GAF machine
Member
"so he built an empire" -- emperor Kaz Hirai
An epic feat that saw the near improbable defeat of Nintendo and Sega at the height of their power, forever changed gaming's landscape and with it, Sony Group Corp.'s fortunes. So influential was his empire, that it drew in ~60% of Sony's revenue at one point; but Kutaragi's tale isn't just one about how he led a small cadre into lands ruled by Nintendo and Sega to wage a vengeful war for console supremacy. His is also a tale of him making legacy provisions before passing the scepter to Hirai that continue to guide the empire's direction and sustain the empire's vigor across various interests:
Hardware -- After witnessing System G's brilliance, Kutaragi embraced a real-time philosophy for 3D CG that his PS1 would be first to establish in the industry as "graphics synthesis". For PS2, he extended the philosophy to physics/simulated expressions, which he dubbed "emotion synthesis". His PS3 unified the two concepts (actually three concepts; "sound synthesis" came first, PS3 enriched it) with stereoscopic 3D and presented it as "reality synthesis", which brought about the prospect of VR on PS4. The emphasis Kutaragi placed on real-time performance drove him to architect radical hardware. None more radical and special than PS3. It was first to embody all four real-time concepts and has been the source of inspiration behind PS hardware customization ever since. Kutaragi took a big risk when he traded ease of development for a chip faster than Moore's Law, but that trade handed PS3 and successive generations of PS consoles an ingenious weapon with which to combat more powerful consoles. SPURS, the SPU Runtime System enhanced with STM was the weapon. It was forged in the fire of CELL's asynchronous compute engine architecture. More specifically, in the fire of SPU direct memory access (DMA). Such was the weapon's effectiveness that ranger Mark Cerny (he helped forge it, knows it well) sought to arm PS4's GPU with it. SIE struck a deal with AMD's merchants to equip PS4s, and in exchange AMD got to market SPURS as 'ACE' for their GPU line. A fair deal, but AMD's GPUs can't "replicate the SPU Runtime System (SPURS) of the PS3" in the same way that PS4, PS4 Pro and PS5's GPUs can. That's because AMD GPUs lack a hardware implementation of STM. The SPU task manager (STM) makes SPURS highly efficient (entries [0012], [0015] and [0016] explain what and why; [0070], [0071] and [0072] touch on non-CELL implementations and STM results). It allows an SPU to treat it's local store (i.e., memory) like cache, and select only the tasks that match its current program code as a way 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 the 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. With SPURS/STM, PS4, PS4 Pro and PS5 are forces to be reckoned with. Without it, they're easy marks that have no effective counter against One X and Series X's brute-force GPU asynchronous compute or Series X's Velocity architecture. That despite D3D12's Multi-engine Resource/Enhanced Barrier hindering both Xbox GPUs by requiring excessive cache flushes after they complete their work. The popular opinion is that PS3 was a failure, but profitability and market share aren't the only measures of success. The unpopular fact is that PS3 was a triumph of engineering and has become the saving grace of current/future PS consoles. They 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. 'The Road to' began with CELL. It continues to prove itself a valuable resource and wise expenditure of the empire's treasure.
Software -- Kutaragi's crucible (i.e., a precursor CPU that's GPU-like) tested the mettle of PS Studios' ranks and steeled them for battle in the age of GPU asynchronous compute programming years before the dawn. For two generations, Redmond's assaults have been repelled by happy warriors (e.g., the ICE Team) highly skilled in the art of PS3. Mastery of Packing fp16 shader operations on SPUs and checkerboard MLAA on SPUs were trials for fp16 'Rapid Packed Math' usage and checkerboard rendering on PS4 Pro. PS3 pushed PS Studios' elite to be the most cunning at their craft. The evidence of history gives credence to the claim. Kutaragi is the only CEO (or system architect for that matter) to ever assure SIE's hardware and programmer competence ahead of future trends. The grateful warriors salute him.
