• 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.

New Sony patent: "Spoofing CPUID For Backwards Compatibility" (filed by Mark Cerny)

Markio128

Member
Mark Cerny reminds me of Prof Brian Cox. I listen to both of them, utterly engaged, believing, just for a brief moment, that I know what the fuck they are talking about.
In reality….
Hot Shots Idiot GIF
 
  • LOL
Reactions: Isa

Knightime_X

Member
Whatever helps you sleep at night

You're simultaneously arguing that they didn't have enough time to remove your imagined used game blocking technology from games at launch but also had no reason to patch other things.

How about you bring back some proof? I'm sure there are plenty of articles around about how these unpatched games don't work on offline consoles to back you up?
So what was the point of the patent then?
Shits and giggles?
 

Knightime_X

Member
Patents are made for a lot of things that are never created

Still waiting to see your evidence that it was used on launch games
Still waiting on YOUR evidence that it never was.
There is more evidence that they had intentions than your defense stating it wasn't.
If this was a murder case it would soo not end well for sony.
 
Last edited:
I remember being able to play Mortal Kombat 3 (PS1 version) on my PS3 super slim. That was awesome. As far as why they were able to do so, I know PS2 had dedicated hardware to carry it out and I know that at least the first generation/phat PS3 had dedicated hardware to carry out PS1 and PS2 compatibility. I assume it was cheap enough to maintain it for later PS3 iterations for PS1 compatibility at least. I'm sure if their backwards compatibility solution had been less hardware dependent, they would have kept it for PS4.
Yea, when I first got my PS3 I figured the BC for PS1/PS2 had been removed from all consoles. When I realized that my PS3 could play PSX games, I immediately went on Amazon and got Chrono Cross and the FF 5&6 Anthology. Funny to look back cause those games were available still in brand new condition, would have been 2008 or so. Good luck trying that now.

And actually, the PS3 BC for PSX I know is different from PS2, cause Vampire Hunter D doesn't emulate correctly on a PS3 but it plays just fine on a PS2. Interesting....

On a completely random note, a few years back at an old job, we were collecting gifts for a family in need. One of the wish lists had Chrono Cross (playstation) on it. This was like 2018, lmfao, couldn't believe it. I actually gave my copy to the person and ordered them a PSX memory card. That dude was probably ecstatic that Christmas.
 

Saucy Papi

Member
Yea, when I first got my PS3 I figured the BC for PS1/PS2 had been removed from all consoles. When I realized that my PS3 could play PSX games, I immediately went on Amazon and got Chrono Cross and the FF 5&6 Anthology. Funny to look back cause those games were available still in brand new condition, would have been 2008 or so. Good luck trying that now.

And actually, the PS3 BC for PSX I know is different from PS2, cause Vampire Hunter D doesn't emulate correctly on a PS3 but it plays just fine on a PS2. Interesting....

On a completely random note, a few years back at an old job, we were collecting gifts for a family in need. One of the wish lists had Chrono Cross (playstation) on it. This was like 2018, lmfao, couldn't believe it. I actually gave my copy to the person and ordered them a PSX memory card. That dude was probably ecstatic that Christmas.
That's an awesome story! I'm sure you were able to make their day on Christmas 😁

I'm assuming you're referring to either the slim or super slim iterations of the PS3? Somebody quoted me and let me know that later iterations of the PS3 actually used software emulation for PS1 which would have definitely caused some issues. I'm sure hardware emulation worked perfectly fine on the original iteration of the phat model.
 
That's an awesome story! I'm sure you were able to make their day on Christmas 😁

I'm assuming you're referring to either the slim or super slim iterations of the PS3? Somebody quoted me and let me know that later iterations of the PS3 actually used software emulation for PS1 which would have definitely caused some issues. I'm sure hardware emulation worked perfectly fine on the original iteration of the phat model.
Mine is the fat version where they removed the PS2 BC and reduced the USB slots to 2, etc. I got mine in late 2008.
 

Saucy Papi

Member
Mine is the fat version where they removed the PS2 BC and reduced the USB slots to 2, etc. I got mine in late 2008.
Ah okay, I thought they had moved to software emulation for the PS2 GPU in fat 2.0 and later removed it altogether in the slim and super slim versions but I guess I was wrong. I guess they removed it well before then.
 

