PS3 Project RSX Boost: Overclock your Retail PS3 RSX Speeds (ps3 cfw only)

[aldostools]

Can someone please point me to the right post, dev wiki link...or whatever so I can learn to make the LV1 needed to add specific clocks to the MFW builder?

I want a 450/800 and so on.

Check ps3mfw-builder-0.2.1-mod by @haxxxen
https://github.com/haxxxen/ps3mfw-builder-0.2.1-mod/blob/master/tasks/01_patch_rsx_oc.tcl

 

[Smugg]

*900MHz memory on 65nm is possible and completely stable on mine.
650 on core is accurate, even if some 20XX can do more (700) depending on the NEC Tokins health.

25XX models have 3 different RSX revisions. What would be interesting is to know exactly which revision we have in each model, associate them with dates and see how they respond to OC.
View attachment 43141

Also, if we put a EADP-220BB or APS-250 power supply in a 25XX model, I wonder if they would be able to handle a bit more (since they should bring more stability overall).
EADP 200DB / APS-270 (21XX and 25XX) : 12V 16A (192W)
EADP-220BB / APS-250 (20XX) : 12V 18A (216W)

I'm currently trying to find a 25XX slim btw.

my 2501a has a CXD5300DGB from August 2010. just starting the OC im at 600 core 825 memory, working up in 25mhz and testing. 700mhz on core is probably the highest i should go?

 

[Tanzu15]

I've updated the picture, "1-882-770-21" is there indeed.
"APS-270" or "EADP 200DB" PSUs are what you will find in every 21XX and 25XX slims across the world.
Those I speak of are only present in the 20XX models (the region doesn't matter) and can deliver more(/handle more) since they were designed to work with the demanding 65nm RSX.

See :
View attachment 43142 (I'm from France btw)

Would it be safe to put higher psu? And would it help at all with OC? I find nothing on forums and Reddit saying certain power supplies are better. I spent hours and no such info exists.

my 2501a has a CXD5300DGB from August 2010. just starting the OC im at 600 core 825 memory, working up in 25mhz and testing. 700mhz on core is probably the highest i should go?

Bro you can do 700/950 easy. Just skip to there. Then start adding 50 to the core and test.

 

[RIP-Felix]

Check ps3mfw-builder-0.2.1-mod by @haxxxen
https://github.com/haxxxen/ps3mfw-builder-0.2.1-mod/blob/master/tasks/01_patch_rsx_oc.tcl

Wow, sereiously? Just keep adding if statments...lol. Seems like there would be a more efficient procedural mechanism. I'm not complaining tho...I can work with it.

 

[Smugg]

This is where im at so far. Pretty nice increase in frames over base OC. Testing in the beginning of Diablo 3

 

[Tanzu15]

This is where im at so far. Pretty nice increase in frames over base OC. Testing in the beginning of Diablo 3

I can't see that image at all. Are you using 700/950?

 

[Smugg]

I can't see that image at all. Are you using 700/950?

i got to 700 / 850 before i quit for the night.

 

[Tanzu15]

i got to 700 / 850 before i quit for the night.

Oh ok. Well you should be able to do 950 memory easily.

 

[RIP-Felix]

Error log:

Executing command file copy {*}$args
******** Running task 01_patch_rsx_oc **********
Modifying CORE_OS file: lv1.self
Unpacking CORE_OS files.....
unpkg-ing file: CORE_OS_PACKAGE.pkg
Executing command unpkg $pkg $dest
Executing shell pkgtool.exe -debug yes -action decrypt -type pkg -in {C:\Users\Luan\AppData\Local\Temp\PS3MFW\PS3MFW-MFW\update_files\CORE_OS_PACKAGE.pkg} -out {C:\Users\Luan\AppData\Local\Temp\PS3MFW\PS3MFW-MFW\update_files\CORE_OS_PACKAGE.unpkg}

cosunpkg-ing file: content
Executing command  cosunpkg $pkg $dest
Executing shell cosunpkg {C:\Users\Luan\AppData\Local\Temp\PS3MFW\PS3MFW-MFW\update_files\CORE_OS_PACKAGE.unpkg\content} {C:\Users\Luan\AppData\Local\Temp\PS3MFW\PS3MFW-MFW\update_files\CORE_OS_PACKAGE}



CORE_OS unpacked
Modifying self/sprx lv1.self
Importing SELF-HDR info from file: lv1.self
Executing command set buffer [shellex ${::SCETOOL} -w $in]
Executing shellex scetool -w C:/Users/Luan/AppData/Local/Temp/PS3MFW/PS3MFW-MFW/update_files/CORE_OS_PACKAGE/lv1.self