Network -- As SCEI's emperor with broad authority inside Sony Group Corp.'s larger empire, Kutaragi the Great commissioned the PSNP's construction to commune PS users, traffic e-commerce and provide a foothold in network services across platforms for generations to come. It was his long-held vision brought to emperor Jim Ryan's remembrance in later years: "I think he foresaw the network era and was talking about the network era many, many years before it became reality"... "he talked about this before we could really comprehend what he was saying". His visionary leadership and persistence (he tried with PS BB in Japan, but it failed to gain traction; then he tried with PSN) are the reasons for PSN's appearance on SIE's timeline. The network's impact can't be fully appraised, but a partial appraisal is that PSN has been vital to most growth vectors. The platform has generated billions in profit since operation and is the enabler of SIE's expansion into cloud, PC and mobile gaming, which will generate billions more in perpetuity. In the Web3 era, PSN will host P2P transactions and records of interoperable distributed ledgers (i.e., interoperable blockchains). The latter is an inevitability that Kutaragi didn't foresee, but he foresaw the former and it ties into the latter. CELL was conceived to be "a bus for peer-to-peer computing" (translated by DeepL) on a network; which a scribe worded as 'a peer-to-peer web that would transform every home into a kind of domestic (and legal) Napster'. Kutaragi had an uncanny intuition about the future of PS3's tech, but 7 years wasn't enough time for that future to emerge. He still speaks of the P2P model from time to time; but as it relates to blockchain and NFTs. PS3 didn't bring P2P transactions to pass, but a future PS console will.
Multimedia-as-a-Service/Transmedia -- As a point of clarity, Kutaragi declared SCE/SIE an entertainment company and laid out a timeless '21st Century Entertainment' strategy that draws from all corners of Sony Group Corp. to widen PlayStation's appeal beyond gaming. His vision was to wield popular entertainment as an instrument with which to further spread PS's influence, under the declaration that PlayStation is for entertainment, not just games. A scribe privy to Kutaragi's declaration noted: 'games and other services for PlayStation. His goal is to establish the PlayStation as a separate brand, not part of a dedicated Sony network for the online distribution of music, movies, and TV shows'. SIE's latest proposal to transition PS Plus into "Services 3.0" and scale it out in partnership with leading subscription service providers is testament to his declaration and a re-commitment to his entertainment strategy. It's also part of a larger aim to transform SIE into a transmedia superpower through collaboration with other Sony Group Corp. divisions, in accordance with his strategy.
GaaS/Live Service -- As the scribe noted, GaaS was an objective of Kutaragi's entertainment services ambition. SIE pursuing live service games now was always a matter of when, not if.
GaaS/Game Streaming -- Kutaragi saw 'the Cloud' gathering from afar and readied PS3 for a ubiquitous game streaming future before it made sense to the masses. Prior to PS2's US launch, he riddled that PS3 would "totally disappear as a console, as a shape". At 9:51 - 9:52 Tokyo time (2:51 - 2:52 BST if Tokyo time isn't shown) of his final keynote as emperor, he justified PS3's engineering as necessary to take advantage of a speedier future network that he predicted would be fast enough to reliably stream video for services in "ten years time" (i.e., by year 2016, ten years from his 2006 keynote). PS Now launched in major territories from 2014 - 2016 as a nebulous PS3 requiring a 5mbps connection speed for 720p streaming. As it turns out, one SPU can encode a 720p stream at 5mbps. By this metric, every initial PS Now subscriber and every PS3 game that's ever been streamed via Plus Premium is a credit to him. Premium's PS3 cloud servers are as much a credit to him as they are to David Perry, Steve Perlman and the noble scout, Eric Lempel. SIE was first to act on Gaikai (streamed games to any webpage/browser), OnLive (used perceptual tweaks to mask latency) and iSIZE (used AI-based perceptual pre-encoding to mask latency) because Kutaragi was first to know: "In the future, broadband will connect all appliances - console, TV, phone, PC, everything. Then exclusivity means nothing".
PC -- Kutaragi considered PC neither friend nor foe to PS consoles; but his notion that PS exclusives would come to PC in the then-future via streaming means he knew that PC would have a role in the empire's scheme to seize greater mind share. The eventuality of letting Plus Premium members stream first-party titles they purchase for PC (e.g., Helldivers 2 via Steam Cloud Play developer opt-in) to a TV, phone, tablet or PS Portal follows his logic. If followed to its logical conclusion, then Steam will eventually have a role to play in SIE's scheme, too. The titan in Santa Clara may have to defend against Geforce Now falling to Plus Premium's sword one day.