Yoboman

Member
Still waiting on YOUR evidence that it never was.
There is more evidence that they had intentions than your defense stating it wasn't.
If this was a murder case it would soo not end well for sony.
My evidence is that there was never an issue of PS4 launch games being unable to be played because of DRM. There were countless people on the look out for this at the last gen launch. You're making the claim in the first place, but keep going in circles without a shred of fact
 

Knightime_X

Member
My evidence is that there was never an issue of PS4 launch games being unable to be played because of DRM. There were countless people on the look out for this at the last gen launch. You're making the claim in the first place, but keep going in circles without a shred of fact
I never said the games were unplayable.
 
I know.

But if Microsoft has to do some extra for their BC program, this might be an obstacle for their plans.
So how do we know the same doesn't apply to Microsoft? Is there proof their patents didnt create obstacles for Sony creating backwards compatibility on their consoles? Or are we just going off blind faith believing that Sony wants user to pay to play games they already own and allow Microsoft to reap the benefits?
 

Saucy Papi

Member
Are these guys for real?

REALLY???

WE'VE BEEN DOING THAT IN THE HACKINTOSH COMMUNITY FOR YEARS!! (since hackintoshing became a thing in 2006)


And that's the problem: they DIDN'T design it in the first place. They just took it. They used what we've been doing in the hackintosh community for years to apply it for PS4 back-compat and then patented it, which is a really shitty move if you ask me.

Also, look up for the Clover Bootloader or the more recent OpenCore bootloader: both of them are infinetely more complex than the bullshit CPU spoofing that Cerny came up with, specially OpenCore, since it can feed the OS with spoofed CPU-ID, GPU-ID, modified USB pathways and connectors, etc (in short, spoof everything in software).

Heck, the computer I'm writing this post right now is my potato Lenovo m700 SFF with a Core i5-6400 and Radeon RX550, which I spoofed via DSDT file and config.plist file into an iMacPro1,1. That means the OS believes I have a Xeon CPU (which I don't), with custom thermals and a Radeon RX560 so it can have native metal and hardware DRM decoding.



What Sony did with that patent is acknowledging they didn't have a better solution to come up with in issuing PS4 back-compat, so they had to retort to open-source community-driven projects to get their bearings together.

Oh well...
How feasible do you think their current BC solution is long term? I ask because I know Microsoft created an abstraction layer for XB1 and forward consoles, so (theoretically) all they need to do is update HAL to maintain compatibility as long as they maintain an x86 instruction set and GPU compute. Sony on the other hand seems to be much more hardware dependent for their solution which makes it more difficult to maintain compatibility the farther you get from the original design. I say that with pure speculation since I have no insider information but I remember Cerny essentially talking about how the much higher IPC count of the Zen 2 CPU was causing issues with compatibility. My guess is that AMD implemented a hardware solution for the Jaguar cores.
 
Last edited:

Knightime_X

Member
You literally said they had to be patched to remove DRM

If DRM was there then every game from launch would be unplayable on reselling or putting in a second console. Why has nobody ever reported on it?
Or disable it when you first initialize the console then patch a firmware so the games never call for DRM to activate.

Wasnt internet connection required to initialize ps4 in 2013?
 

skybaby

Member
To facilitate backwards compatibility, a computing device may respond to a call from an application for information regarding a processor on the computing device by returning information regarding a different processor than the processor on the computing device

Isn't that quite literally patenting all kinds of emulation?
 

Knightime_X

Member
I think this is where there might be some confusion. Can you elaborate on what you meant by "needed an immediate patch before playing"?
I clearly remember every physical game that I had early on requiring a rather short download and the speed and duration of this download was consistent with every game.
Eventually I went fully digital and stopped paying attention so I don't know when this ended.
 

MarkMe2525

Member
The cell also used PPC cores with In Order Execution, versus the PS5's x86-64 cores with Out Of Order Execution.

Overcoming those hardware architectural differences with such a small difference in clock speed makes direct emulation virtually impossible.

Even JiT recompilation may be just about doable but will struggle practically in some cases on some workloads.
Correct me if I'm wrong but Xbox one emulated Xbox 360's 3 3.2ghz in order ppc cores with a tablet CPU with a much lower clockspeed. Sony could figure something out if it was a priority to them.

Maybe this is what you meant when you said "direct emulation"?
 
