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

pawelekd9 said:
External PSU, step-down converter 12V>5V inside.
Click to expand...
Your project is very cool, and the source being external helps a lot to lower its temperature, but just out of curiosity, you must have noticed this of course, but did you check the maximum current supplied by 12V and 5V?
using the APS 226 as a base, we have a maximum current of 32A on the 12V line and 3A on the 5V.

I was curious because, at least where I live, it is difficult to find a quality source that provides such a high current.
 
_XBrD4shDOE_808_ said:
Your project is very cool, and the source being external helps a lot to lower its temperature, but just out of curiosity, you must have noticed this of course, but did you check the maximum current supplied by 12V and 5V?
using the APS 226 as a base, we have a maximum current of 32A on the 12V line and 3A on the 5V.

I was curious because, at least where I live, it is difficult to find a quality source that provides such a high current.
Click to expand...
I didn't check actual power consumption from grid or current on 12V and 5V rail. Dell D220P-01 is rated only up to 18A on 12V rail, Step-Down Converter has maximum current of 5A. Powered On Time according to webMAN MOD is 42 days and haven't got crashes cause by my power supply setup. Yesterday I have been playing GOW Ascension with overclock for 1~2h with no issue. I'm planning to measure some currents and post my results on the forum with better description. For now, overclock seems stable without any issues.
 
sandungas said:
I "fixed" a word in your quote, did you mean the lv1 is the responsible to read that value from syscon EEPROM, right ?
In other words... the lock in the latest firmwares happens because the old versions of lv1 was reading the CELL clocks from syscon, but the new lv1 versions are not reading it
Click to expand...
CELL and RSX clock settings are always stored in the Syscon EEPROM.
CELL reads it's own clock settings before the config ring is send.
Lv1 uses the EEPROM values to initialize RSX, but versions >0.85 removed that and instead use a fallback value. I can patch that but it's easier to patch the hardcoded value.
 
M4j0r said:
CELL and RSX clock settings are always stored in the Syscon EEPROM.
CELL reads it's own clock settings before the config ring is send.
Click to expand...
So a CELL overclock made at syscon remains after the instalaltion of a OFW, right ?
M4j0r said:
Lv1 uses the EEPROM values to initialize RSX, but versions >0.85 removed that and instead use a fallback value. I can patch that but it's easier to patch the hardcoded value.
Click to expand...
Ok, so the feature can be restored with a lv1 patch, and it would allow to change CELL and RSX clock settings by reprogramming syscon, i guess this could be handy but only for people with an special setup (with the syscon UART-2-USB programmer installed permanently), so it would take only of a couple of minutes to change the clocks
 
Hi All, I've performed a benchmark using Bioshock Infinite with the framerate unlocked measuring both framerate and power consumption between stock vs OC clocks, i've attached my testing methodology below:


Testing specs:

Ambient Temp: 22c
PS3 Model: CECH-2503A (Slim)
Storage: 500GB 7200RPM HDD
Resolution Set on PS3: 1080p
Game Render Resolution: 720p
Fan speed: 35%
Framerate is unlocked via menus (can hit over 100FPS if looking into the sky)

Conclusion: The increase in temperatures and power consumption is negligible (less than 5%, around +2c & 3W) compared to the substantial performance uplift of 15%-20% across the board. No visual artifacts were found.
 
Crysis 3 Benchmark, another title with large gains:


Conclusion: The increase in temperatures and power consumption is negligible (less than 5%) compared to the performance uplift of 15%-20% across the board. As this game spends pretty much all of it's time below it's 30 FPS cap, the upgrade in performance is tangible during gameplay and can be felt even without monitoring software. The only visual artifact I noticed was a 'ring' appearing around phyco's face briefly during the cutscene, appears to be related to LOD's. Not sure if this definitely related to the overclock but not very noticeable either way.

Testing methodology:

Ambient Temp: 23c on stock test & 22c on OC test.
PS3 Model: CECH-2503A (Slim)
Storage: 500GB 7200RPM HDD
Resolution Set on PS3: 1080p
Game Render Resolution: 1024x720
Fan speed: 35%

I tested the loading times from a cold boot and they were identical (well the OC loaded 0.3 seconds quicker if you want to be precise, margin of error stuff).
 
blondbassist said:
Hi All, I've performed a benchmark using Bioshock Infinite with the framerate unlocked measuring both framerate and power consumption between stock vs OC clocks, i've attached my testing methodology below:


Testing specs:

Ambient Temp: 22c
PS3 Model: CECH-2503A (Slim)
Storage: 500GB 7200RPM HDD
Resolution Set on PS3: 1080p
Game Render Resolution: 720p
Fan speed: 35%
Framerate is unlocked via menus (can hit over 100FPS if looking into the sky)

Conclusion: The increase in temperatures and power consumption is negligible (less than 5%, around +2c & 3W) compared to the substantial performance uplift of 15%-20% across the board. No visual artifacts were found.
Click to expand...
Good job bud very helpful, I would like to see this test on GTA IV, Infamous 2

or some exclusive games that only available on ps3
 
I have been looking at this thread for quite some time now and decided to give it a go.
Just done de-lidding my CECH-2000B to optimize the temps which went well, flashed FMW 4.89 - 600/750MHz and it is stable so far.
Is it possible to unlock the framerate of any games? especially for those locked at 30FPS.
Thanks

Edit: Drakengard 3 benefits greatly from the overclock! Played thru chapter 4 and it never dropped below 27fps.
 