Mobile -- With the threat of mobile's march into gaming looming large on the horizon, Kutaragi began work to fashion a shield that PS consoles could use in the inevitable fight. Scrolls from the period reveal that before he relinquished the throne, a PSP DualShock dock and remote play controller (under his name) that served music, movies, TV shows, mini games (possibly PS Mobile for Android games) and e-mails were contemplated. The remote play controller arose out of his suspicion that mobile phones would eventually be an 'enemy' of PS consoles. In his absence, the remote play controller went lost to time in the archives; until it was publicized as a new USPTO filing in 2019. Four years later, it was pared-down and launched as 'PlayStation Portal'. For now it's touted as just a remote player, but despite having ~6GB storage (guessing here, I think it's that small because SIE plans to update Portal with a machine learning-based downloader that "cloud delivers"/auto deletes individual mobile game levels and areas as necessary for continuous local play or resumption of local play; hence Portal's rumored codename: 'Q' 'Lite'), it's only a few approved media/VoWifi apps away from also being the single entertainment/communications device he spoke of in the '90s (emperor Andrew House recalled: "I remember him talking about there was going to be a convergence between all forms of media and entertainment into a single communications device"). The head craftsman Hideaki Nishino and his smiths crafted Portal's appearance from a DualSense perspective, but Kutaragi hammered out the specifics of Portal's motion sensing, remote play, cloud gaming and mobile aspects years in advance.
AI -- Long before LLMs and autonomous AI were hot topics, Kutaragi's take was that "one Cell or cluster of Cells can also function as a server; but whereas the present Internet mainly handles characters, applications on the Cell network will also handle semantics and reasoning". He always thought ahead and occasionally pushed to do more than necessary for the glory of the empire. CELL was one such push, done in part to spark an AI revolution within SIE. A couple years after PS3's launch, the neighboring state of IBM presented its approach to training neural networks on CELL processors and a workaroundfor SPEs to compute training functions when bandwidth is limited. The work didn't yield anything tangible for PS3 games, but that doesn't negate the fact that CELL foreshadowed AI accelerators. Its array of SPEs are SIMD dataflow units with support for 8-bit operations and SRAM connected to a high-bandwidth on-chip network that allows SPEs to access each others SRAMs directly. These are attributes of dedicated AI accelerators from IBM, Graphcore and others (last paragraph). "Enormous applications that require image processing and large-volume computing" were other CELL use cases he considered. A real-world example is Toshiba's SpursEngine variant. It was used for Super Resolution video up-conversion in the image processing domain. Almost two decades later, here comes PS5 Pro with talk of 8-bit computation for ML-based Super Resolution up-scaling. PS5 Pro's AI acceleration hardware is still a mystery, but if "past is prologue" then the hardware will be Tempest Engine-like; which means its design will also hark back to the new age chip Kutaragi envisioned.
Metaverse -- Ryan saw PS Home as "a very early manifestation of a platform metaverse". Kutaragi saw it as "only the beginning of what will come". What was yet to come (and still is) was a realm of "make.believe-as-a-service" illusions and spectacles to enchant the people. This realm wasn't a metaverse experience, it was a blended 'Cyber Society' experience. A mixed realm that "combines the real and the imaginary" with music, movies, games and other entertainment. The power to muster the realm's alchemy in cyberspace was to come from a CELL network. Likely, a network of petaFLOP class server room CELL supercomputers connected to PS3 consoles distributed across the land. Decades on, his grand vision of a 'Cyber Society' now aligns with emperor Kenichiro Yoshida's perception of the so-called "metaverse" and Sony's approach to it (Q2., Q3. and Q4). Yoshida's "metaverse" proclamations of today are no different than Kutaragi's 'Cyber Society' proclamations of yesteryear, which were roundly dismissed as "krazy" talk at the time. As time has shown, he wasn't crazy. He was clairvoyant and well-versed in technologies. One of his earliest insights into a MR cyberverse was recorded by a scribe who wrote that while he was waving his arm, he said: "This, this is the real broadband interface" (last sentence of the article Onslaught posted). The alchemy of reality interfaces have come in phases throughout the empire's existence. Phase 1 was EyeToy for Play, Groove, Chat and Kinetic, PS Eye with EyePet in S3D and PS Move for PS3 Reality Synthesis (i.e. Stereoscopic 3D worlds rendered in real-time by CELL SPUs and the RSX GPU), which served as a PS3 provided VR testbed and ultimately gave rise to PS VR. PS Camera and its associated Asobi/AR Bots, PS HD Camera with PS VR2/Sense controllers and social VR experimentation were Phase 2. Phase 3 is in follow ups to these peripherals and experiences. In particular, the next PS HMD (PS XR?) which the upcoming Sony XR HMD already solves for. In the Phase 3 era, Kutaragi's 'Cyber Society' will no longer be a concept of his imagination; rather a reality that his consoles (PS2, PS3) encouraged Magic Lab's magicians and Sony Electronics' wizards to formulate.