Last edited:

Fafalada

Fafracer forever
I'm assuming you're referring to either the slim or super slim iterations of the PS3? Somebody quoted me and let me know that later iterations of the PS3 actually used software emulation for PS1 which would have definitely caused some issues. I'm sure hardware emulation worked perfectly fine on the original iteration of the phat model.
On second thought - I was referencing evolution of PS3->PS2 BC (Launch had CPU+GPU, Europe Launch removed the CPU & 32MB of RDRam, 40GB launch removed the GPU).
I'm about 95% sure as this was 15 years ago - but on further thought I don't think PS3 ever had any PS1 hardware - basically it was fully software BC throughout.

PS2 evolution was: CPU in hardware, GPU software at launch. CPU in software ~4 years later (revision 7xxx onwards) - when I/O complex was replaced by a new powerPC chip.
 

Danknugz

Member
Yeah, they even had a patent to block secondhand games shortly before PS4 release:
wow, i never heard of this. am i reading this wrong or did they try to patent something f that would prevent people from shari g owned games? how is this any different from what MS tried with xbox one and sony literally made fun of them for it? that was on the prior gen for ps3, and this is for ps4?
 

Saucy Papi

Member
On second thought - I was referencing evolution of PS3->PS2 BC (Launch had CPU+GPU, Europe Launch removed the CPU & 32MB of RDRam, 40GB launch removed the GPU).
I'm about 95% sure as this was 15 years ago - but on further thought I don't think PS3 ever had any PS1 hardware - basically it was fully software BC throughout.

PS2 evolution was: CPU in hardware, GPU software at launch. CPU in software ~4 years later (revision 7xxx onwards) - when I/O complex was replaced by a new powerPC chip.
Gotcha, so it never included additional hardware for PS1 BC, just PS2. I thought the original phat model included both hardware emulation for both the CPU and GPU for PS2 and then moved to software emulation for the GPU while maintaining hardware emulation for the CPU in the second iteration of the phat model but I guess I was wrong. I'm surprised they accomplished CPU software emulation at all. I guess that's why I would see people complaining about BC performance in later iterations.
 

Dr Bass

Member
I have a feeling that Sony just patents overzealously even for stuff with prior art. The issue with the system is that the real test of patents comes down to law suits or private and commercial agreement between companies.
It's not a Sony thing. All tech companies on the planet do this, and yeah it's completely messed up.
 

JackMcGunns

Member
Yeah, they even had a patent to block secondhand games shortly before PS4 release:


I heard about this in 2013 with the whole used games fiasco, but Sony waited for Microsoft to balk and walked out of there like nothing happened lmao.

 

DrAspirino

Banned
How feasible do you think their current BC solution is long term? I ask because I know Microsoft created an abstraction layer for XB1 and forward consoles, so (theoretically) all they need to do is update HAL to maintain compatibility as long as they maintain an x86 instruction set and GPU compute. Sony on the other hand seems to be much more hardware dependent for their solution which makes it more difficult to maintain compatibility the farther you get from the original design. I say that with pure speculation since I have no insider information but I remember Cerny essentially talking about how the much higher IPC count of the Zen 2 CPU was causing issues with compatibility. My guess is that AMD implemented a hardware solution for the Jaguar cores.
You are correct.

Sony's solution on BC is, as I said before, low-effort / low-gain, which means it isn't feasible in the long term. Also, Sony APIs are much more hardware-dependant than Vulkan or DirectX, and their entire OS philosophy and system structure is built around a VERY specific hardware, rather than having room for growth. It's as Sony designs their consoles as the last ones they're ever going to make, without thinking that maybe they'll sell well enough to make a PS6 or PS7 and people will be demanding back-compat in those as well.

If you want to know wether Sony's approach has a future, just look at the hackintosh scene: the moment Apple said they were ditching x86 to go with their own CPUs, was the moment we knew we were screwed, since there's only so much you can do with hardware-spoofing.
 

Saucy Papi

Member
You are correct.

Sony's solution on BC is, as I said before, low-effort / low-gain, which means it isn't feasible in the long term. Also, Sony APIs are much more hardware-dependant than Vulkan or DirectX, and their entire OS philosophy and system structure is built around a VERY specific hardware, rather than having room for growth. It's as Sony designs their consoles as the last ones they're ever going to make, without thinking that maybe they'll sell well enough to make a PS6 or PS7 and people will be demanding back-compat in those as well.

