platforms (MS unified their PC and Xbox SDKs through the GDK with DX12U, so the process you've been describing is likely only mostly true for Xbox game development. Even then, there are likely quite a few exceptions).
 
You're conflating AAA and GaaS now? You're talking about completely different models. How far into production was the Spider-Man GaaS? You have no idea. You only know it existed because it was part of a leak, but even the rest of the leak suggests that it wasn't even in production. The fact that you realize the studios decided they didn't want to make these games should allow you realize that PC wasn't a factor in them not making them.

You do realize that there are AAA GAAS and AA GAAS, right? Or do you think Foamstars is a AAA title? What about Fall Guys? Didn't think so.

Yes we don't know how far the Spiderman GAAS was in development. But there was enough for some leaked concept footage to come out.

PC was a factor for SIE to greenlight 12 GAAS, that was my point. Will they ever come out and say it? No. But the fact all the GAAS were targeting simultaneous PC releases, shows that they saw GAAS as a game type to justify a push into the PC space. If the games have been getting cancelled, then among other reasons I'd suggest SIE rethinking their approach to PC is among them.

You've got yourself a major comprehension issue. Would Until Dawn have sold better had it not prioritized PC Day 1? Make that make sense...

Potentially. The game was a technical disaster at launch on PC, which contributed to low sales on Steam. Thing is, that talk might've also made some prospective PS5 owners hold off on buying the game, perhaps mistakenly thinking the PS5 version had similar performance issues. Even if that wasn't true, the stink of the PC version risked bleeding over into perceptions for the PS5 iteration.

Whereas if the game were PS5-only, maybe more features and bonus content could've been put into the PS5 version, or maybe a PS4 version could've been offered in place of PC with a bit of new features & content, and a sensible upgrade path for PS4/PS5 owners who already owned the original game. That could've likely made up for lack of PC or even been a better combo for sales long-term.

You don't seem to understand that the cost to port to PC versus the returns is what makes porting to PC make sense. It doesn't mean a game that won't sell will all of sudden be a success.

If the returns from PC are coming at the expense of returns off the console version (due to a growing number of console owners who have both platforms deciding to wait for the PC version), then the PC profits basically get negated and then some.

Has that been happening? As of right now, most likely not. Is it a 0% probability? Definitely not. Considering them putting HFW on PS+ hurt B2P sales of the game on PS4/PS5, there is no reason to try claiming SIE putting games on Steam 1 or so years after console doesn't have some negative impact on console sales of that game. Even if the impact is less (compared to the HFW/PS+ stuff), it would still be there.

I'm not even going to bother looking it up...

Too bad for you :/

MAUs can only increase if they're selling PS5s to new users if the PS4 users are still there. Whether it is 10 hours or 10 minutes means that someone is actively using their PS4/5. That they use it at least once monthly suggests that they're active users. You're hopeless.

You do know people can use PS4s and PS5s simultaneously, right?

Besides that, you're just proving my other point: MAU as a metric is very loose and malleable, by design. And that's not limited to SIE; it goes for ALL publishers & platforms that use MAU. It's a reason MAU never really impresses me.

You can have multiple user profiles. So tell me what percentage of owners have multiple profiles. You're talking about probably less than 1% but you'll walk off the ledge to make your bullshit argument.

No dude, it is definitely more than 1%. How much more exactly, I obviously can't provide that data. It's not published; if anyone knows it would be Sony/SIE.

But I can tell you anyone with a PS5 or PS4 who also has a family or siblings...likely has multiple profiles. If you're a FGC head, chances are you probably have multiple profiles. Either of those alone is easily more than 1%, should go without saying.

How many times do you have to be wrong about things before you shut up? You already moved the goalpost that MAU won't fall until PS6. It's clownish at this point.

How many times do you have to avoid the topic to take cheap pot-shots because you're getting outdebated by someone on a gaming forum?

Goalposts were never moved, simply clarified.

It literally is tied to console sales, you'll get there one day.

It's not 1:1 linear association. That was my point, not that console sales didn't factor in.
 
Last edited:

yurinka

Member
No it didn't. The total costs also include amounts the leaks wouldn't reflect because they would've been baked into the initial development of the game, i.e engine overhauls to support other SDKs like DX12U.
No, as I said, as seen in the leaked documents the original PS5 game and the port are separate projects with separate budgets.