Frankly, Kutaragi was wise beyond his years on the throne. The passage of time, brand/IP maturation and tech advancement across industries have affirmed his sweeping vision; and have allowed his successors to see what he saw decades earlier as they look to further build on the foundation he laid, while moving to extend the empire's reach into areas of interest he identified beyond the traditional pillars set in place during his rule.
His greatest success is that he made the success of every SCE/SIE emperor before (Toshio Ozawa, Shigeo Maruyama, Teruhisa Tokunaka, etc.) and after him possible. Those who came before are beneficiaries of his gamble to create real-time computer entertainment. Those who came/come after are inheritors of provisions he made to ensure the empire's survival and competitiveness.
Past, present and future emperors can say they cut costs, streamlined operations, consolidated silos, improved communications, avoided pitfalls, entered partnerships, made exclusive deals, approved major acquisitions, pivoted to other areas for growth and led the empire through challenging circumstances on the way to greater expansion and wealth; but Kutaragi is the only one who can say that he did most if not all the same and defined PlayStation's past, present and future.
Here's to Ken Kutaragi, the man, the myth, the legend, the visionary, "the Oracle", The G.O.A.T. May he continue to live long.
Long live the empire.
An epic feat that saw the near improbable defeat of Nintendo and Sega at the height of their power, forever changed gaming's landscape and with it, Sony Group Corp.'s fortunes. So influential was his empire, that it drew in ~60% of Sony's revenue at one point; but Kutaragi's tale isn't just one about how he led a small cadre into lands ruled by Nintendo and Sega to wage a vengeful war for console supremacy. His is also a tale of him making legacy provisions before passing the scepter to Hirai that continue to guide the empire's direction and sustain the empire's vigor across various interests:
Hardware -- After witnessing System G's brilliance, Kutaragi embraced a real-time philosophy for 3D CG that his PS1 would be first to establish in the industry as "graphics synthesis". For PS2, he extended the philosophy to physics/simulated expressions, which he dubbed "emotion synthesis". His PS3 unified the two concepts (actually three concepts; "sound synthesis" came first, PS3 enriched it) with stereoscopic 3D and presented it as "reality synthesis", which brought about the prospect of VR on PS4. The emphasis Kutaragi placed on real-time performance drove him to architect radical hardware. None more radical and special than PS3. It was first to embody all four real-time concepts and has been the source of inspiration behind PS hardware customization ever since. Kutaragi took a big risk when he traded ease of development for a chip faster than Moore's Law, but that trade handed PS3 and successive generations of PS consoles an ingenious weapon with which to combat more powerful consoles. SPURS, the SPU Runtime System enhanced with STM was the weapon. It was forged in the fire of CELL's asynchronous compute engine architecture. More specifically, in the fire of SPU direct memory access (DMA). Such was the weapon's effectiveness that ranger Mark Cerny (he helped forge it, knows it well) sought to arm PS4's GPU with it. SIE struck a deal with AMD's merchants to equip PS4s, and in exchange AMD got to market SPURS as 'ACE' for their GPU line. A fair deal, but AMD's GPUs can't "replicate the SPU Runtime System (SPURS) of the PS3" in the same way that PS4, PS4 Pro and PS5's GPUs can. That's because AMD GPUs lack a hardware implementation of STM. The SPU task manager (STM) makes SPURS highly efficient (entries [0012], [0015] and [0016] explain what and why; [0070], [0071] and [0072] touch on non-CELL implementations and STM results). It allows an SPU to treat it's local store (i.e., memory) like cache, and select only the tasks that match its current program code as a way 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 the 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. With SPURS/STM, PS4, PS4 Pro and PS5 are forces to be reckoned with. Without it, they're easy marks that have no effective counter against One X and Series X's brute-force GPU asynchronous compute or Series X's Velocity architecture. That despite D3D12's Multi-engine Resource/Enhanced Barrier hindering both Xbox GPUs by requiring excessive cache flushes after they complete their work. The popular opinion is that PS3 was a failure, but profitability and market share aren't the only measures of success. The unpopular fact is that PS3 was a triumph of engineering and has become the saving grace of current/future PS consoles. They 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. 'The Road to' began with CELL. It continues to prove itself a valuable resource and wise expenditure of the empire's treasure.
