Support NeoGAF

[Bradamante3D]

Hardware -- The popular opinion is that PS3 was a failure, but profit and market share aren't the only measures of success. Technological success can be a measure, but it isn't valued in quarterly/yearly reports and "console wars". Nonetheless, Kutaragi knowingly sacrificed PS3’s near-term profit and market share (he was “aware that with all these technologies, the PS3 can't be offered at a price that's targeted towards households”) for technological success, mid-term profit and maket share. It didn't go exactly as was hoped (didn't make bank and sell 100 million units), but in the end PS3's sacrificial fate made asynchronous compute engine architecture mainstay tech on PC and console. It's the major aspect of CELL that ranger Mark Cerny knows well, now marketed as 'ACE' for AMD GPUs. These GPUs closely mimic CELL's task submission/job execution (i.e., SPURS), but PS4, PS4 Pro and PS5’s GPUs mimic CELL the closest. They “replicate” CELL’s SPU Runtime System (SPURS) and are SPU task manager (STM)-style efficient. STM (entries [0012], [0015] and [0016] explain what and why; [0070], [0071] and [0072] touch on non-CELL implementations and STM results) allows an SPU to load portions of task queues into its local store (i.e., memory) used as “cache”, select the tasks that match its current program code and return the ones that don't to main memory to minimize costly context switching/"cache" flushing. PS4/PS4 Pro's GPUs imitate this process by means of a "volatile bit" that selectively invalidates cache lines to reduce cache flushing when L2 is used for compute and graphics at the same time. PS5 keeps with PS4/PS4 Pro’s volatile bit process and has a similar process for its I/O complex. Coherency engines in the complex tell cache scrubbers on the GPU which address ranges to scrub away so as to avoid full cache flushes when the SSD is read. Without these STM-inspired efficiency measures, PS4 Pro and PS5 are easy marks with no way to effectively counter the brute-force asynchronous GPU compute of One X and Series X or outduel SX’s Velocity architecture. That despite both Xboxes being hobbled by D3D12’s Multi-engine Resource/Enhanced Barrier requiring full GPU cache flushes after work is completed. The unpopular fact is that CELL’s system architecture is the saving grace of post-PS3 era PS consoles. They and future PS consoles are Kutaragi's standard-bearers. His due honor for attempting to make CELL a fixture in the world of computing, which unintentionally yet consequently gave Cerny/SIE the privilege of putting certain CELL related features on AMD's GPU roadmap for the benefit of AMD, SIE and industry. Fate, it seems, is not without a sense of irony. 'The Road to' begins with CELL tech/PS3.

A lot of trollish nonsense there, attributing innovations to either Kutaragi or the CELL chip, that were made either in the RISC space, or the PowerPC group or IBM.
Also, what are all the hardware innovations good for, if even the most loyal Japanese developers were already frustrated witht he PS2 platform, and were then adopting the XBox 360 during the PS3 era - where they saw the trends of the PS2 era continue, instead of being corrected?

 

[VictorVectors]

Maybe I should remove GAF from my ad block whitelist.

lol

It's only natural to feel nostalgic about the peak times of the platform.

If the Sony board reinstalls a profitable CEO instead of a political one, Playstation has every right to regain some of that lightning in a bottle again

 

[Yoboman]

There won't be no playstation with Ken Kutaragi, and PS3 was still great console with great games. Hermen closed Japan Studio and bought Firewalk and Haven.

In reality he restructured Japan Studios around Asobi and it paid off

Also bought Insomniac, Housemarque and Bluepoint

 

[Negotiator]

Isn't this the nutter that almost floored the company by insisting on the cell processor in the PS3.

Without Cell, the GPGPU concept wouldn't exist today.

Even AI wouldn't exist... nVidia wouldn't be a 3 trillion company (they used to have a close partnership with Sony).

Cell deserves a lot more credit than flak.

 

[Ivan]

Just imagine what could Ken do today and how exciting would every new playstation be.... What a LEGEND.

 

[Moses85]

 

[GAF machine]

A lot of trollish nonsense there, attributing innovations to either Kutaragi or the CELL chip, that were made either in the RISC space, or the PowerPC group or IBM.

Huh? I'm not attributing RISC, PowerPC and IBM innovations to Kutaragi. I'm attributing CELL to Kutaragi. The chip doesn’t exist, if not for him (see Acknowledgements).

As for CELL's innovation, it isn't in RISC/PowerPC. It's in the novel system architecture that he broke STI's brains to create:

‘Going multicore solved some of the heat and power problems facing Cell, but it didn't solve everything. Kutaragi was incredibly demanding and repeatedly sent Kahle back to the drawing board. At one point about a year into the project, Kutaragi made the team scrap the whole system structure and start over nearly from scratch. Another time Kutaragi decided he wanted two more cores. Why? "He just wanted to squeeze the engineering team," explains Masakazu Suzuoki, Sony's top Cell engineer, wringing his hands as if strangling a snake. "It hurt your head," Kahle recalls’. - CNNMoney

Also, what are all the hardware innovations good for, if even the most loyal Japanese developers were already frustrated witht he PS2 platform, and were then adopting the XBox 360 during the PS3 era - where they saw the trends of the PS2 era continue, instead of being corrected?

There was nothing to correct. Kutaragi wasn’t interested in making a “dumb PC".

His interest was in giving his consoles (PS1, PS2 and PS3) bespoke higher frequency cores on the CPU that assist a lower frequency GPU with graphics processing in order to spare GPU resources and increase GPU performance.

This came with a steeper learning curve, but it was his way of doing more with less.