The engine overhaul, if needed, is done in the porting project, because it's part of the job of a port.

Even Sony said multiple times that they had external teams for the PC ports to make sure the lead dev teams like Insomniac could focus only on PS and not worry about PC, and also the (Nixxes) porting team talking about their job with Spider-Man.

Again, you don't make a port for Spiderman 2 to Windows & Steam in 15 months unless you already set up the engine and pipeline for that port ahead of time...which would be baked into the costs of the original budget. This is just common sense.
If TLOU was released 10 years later on PC wasn't because the port took 10 years to be made. It's because Sony strategically thought that the point to release it was then.

The engine -most of it- was already ported by Nixxes in their previous Spider-Man and Ratchet ports. And as I said, even in the console games that never get released in PC, they have internal a rough PC version to test and debug the games there specially in the early stages.

That don't have all the scalability, all the GPUs/controllers/etc. support, recent DX12 features or upscaling stuff, several PC or console specific things etc.: it's just what they need to test and debug the console game on the PCs of the devs only, they don't do anything to run it in all PCs as laters the porters will need to do.

No, you saw the costs for what was allotted to studios like Nixxes to bring the game to PC, but that would've included various assets and code already set with some degree of compatibility on that platform, baked into the costs amortized to the PS5 version.
We saw the budget of the PS version, which means all the money they spent to make that version in all the studios and Sony teams who worked on it.

And we saw other budget for the PC port, which means all the money they spent on the PC port in all the studios and Sony teams who worked on it. Which basically is Nixxes, a handful guys from Insomniac (pretty a person from tools/engine to solve questions or support, someone from UI to solve questions or support, and a producer to send them material and overview everything) and then the marketing/PR/CS stuff.

It only costs teams like Nixxes a handful of million to finish up work on the port and get it ready for Windows & Steam; other actual costs (i.e developing various textures and assets to future-proof for eventual PC release with various scaling options) would have been done while working on the PS5 version.
Insomniac makes the game for PS, which includes having an internal rough, basic PC version for test and debug purposes, but that misses most things required to be a commercial game, like performing decently on a consumer level PC.

Nixxes makes the port: prepares it for PCs without shared memory, for Nvidia and AMD modern renderers, makes it scalable adding different texture packs and extra visual settings, adapts things like AA or RT, adds support for many controllers and keyboard controls, adapts it to multiple resolutions and its UI, changes the PS OS specific things for the Steam specific things (trophies, etc), fixes whatever is needed and a gazillion things more.

ND themselves said they retooled their engine pipeline to better accommodate PC. They've said that. It's on record.
No, they didn't. And Iron Galaxy ported their games, not ND.

So now you're saying 'computers' instead of PC? Because you do realize a PC is an IBM-compatible, right? A computer can be anything. The Amiga was a computer. The 68K was a computer. SHARC systems are computers. SPARC systems are computers. The Atari Falcon is a computer. NONE of those are PCs as in being IBM-compatibles with x86/x86-64 or Windows at their core, though.
Because -at least some teams- in some cases didn't use IBM x86 PCs to make their 8 and 16 bit console games, they did use other computers. Like the Commodore Amiga you mentioned. I'm personal friend of some of them. One of them worked in GB, NES, MS, GG, SNES, GBC, GBA games, including the first console game ever made in our contry (a GB game).

As an example, back then for GB they had a GB emulator. They had to manually erase and flash the EEPROMs, which took time, each time they wanted to test it in a GB (at the beggining they didn't have official devkits/testkits, had to make everything by themselves).

Why are you using them interchangeably, when PCs historically meant IBM-compatibles running DOS/Windows on x86 processors?
PC means "personal computer" and today they are the same. But yes, back in the 70s-90s there were other computers that weren't 'PCs' (x86).

I'm debunking your idea that ALL development is done on computers, especially PCs, especially since you're probably insinuating Windows-based systems.
Many coders use Linux instead of Windows but yes, all games (console, PC, mobile) are made today using computers (or 'PCs' if you want to be specific, sometimes in Mac if you don't consider it a PC). Specially in the early stages of the development.