Software -- Kutaragi's crucible (i.e., a precursor CPU that's GPU-like) tested the mettle of PS Studios' ranks and steeled them for battle in the age of GPU asynchronous compute programming years before the dawn. For two generations, Redmond's assaults have been repelled by happy warriors (e.g., the ICE Team) highly skilled in the art of PS3. Mastery of Packing fp16 shader operations on SPUs and checkerboard MLAA on SPUs were trials for fp16 'Rapid Packed Math' usage and checkerboard rendering on PS4 Pro. PS3 pushed PS Studios' elite to be the most cunning at their craft. The evidence of history gives credence to the claim. Kutaragi is the only CEO (or system architect for that matter) to ever assure SIE's hardware and programmer competence ahead of future trends. The grateful warriors salute him.
Network -- As SCEI's emperor with broad authority inside Sony Group Corp.'s larger empire, Kutaragi the Great commissioned the PSNP's construction to commune PS users, traffic e-commerce and provide a foothold in network services across platforms for generations to come. It was his long-held vision brought to emperor Jim Ryan's remembrance in later years: "I think he foresaw the network era and was talking about the network era many, many years before it became reality"... "he talked about this before we could really comprehend what he was saying". His visionary leadership and persistence (he tried with PS BB in Japan, but it failed to gain traction; then he tried with PSN) are the reasons for PSN's appearance on SIE's timeline. The network's impact can't be fully appraised, but a partial appraisal is that PSN has been vital to most growth vectors. The platform has generated billions in profit since operation and is the enabler of SIE's expansion into cloud, PC and mobile gaming, which will generate billions more in perpetuity. In the Web3 era, PSN will host P2P transactions and records of interoperable distributed ledgers (i.e., interoperable blockchains). The latter is an inevitability that Kutaragi didn't foresee, but he foresaw the former and it ties into the latter. CELL was conceived to be "a bus for peer-to-peer computing" (translated by DeepL) on a network; which a scribe worded as 'a peer-to-peer web that would transform every home into a kind of domestic (and legal) Napster'. Kutaragi had an uncanny intuition about the future of PS3's tech, but 7 years wasn't enough time for that future to emerge. He still speaks of the P2P model from time to time; but as it relates to blockchain and NFTs. PS3 didn't bring P2P transactions to pass, but a future PS console will.
Multimedia-as-a-Service/Transmedia -- As a point of clarity, Kutaragi declared SCE/SIE an entertainment company and laid out a timeless '21st Century Entertainment' strategy that draws from all corners of Sony Group Corp. to widen PlayStation's appeal beyond gaming. His vision was to wield popular entertainment as an instrument with which to further spread PS's influence, under the declaration that PlayStation is for entertainment, not just games. A scribe privy to Kutaragi's declaration noted: 'games and other services for PlayStation. His goal is to establish the PlayStation as a separate brand, not part of a dedicated Sony network for the online distribution of music, movies, and TV shows'. SIE's latest proposal to transition PS Plus into "Services 3.0" and scale it out in partnership with leading subscription service providers is testament to his declaration and a re-commitment to his entertainment strategy. It's also part of a larger aim to transform SIE into a transmedia superpower through collaboration with other Sony Group Corp. divisions, in accordance with his strategy.