If you want to know wether Sony's approach has a future, just look at the hackintosh scene: the moment Apple said they were ditching x86 to go with their own CPUs, was the moment we knew we were screwed, since there's only so much you can do with hardware-spoofing.
Cool, thanks for the response! That's exactly why I sold my MacBook Pro once I started hearing rumors from reputable sources that they were going to dump x86. I was like "well, wait a minute. So that means I won't be able to use Boot Camp to boot into Windows anymore? Screw that!". Gaming support is pretty crappy on macOS when compared to Windows.

As far as PlayStation legacy support goes, this is exactly why I moved away from PS into PC this generation. I guess it should have been a tip off when the PS4 Pro had nothing more than a bifurcated GPU but Cerny talking about how the higher IPC was causing compatibility issues on PS5 really drove it home for me.
 
This would explain why every ps4 game early on needed an immediate patch before playing.
To quietly remove the very thing they were making fun of xbox one doing.
I recall the first night with the PS4... No need for the patches.

Go work with the nyt or CNN if you want to make up stuff.
 

Fafalada

Fafracer forever
Also, Sony APIs are much more hardware-dependant than Vulkan or DirectX
Strictly speaking that's true - Gnm is basically custom vendor(AMD) API, much in the same vein Switch uses modified NVidia APIs. As far as compatibility goes - that tends to couple closer to a single vendor.
That being said - XBox APIs do (and have done so throughout) also expose direct access to features specific to their consoles. Some of the modern features this gen (like their implementation of SFS) will also be increasingly difficult to emulate/replicate if switching to a different vendor in the future.

BC is not something I'd take for granted if any of them chose to switch vendors in the future again. At minimum - it will again come at a substantially higher cost than this recent transition.
 
Last edited:

Fafalada

Fafracer forever
I thought the original phat model included both hardware emulation for both the CPU and GPU for PS2 and then moved to software emulation for the GPU while maintaining hardware emulation for the CPU in the second iteration of the phat model but I guess I was wrong. I'm surprised they accomplished CPU software emulation at all. I guess that's why I would see people complaining about BC performance in later iterations.
Yea it was reverse - GPU was the last thing emulated (and the hardest). Performance was mixed IME - some emulated titles actually ran better, others worse, that persisted into when GPU was emulated too. A lot like og XBox on 360, or 360 on base X1 - software emulation at scale is a tricky beast.
 
I get that but the patent is proof of their intentions.
Had it succeeded with Xbox the future of gaming would look quite different today.

The patent is proof it was a consideration. Just as the final announcement was proof they decided against moving in that direction.

Also, MS's plan never would have succeeded. The market wholesale rejected it. MS bore the brunt because they were the only one at the end foolish enough to decide on going down that route.

But yeah, of course, the whole industry was pushing for it. It was clearly pressure applied from the publishers who wanted to curb the insane resale profits Gamestop were generating at the expense of new game sales.

Recompilation doesn't require 'software wrappers' anymore than anything else, it's just an optimization - raw interpreters are painfully slow. Even PS2 emulated PS1 with JIT for the CPU code.
Besides how 'direct' are we talking about here. Interpreting CPU/GPU etc. instructions is still just another form of virtualization, you should go to logic-gate level to truly 'emulate' the original hardware, and good luck doing that in software for anything less than 20 years old.

Thanks for the correction. I don't assume to be an expert in this area, but it's pretty clear you have direct expertise so I welcome your input.

It can impact emulation (and performance), as reordering FP instructions literally changes the outcome of computations, which can lead to extremely hard to track down bugs. Of course every CPU can enforce order of execution, but it comes at a (potentially severe) performance penalty.
Eg. European launch PS3 PS2 emulator (and onwards) employed speed-hacks where accuracy wasn't needed, but for some titles it ran in 'slow-mode' - like the one I mention above (both order of execution and some other quirky behaviors of PS2 FP that was rather liberal with IEEE compliancy). And that was when both architectures were in-order (the ISA changes alone can lead to differences in ordering of how math gets processed).

This is very interesting. Thanks for sharing.

The whole industry was onboard and the moment MS opened their mouth about it and they all saw the backlash, everyone scrapped it and hung MS out to dry. It was a huge error in judgment on the part of MS but anyone who thought this was their idea and they were the only ones doing it was out of their mind. Prior to this, EA and other pubs started requiring online passes in order to snatch up something from a second hard sale. This was likely the ultimate goal for the industry. Glad MS caved and everyone scrapped this terrible idea.

For sure, it was the publishers pushing for this. MS bent over backwards to do it at their own detriment ultimately. They paid the rightful price for their error in judgement.
 
Top Bottom