Last edited:
Gran Turismo 6 Benchmark:


Conclusion: Performance uplift of around 10%-15% across the board. GT6 almost never hits it's 60fps target even whilst overclocked but the increase in performance is very much welcome with no visual artififacts found. When your PS3 is set to 720p output down from 1080p (or you modify the games PARAM.SFO to such) GT6 will render in 1280x720p instead of 1440x1080 this alone will yield signifant performance increases even without an overclock, but have yet to test this with an overclock active.

Testing Methodology:
Ambient Temp: 21c on stock test & 22c on OC test.
PS3 Model: CECH-2503A (Slim)
Storage: 500GB 7200RPM HDD
Resolution Set on PS3: 1920x1080
Game Render Resolution: 1440x1080
Fan speed: 35%
 
Merhaba arkadaşlar RSX Clock 650 Mhz'de stabil çalışıyor mu cevaplarsanız sevinirim.

English:
Hello friends, does the RSX Clock work stably at 650 Mhz?
 
Last edited by a moderator:
So I have a 28nm RSX super slim coming and I was watching this teardown video. Holy hell, if it's direct IC cooling with a copper heatpipe that means we can get some liquid metal action going and it should run EXTREMELY cool. Yes I'm aware we can't overclock the super slim YET but this looks promising to further overhead for overclocking.

 
zebular said:
So I have a 28nm RSX super slim coming and I was watching this teardown video. Holy hell, if it's direct IC cooling with a copper heatpipe that means we can get some liquid metal action going and it should run EXTREMELY cool. Yes I'm aware we can't overclock the super slim YET but this looks promising to further overhead for overclocking.

Click to expand...

Yes, I'm very much excited at the idea of getting LV1 access on these and overclocking them. It's a shame Sony never ended up die shrinking the CELL to 22nm, would of been crazy efficient in power and thermals.
 
I encountered a glitch in ASURA'S WRATH [BLUS30721] chapter 5 while fighting Vajra Wyzen.
R2 input does not register during a QTE which puts the fight and cut scenes in a loop (or until death of Asura).
System used CECH-2000B 4.89 EVILNAT 600/750 Mhz. I tried with 2 different controllers although there was no issues before reaching that part of the game.
I confirmed that the glitch does not appear on CECHA01 4.89 EVILNAT 500/650 Mhz. The fight scene plays thru as intended.
Same ISO file was used on both systems.
 
Vendest said:
I encountered a glitch in ASURA'S WRATH [BLUS30721] chapter 5 while fighting Vajra Wyzen.
R2 input does not register during a QTE which puts the fight and cut scenes in a loop (or until death of Asura).
System used CECH-2000B 4.89 EVILNAT 600/750 Mhz. I tried with 2 different controllers although there was no issues before reaching that part of the game.
I confirmed that the glitch does not appear on CECHA01 4.89 EVILNAT 500/650 Mhz. The fight scene plays thru as intended.
Same ISO file was used on both systems.
Click to expand...
that suck because Asuras wrath was one of the games that really need the fps boost

maybe try pal version !
 
Vendest said:
I encountered a glitch in ASURA'S WRATH [BLUS30721] chapter 5 while fighting Vajra Wyzen.
R2 input does not register during a QTE which puts the fight and cut scenes in a loop (or until death of Asura).
System used CECH-2000B 4.89 EVILNAT 600/750 Mhz. I tried with 2 different controllers although there was no issues before reaching that part of the game.
I confirmed that the glitch does not appear on CECHA01 4.89 EVILNAT 500/650 Mhz. The fight scene plays thru as intended.
Same ISO file was used on both systems.
Click to expand...
Hello vendest actually i was testing that version that is the same one that i have and i managed to clear chapter 5 with no issues at all, try to have installed the latest updates of the Game and also try to do a database check along with filesystem restore process, that one could help you
 
Last edited:
ashura said:
Hello vendest actually i was testing that version that is the same one that i have and i managed to clear chapter 5 with no issues at allá, try yo hace installed the latest updates of the Game and also try to do a database check along with filesystem restore process, that one could help you
Click to expand...

Just to confirm that you have tested this game on 4.89 EVILNAT 600/750 Mhz? If so then the overclock does not cause this glitch.
 
Vendest said:
Just to confirm that you have tested this game on 4.89 EVILNAT 600/750 Mhz? If so then the overclock does not cause this glitch.
Click to expand...
correct, thats right i do have the same type of model of slim console and its working fine on 4.89 evilnat 600/750 mhz
 
You must log in or register to reply here.

Similar threads

Back
Top