GaaS/Live Service -- As the scribe noted, GaaS was an objective of Kutaragi's entertainment services ambition. SIE pursuing live service games now was always a matter of when, not if.
GaaS/Game Streaming -- Kutaragi saw 'the Cloud' gathering from afar and readied PS3 for a ubiquitous game streaming future before it made sense to the masses. Prior to PS2's US launch, he riddled that PS3 would "totally disappear as a console, as a shape". At 9:51 - 9:52 Tokyo time (2:51 - 2:52 BST if Tokyo time isn't shown) of his final keynote as emperor, he justified PS3's engineering as necessary to take advantage of a speedier future network that he predicted would be fast enough to reliably stream video for services in "ten years time" (i.e., by year 2016, ten years from his 2006 keynote). PS Now launched in major territories from 2014 - 2016 as a nebulous PS3 requiring a 5mbps connection speed for 720p streaming. As it turns out, one SPU can encode a 720p stream at 5mbps. By this metric, every initial PS Now subscriber and every PS3 game that's ever been streamed via Plus Premium is a credit to him. Premium's PS3 cloud servers are as much a credit to him as they are to David Perry, Steve Perlman and the noble scout, Eric Lempel. SIE was first to act on Gaikai (streamed games to any webpage/browser), OnLive (used perceptual tweaks to mask latency) and iSIZE (used AI-based perceptual pre-encoding to mask latency) because Kutaragi was first to know: "In the future, broadband will connect all appliances - console, TV, phone, PC, everything. Then exclusivity means nothing".
PC -- Kutaragi considered PC neither friend nor foe to PS consoles; but his notion that PS exclusives would come to PC in the then-future via streaming means he knew that PC would have a role in the empire's scheme to seize greater mind share. The eventuality of letting Plus Premium members stream first-party titles they purchase for PC (e.g., Helldivers 2 via Steam Cloud Play developer opt-in) to a TV, phone, tablet or PS Portal follows his logic. If followed to its logical conclusion, then Steam will eventually have a role to play in SIE's scheme, too. The titan in Santa Clara may have to defend against Geforce Now falling to Plus Premium's sword one day.
Mobile -- With the threat of mobile's march into gaming looming large on the horizon, Kutaragi began work to fashion a shield that PS consoles could use in the inevitable fight. Scrolls from the period reveal that before he relinquished the throne, a PSP DualShock dock and remote play controller (under his name) that served music, movies, TV shows, mini games (possibly PS Mobile for Android games) and e-mails were contemplated. The remote play controller arose out of his suspicion that mobile phones would eventually be an 'enemy' of PS consoles. In his absence, the remote play controller went lost to time in the archives; until it was publicized as a new USPTO filing in 2019. Four years later, it was pared-down and launched as 'PlayStation Portal'. For now it's touted as just a remote player, but despite having ~6GB storage (guessing here, I think it's that small because SIE plans to update Portal with a machine learning-based downloader that "cloud delivers"/auto deletes individual mobile game levels and areas as necessary for continuous local play or resumption of local play; hence Portal's rumored codename: 'Q' 'Lite'), it's only a few approved media/VoWifi apps away from also being the single entertainment/communications device he spoke of in the '90s (emperor Andrew House recalled: "I remember him talking about there was going to be a convergence between all forms of media and entertainment into a single communications device"). The head craftsman Hideaki Nishino and his smiths crafted Portal's appearance from a DualSense perspective, but Kutaragi hammered out the specifics of Portal's motion sensing, remote play, cloud gaming and mobile aspects years in advance.