It's in a more advanced part of the development, closer to the end, when you need to implement the console specific stuff and then using your PC, you connect remotely to the devkit, which in theory it is mandated to be in a secure room as demand some console maker to compile and debug there the console build.

Which can be tested on a testkit, which is basically a modded retail console prepared to run unsigned games (in the past were loaded via CD-burned disc or via network, nowadays everybody uses the network). They are cheaper and are the ones that use the testers. In case of Xbox, you can mod a retail console bought in a normal store to turn it into a testkit.

That part is false. If such were the case, we wouldn't need specific hardware SDKs like what the PS5 itself has.
Hardware SDKs don't exist.

As its name says, SDK are Software Development Kits, a group of tools, apps, libraries etc. you use in your PC to make games for a platform.

As an example, recently I made a Mega Drive game for a gamejam on my laptop using the SGDK, a fan made, open source SDK to make Mega Drive games. It includes stuff to compile the code, adapt art or sound resources and an emulator with useful debugging stuff. The result is a ROM you can use in emulators, in an Everdrive or similar to play it in real consoles or, as a few weeks a friend did, burn it into EEPROM cartridges.


stage1.png

GdvW9quXcAACXaG

GjTGKgtX0AAoUqK

GjTGKgrXAAAtjTs


I think you're letting your anecdotal experience working on PCs conflate your understanding of the actual dev scene, especially for AAA games, and especially when going further back in older generations. It also kind of has a recency bias on your part.
I work as gamedev since June 2005. Initally cell phone games, later also PC downloadable games before Steam even was a thing. Later PC browser games and smartphone games. Got bought by a top AAA publisher that obviously also makes console and PC games, getting merged with an older consoles+PC games studio of that company that had in our city. Beyond devkits, testkits or prototypes of things like VR headsets, as an example I saw some top AAA game running in real time on a PC -which had can't remember if 24GB or 32GB of RAM like 10 years ago- a couple days after it was revealed in a E3 conference (it was of my company but I didn't work on it).

Later I left to do indie PC & console games (one of them for VR) supporting other teams or making our own stuff, while also being a gamedev speaker at universities or mentor in gamedev incubators, or museum curator for gaming history exhibitions or conference (mainly focused in national games, sometimes interviewing or hosting round tables to some of the pioneers of my country).

Obviously during all these years I made many friends or had coworkers who worked in many places. As an example, the guy who coorganized with me the 'private/petit-comité' gamejam where my team did the game posted above, is the same that burned the game. That guy worked as 2nd party for Sony in the PSP/Vita/PS3 times. In the gamejam we had people from an Activision studio, from a Blizzard team, IOI, CCP Games and more. There was a guy from a NA Sony studio who wasn't there because couldn't take some days off.

Right. The engine already used commercially on PC = part of the initial version's development was spend in adjusting the engine to support a future PC port. That's a cost attributable to PC which is not going to be reflected in the cost of the port coming after the initial game's release!
Pretty likely you'd need a NASA PC to run the game decently, and that builld pretty likely had AMD specific code not ready to run on Nvidia cards and things like that. Who knows how the code of games like these PS4 games was.

These internal PC debug builds normally are pretty basic buggy etc. Need a lot of work to port it to PC as a commercial game, this is why they hire specialist dedicated studios to do so, who may spend like a year working on it.

Why are you bringing up UE or Unreal?
To show as example things you may know. And well, UE is the engine of Days Gone, Returnal and Sackboy adventure.

When I said that not all game dev takes place on a PC, I obviously would've by association suggested that console-specific engines would've been included in that statement.

I'm not talking about engines like Unreal or Unity because I know chances are that's going to be a console/PC multiplat from the jump. But even in those cases, you will have a lot of games that still heavily utilize the console-based devkit hardware, and not just at the "tail-end" of the development like you're suggesting. The reason why is because various parts of the game logic need to be tested early on to make sure there is performance on the target hardware (the console) to accommodate those engine features.
Yes, some parts of the game need to be tested early on devkits, and they are. Particularly engine specific things, normally related to visual tech stuff. Some guys, particularly engine coders, use devkits for these things.

But most of the stuff, mosty the things that are done by the people who make the game itself and not the engine, normally don't need the devkit until pretty late.

It may be some teams who everybody may use it since the start, but it would be pretty unproductive. Specially back in the time when you had to burn builds in a CD to test it.

You're treating it like a waterfall process; I've done database design in the past and even from that limited time I did, you learn that the waterfall method is inefficient.
Yes, waterfall has been outdated for way over a decade.

So with game development, a lot of studios working on console are CONSTANTLY going back-and-forth between working on PC (or computer) hardware, and working on the game via console devkits, which in most cases might be hooked up to a computer (not necessarily Windows-based or even x86-based i.e some might be running ARM) via ethernet or some PCIe linker cable, running SDK packages provided by the platform holder.
Well, nowadays some people work on their PC from home and connects via VPN to the office and compile remotely in the devkit. And some people plays these debug builds via streaming.
 

yurinka

Member
Astro Bot might be a bigger success on Switch than on PS5.
Well, in Nintendo terms Astro's Playroom sold 75M copies on PS5 (just kidding).

The problem with Nintendo is that most 3rd party games sell like shit there. Few 3rd party Switch games sold more than let's say 5M copies.

I think Astro Bot will end selling over 5M copies on PS5, and that in Switch could be maybe sell a couple more. Which would be great for them, but I'd prefer to keep Astro console exclusive.
 

Mibu no ookami

Demoted Member® Pro™
In a sense, Playstation fans should thank the PC players for helping to sustain Sony's Playstation business.

You're trying to be funny, but I think you're absolutely correct. PC gamers buying PS games means these game are more sustainable on console. It's a win for PS gamers who want the games they want.

It's the same with the transmedia. If the transmedia is successful enough, Sony have no choice but to make new games even if the original studios and/or staff aren't around.
 

Nickolaidas

Member
He's %100 right, but the lede is also kinda buried.

He specifically mentioned China as the biggest reason Sony are doing this.

China has been a major reason of PC Gaming's rise, just look at any big game's CCU timeline, it always starts to rise around the time Chinese people get back home from school/work, and it starts going down at 12AM China time when they go to sleep.
Seems like the Japanese are all about mobile gaming (and I mean that literally - they want to game as they walk, take the subway, wait in the reception, etc.), while the Chinese are all about gaming at home, with the best hardware and the most hacking/modding capabilities.

I've seen a few Netflix Korean and Chinese shows and whenever they show a character who is a gamer, they're always gaming on PC, and almost always play a PvP online game.
 
Last edited:
I just wish Sony would remaster games that need it like Infamous for example.

Instead of the pointless Horizon, Days Gone, Last of Us, shovelware remasters.
 
Last edited:

Zathalus

Member
No, as I said, as seen in the leaked documents the original PS5 game and the port are separate projects with separate budgets.

The engine overhaul, if needed, is done in the porting project, because it's part of the job of a port.

Even Sony said multiple times that they had external teams for the PC ports to make sure the lead dev teams like Insomniac could focus only on PS and not worry about PC, and also the (Nixxes) porting team talking about their job with Spider-Man.
Also confirmed here:


Relevant section confirming the porting developers (all four of them) getting it fully working on DX12:

“Digital Foundry: So when did the project to start porting God of War Ragnarok begin?

Matt DeWald: Yeah, so we've been working on it for about 18 to 24 months now. The majority of it before Ragnarok shipped was just getting the engine up to snuff on DX12. Once Ragnarok shipped fully, that's when we were able to fully transition over and start the true porting process. Then when Valhalla finished, that's when we were able to pull in all that content as well and wrap it up here. So a little bit between 18 to 24 months.”
 

Mibu no ookami

Demoted Member® Pro™
Also confirmed here:


Relevant section confirming the porting developers (all four of them) getting it fully working on DX12:

“Digital Foundry: So when did the project to start porting God of War Ragnarok begin?

Matt DeWald: Yeah, so we've been working on it for about 18 to 24 months now. The majority of it before Ragnarok shipped was just getting the engine up to snuff on DX12. Once Ragnarok shipped fully, that's when we were able to fully transition over and start the true porting process. Then when Valhalla finished, that's when we were able to pull in all that content as well and wrap it up here. So a little bit between 18 to 24 months.”

thicc_girls_are_teh_best thicc_girls_are_teh_best is ALWAYS wrong... like dead wrong...
 
Top Bottom