!!ERROR!!:  FAILED TO IMPORT SELF-SCE HEADERS FROM FILE: lv1.self
FATAL ERROR: !!ERROR!!:  FAILED TO IMPORT SELF-SCE HEADERS FROM FILE: lv1.self
See C:/Users/Luan/AppData/Local/Temp/PS3MFW/ps3mfw.log for more info
Last lines of log :
*****************
unpacking prog.srvk...
unpacking pkg.srvk...
CORE_OS unpacked
Modifying self/sprx lv1.self
Importing SELF-HDR info from file: lv1.self
Executing command set buffer [shellex ${::SCETOOL} -w $in]
Executing shellex scetool -w C:/Users/Luan/AppData/Local/Temp/PS3MFW/PS3MFW-MFW/update_files/CORE_OS_PACKAGE/lv1.self

!!ERROR!!:  FAILED TO IMPORT SELF-SCE HEADERS FROM FILE: lv1.self
FATAL ERROR: !!ERROR!!:  FAILED TO IMPORT SELF-SCE HEADERS FROM FILE: lv1.self

*****************
Error running script: !!ERROR!!:  FAILED TO IMPORT SELF-SCE HEADERS FROM FILE: lv1.self

I support this it'll help a lot of people. I'm still out but as soon as I get home I will try it again with everything they advised. Also is it possible to set your own custom number(? Eg I no longer see 750/975 and it's the most stable for me but I keep getting another error.
FATAL ERROR: !!ERROR!!: failed to import self sce headers from file:lv1.self

I'm working on it !
I think I've managed to add 975MHz VRAM to the list, but I can't test it myself (on my PS3).
I can make you one if you want.

Ok, I was able to figure out the pattern and add the values I wanted to the combox and search and replace functions. But I get the same error. You guys had above.

When I attempted to build any of the ones previously added, they work fine. It's just the new ones I added. So what was the secret?

***EDIT: SOLVED***
NVM: I'm an idiot and forgot to generate the keys again after DL'ing the latest build from the repo. It works fine after.

 

[Tanzu15]

Ok, I was able to figure out the pattern and add the values I wanted to the combox and search and replace functions. But I get the same error. You guys had above.
View attachment 43157
When I attempted to build any of the ones previously added, they work fine. It's just the new ones I added. So what was the secret?

***EDIT: SOLVED***
NVM: I'm an idiot and forgot to generate the keys again after DL'ing the latest build from the repo. It works fine after.

What are you trying to create?

 

[Mitsu™]

Would it be safe to put higher psu? And would it help at all with OC? I find nothing on forums and Reddit saying certain power supplies are better. I spent hours and no such info exists.

Using a higher-rated PSU in a 25XX will be safe and can only be beneficial. It provides more stable power, which can helps with overclocking by reducing risks of crashes for exemple. It's a fact.

And I remembered this post too :

It's for this very reason that with my 2503A slim I switched out the APS 270 for the 250, it's simply more capable which has numerous advantages;

- more stable power delivery
- shouldn't get as hot
- the extra amps (18 vs 16) will just help the console stay more stable when being pushed hard
- as it now is the extra amperage is useful for overclocking. Especially if someone figures out the CPU which must be taking power from the 12v rail.

 

[Tanzu15]

Using a higher-rated PSU in a 25XX will be safe and can only be beneficial. It provides more stable power, which can helps with overclocking by reducing risks of crashes for exemple. It's a fact.

And I remembered this post too :

So which is the best power supply to use for 25xx to get the most out of an OC? Isn't Effiency % the most important? Or will it be the wattage!

 

[Mitsu™]

So which is the best power supply to use for 25xx to get the most out of an OC? Isn't Effiency % the most important? Or will it be the wattage!

Since both models (for 20XX slims) have the same wattage, higher efficiency is preferable. Efficiency indicates how well the power supply converts electricity from the wall into power the PS3 can use. Higher efficiency means less power is wasted as heat, resulting in more power available for the console. The APS-250 is better because it wastes less power as heat and delivers more stable power, which can help prevent crashes and other issues.

If you can't find the APS-250, the EADP-220BB is still a better option compared to the power supplies found in 25XX slims.

EDIT : My brand """new""" 2504A [Date Code : 0D / 2010 - October-November-December] is here !

 

[Tanzu15]

Since both models (for 20XX slims) have the same wattage, higher efficiency is preferable. Efficiency indicates how well the power supply converts electricity from the wall into power the PS3 can use. Higher efficiency means less power is wasted as heat, resulting in more power available for the console. The APS-250 is better because it wastes less power as heat and delivers more stable power, which can help prevent crashes and other issues.

If you can't find the APS-250, the EADP-220BB is still a better option compared to the power supplies found in 25XX slims.

EDIT : My brand """new""" 2504A [Date Code : 0D / 2010 - October-November-December] is here !
View attachment 43158

I have APS-270s and 200DBs. Oh let's see those Overclocks. Just use 750/950 right away. Try crysis. That slim is automatically gonna do 700/950 fully stable. 750 core is where the test begins. If it artifacts, it lost. And it can only do the 700 core. It kinda sucks to be outside the US. Here in the US we get exact dates when the PS3 was manufactured.

 

[Mitsu™]

I have APS-270s and 200DBs. Oh let's see those Overclocks. Just use 750/950 right away. Try crysis. That slim is automatically gonna do 700/950 fully stable. 750 core is where the test begins. If it artifacts, it lost. And it can only do the 700 core. It kinda sucks to be outside the US. Here in the US we get exact dates when the PS3 was manufactured.

I will clean it first (and repaste, etc...), it's super dusty ! x)
Yeah, date codes are kinda vague unfortunately.
It's annoying that they used so many different dating methods.
"Why make it simple when you can make it complicated ?", I guess.

 

[Tanzu15]

I will clean it first (and repaste, etc...), it's super dusty ! x)
Yeah, date codes are kinda vague unfortunately.
It's annoying that they used so many different dating methods.
"Why make it simple when you can make it complicated ?", I guess.

Yeah I don't know who was responsible for the dog sh!t dating codes. FFS American PS3s did it perfectly.

 

[RIP-Felix]

What are you trying to create?

Toying around with the idea of underclocking the Core, but OC'ing VRAM to make up the difference in FPS.

In our testing, @Nascar1243 found 450MHz core lowered temps 2C with only a -0.3FPS performance hit. I charted the performace curves and calculated VRAM at 800 should compensate.

OFC the type of workload will affect the performance. Shading, water simulation, lighting, polygons, and partical effects are core heavy tasks. Where as streaming data in/out of the frame buffer, speed of loading high resolution textures, assets, etc...is going to benifit from higher VRAM frequency. Since the core underclock is very minor, it may be unnoticable to the average user. And for -2C temp drop that's worth it on 90nm models.

That's the idea anyway.

 

[haxxxen]

450MHz core

interesting approach, but are there no problems? I have underclocked both 400/550 and got serious XMB slowdowns and tripple beep shutdowns. the only positive effect was the temps were equal and bit lower on both, CELL and RSX

 

[Tanzu15]

Toying around with the idea of underclocking the Core, but OC'ing VRAM to make up the difference in FPS.

In our testing, @Nascar1243 found 450MHz core lowered temps 2C with only a -0.3FPS performance hit. I charted the performace curves and calculated VRAM at 800 should compensate.

OFC the type of workload will affect the performance. Shading, water simulation, lighting, polygons, and partical effects are core heavy tasks. Where as streaming data in/out of the frame buffer, speed of loading high resolution textures, assets, etc...is going to benifit from higher VRAM frequency. Since the core underclock is very minor, it may be unnoticable to the average user. And for -2C temp drop that's worth it on 90nm models.

That's the idea anyway.

That sounds pretty interesting. Though underclocking for only a 2c drop? That's not too great. Though, I see how that can make 90nm last a bit longer I guess. Have you tried tinkering or learning more about the Cell Overclock? Cause that's where huge gains would come from. Surprised that's not talked about much.

interesting approach, but are there no problems? I have underclocked both 400/550 and got serious XMB slowdowns and tripple beep shutdowns. the only positive effect was the temps were equal and bit lower on both, CELL and RSX

Is there any hope for a cpu overclock? Cause man this would transform Games.

 

[RIP-Felix]

Besides drop in FPS none that we have noticed. There were some debug units that were clocked lower listed on the dev wiki. Some of them had VID binned silicon (more efficient) with higher clocking potential, but were ran under.

We havent gone lower than the 400 that was already in MFW builder, but now that we know how to generate the LV1 and add the entry to the combox, we could go lower and see where the underclocking limit is.

I nearly bricked my 2501a this weekend. There were ZERO warning signs as 950 VRAM and then it was completly frozen even in safe mode at 975. That's how quick it can go from stable to damn near full soft-brick, even with 25MHz increments.

I was able to recover it by using SYSCON UART to edit the fantable's 1st step fan% to 100% (0x33 to 0xff). After being off and cooled to room temp there was a brief time that it seems stable. Then artifacts begin appearing (pin pricks of light) and become more numerous as it heats up, and finally it froze at about 60% of the 1st part of the update, where it loads before restarting to install. After booting fan to 100% it kept it cool long enough to finish the update, but there were a few artifacts on screen making it hairy.

So PROCEED WITH EXTREEM CAUTION. I do have a teensy, doesnt mean I want to solder all them wires!

That sounds pretty interesting. Though underclocking for only a 2c drop? That's not too great. Though, I see how that can make 90nm last a bit longer I guess. Have you tried tinkering or learning more about the Cell Overclock? Cause that's where huge gains would come from. Surprised that's not talked about much.



Is there any hope for a cpu overclock? Cause man this would transform Games.

haha...What's funny is that people will risk killing their console delidding and using LM for about the same decrease in temps. But if it means loosing 0.3FPS they recoil. And all this requires is an update!

EDIT:
Consider. The PS3 has technically the most powerful hardware in the 7th gen, but because of poor CELL optimization, many crossplatform games ran better on 360. A RSX OC doesn't make those games play significantly better anyway, so what we're talking about here is reducing the temps on a defective 90nm model to help stave off the YLOD. If it barely comes with a hit to performance that's a net benifit. And if we ever cam figure out the CPU OC, that will direcly target the #1 issue with PS3. The CPU bottleneck.