AI -- Long before LLMs and autonomous AI were hot topics, Kutaragi's take was that "one Cell or cluster of Cells can also function as a server; but whereas the present Internet mainly handles characters, applications on the Cell network will also handle semantics and reasoning". He always thought ahead and occasionally pushed to do more than necessary for the glory of the empire. CELL was one such push, done in part to spark an AI revolution within SIE. A couple years after PS3's launch, the neighboring state of IBM presented its approach to training neural networks on CELL processors and a workaroundfor SPEs to compute training functions when bandwidth is limited. The work didn't yield anything tangible for PS3 games, but that doesn't negate the fact that CELL foreshadowed AI accelerators. Its array of SPEs are SIMD dataflow units with support for 8-bit operations and SRAM connected to a high-bandwidth on-chip network that allows SPEs to access each others SRAMs directly. These are attributes of dedicated AI accelerators from IBM, Graphcore and others (last paragraph). "Enormous applications that require image processing and large-volume computing" were other CELL use cases he considered. A real-world example is Toshiba's SpursEngine variant. It was used for Super Resolution video up-conversion in the image processing domain. Almost two decades later, here comes PS5 Pro with talk of 8-bit computation for ML-based Super Resolution up-scaling. PS5 Pro's AI acceleration hardware is still a mystery, but if "past is prologue" then the hardware will be Tempest Engine-like; which means its design will also hark back to the new age chip Kutaragi envisioned.
Metaverse -- Ryan saw PS Home as "a very early manifestation of a platform metaverse". Kutaragi saw it as "only the beginning of what will come". What was yet to come (and still is) was a realm of "make.believe-as-a-service" illusions and spectacles to enchant the people. This realm wasn't a metaverse experience, it was a blended 'Cyber Society' experience. A mixed realm that "combines the real and the imaginary" with music, movies, games and other entertainment. The power to muster the realm's alchemy in cyberspace was to come from a CELL network. Likely, a network of petaFLOP class server room CELL supercomputers connected to PS3 consoles distributed across the land. Decades on, his grand vision of a 'Cyber Society' now aligns with emperor Kenichiro Yoshida's perception of the so-called "metaverse" and Sony's approach to it (Q2., Q3. and Q4). Yoshida's "metaverse" proclamations of today are no different than Kutaragi's 'Cyber Society' proclamations of yesteryear, which were roundly dismissed as "krazy" talk at the time. As time has shown, he wasn't crazy. He was clairvoyant and well-versed in technologies. One of his earliest insights into a MR cyberverse was recorded by a scribe who wrote that while he was waving his arm, he said: "This, this is the real broadband interface" (last sentence of the article Onslaught posted). The alchemy of reality interfaces have come in phases throughout the empire's existence. Phase 1 was EyeToy for Play, Groove, Chat and Kinetic, PS Eye with EyePet in S3D and PS Move for PS3 Reality Synthesis (i.e. Stereoscopic 3D worlds rendered in real-time by CELL SPUs and the RSX GPU), which served as a PS3 provided VR testbed and ultimately gave rise to PS VR. PS Camera and its associated Asobi/AR Bots, PS HD Camera with PS VR2/Sense controllers and social VR experimentation were Phase 2. Phase 3 is in follow ups to these peripherals and experiences. In particular, the next PS HMD (PS XR?) which the upcoming Sony XR HMD already solves for. In the Phase 3 era, Kutaragi's 'Cyber Society' will no longer be a concept of his imagination; rather a reality that his consoles (PS2, PS3) encouraged Magic Lab's magicians and Sony Electronics' wizards to formulate.
Frankly, Kutaragi was wise beyond his years on the throne. The passage of time, brand/IP maturation and tech advancement across industries have affirmed his sweeping vision; and have allowed his successors to see what he saw decades earlier as they look to further build on the foundation he laid, while moving to extend the empire's reach into areas of interest he identified beyond the traditional pillars set in place during his rule.
His greatest success is that he made the success of every SCE/SIE emperor before (Toshio Ozawa, Shigeo Maruyama, Teruhisa Tokunaka, etc.) and after him possible. Those who came before are beneficiaries of his gamble to create real-time computer entertainment. Those who came/come after are inheritors of provisions he made to ensure the empire's survival and competitiveness.
Past, present and future emperors can say they cut costs, streamlined operations, consolidated silos, improved communications, avoided pitfalls, entered partnerships, made exclusive deals, approved major acquisitions, pivoted to other areas for growth and led the empire through challenging circumstances on the way to greater expansion and wealth; but Kutaragi is the only one who can say that he did most if not all the same and defined PlayStation's past, present and future.
Here's to Ken Kutaragi, the man, the myth, the legend, the visionary, "the Oracle", The G.O.A.T. May he continue to live long.
Long live the empire.
Last edited:
