original japanese article / translation by excite.co.jpwatch.impress.co.jp said:The semiconductor strategy of Sony told by Kutaragi Ken
The Sony group which continues investment in the latest semiconductor process for a PlayStation family. Mr. Kutaragi (the Sony vice president / the Sony Computer Entertainment president and CEO) was asked about the semiconductor strategy of this group, and future directivity.
- The following keyword is as a company name entertainment.
[Q] The design of the tip of SCEI is harnessing the advantage of a semiconductor. SHURINKU [shrink,SHuRINKu] [ it ] two generations since a game machine is a five-year cycle. It seems to design the tip on the assumption that it.
[Mr. Kutaragi] Although it may be two generations in PC, there is optical shrink, and there is shrink on a network list etc. and between them has concentration of a circumference tip. Such shrink is repeated. Since the number of harvest per unit investment will increase if and it says what is good, cost should fall.
However, there is also a problem. Although the number of shipment should just be extended by the upward rise to PC ???, the renewal of a platform (game machine) will come by 1 time of the cycle in several. Therefore, a manufacture capacity will remain shortly at the time of peak out (a shipment is stabilizing).
However, remaining is a negative idea. In fact, I do unreasonableness and am making (chip fabrication factory). It becomes the talk what to do by made KYAPA. Extend in the direction where it makes others (tip), such as PSP.
Moreover, a semiconductor (manufacture line) is not crushed for every change of generation. It leaves like Intel (also in case of old line). Then, if it says what can be done, various things can be taken in also to depreciated Fab (factory). This has taken the same strategy as it in which are the latest (process) and begin manufacture of the latest stone (tip), and Intel will finally have and go to manufacture [ of a flash memory ] (if a process depreciates).
[Q] The way of thinking similar to Intel was taken.
[Mr. Kutaragi] The way of thinking of right [ that ] and Sony is not the semiconductor store Mr.-way of thinking but the Intel-way of thinking. If it is the semiconductor-way of thinking and will put together, it will become double pain and a triple handicap. (Per tip) Since a unit price falls, and a selling price falls and a manufacture capacity remains. We are doing the semiconductor and the system enterprise to it. It is a strong point.
If it does so, the volume application of the maximum [ next ] as a system will become a talk of something. One is PC and another is a game machine. - Although it thinks it right to call it computer entertainment apparatus after this, - and one more are pocket apparatus. However, since there is a footprint (restrictions of tip size) in the case of pocket apparatus, although a number (tip) comes out, a tip (DAI area) is seldom used. If it sees from a semiconductor enterprise, it will become the market closed comparatively.
[Q] By "Microprocessor Forum" for '99 years, if computer (Mr. Kutaragi) entertainment industry spreads 100 times 10 times, he remembers the time of carrying out a presentation.
[Mr. Kutaragi] Although it was touch that it was what [untranslated] wholly (smile).
I think that the following keyword is as the name of a company entertainment. If the next huge demand (semiconductor) of PC is the entertainment in a large meaning. For example, it is said from before that communication is the greatest entertainment. It spreads and goes centering on entertainment.
computer entertainment apparatus -- a home and mobile one -- it is that it is portable Speaking theoretically, in the number, exceeding PC, since a home finally goes on a par with the number of TV. furthermore, the portable number -- the number of man -- when it becomes mobile, I want to come to have many sets alone Therefore, by the number, it should go to the place far exceeding PC.
In this way, if it sees, the next of PC will become the time of CE (household electric appliances). Every person comes to have various apparatus not only including entertainment apparatus but a digital camera. Finally, the extreme Big Bang will occur, since all the flats TV will have a processor in inside (semiconductor demand in household electric appliances).
In such inside, an investment scenario becomes very important. This is not visible at all in giving only the semiconductor.
- It is a natural conclusion that Sony has a semiconductor.
[Q] Therefore, it is if it is necessary to have both a system and a semiconductor.
[Mr. Kutaragi] It cannot thrust in, unless it has from process technology. Like the Sony group, if volume application is considered, it is unthinkable to depend on the other company (semiconductor). For example, Mr. Intel only designs CPU and carrying out manufacture consignment cannot consider it in Mr. other company, for example, TSMC. It is a natural conclusion that it is the same as it and Sony has a semiconductor.
Moreover, there is also a background that a semiconductor ratio becomes high from now on. Until now, it was the world of mechatronics and possibly the semiconductor ratio was about 10%. However, the semiconductor ratio in the hard inside which removed a firmware and software becomes 7 - 80 percent after this. If it is all purchased from outside, all become cash out, and because the other company (semiconductor equipment) is depreciated, it will become. However, even if it refunds with much trouble, it does not know whether they continue investment in the latest process.
Since that Mr. Intel can be maintaining competitive power gained profits with them huge every year every year and reinvestment is applied. Since profits are allotted to plant-and-equipment investment and research and development, it has been rapidly stacked by IP (intellectual possessions) and manufacturing capacity is accumulated. It is a wonderful cycle.
[Q] Sony was not in such a cycle until now.
[Mr. Kutaragi] As for the portion of analog mechanism NIKARU, Sony was doing [untranslated phrase] and ASETTO management until now. considerable in a basic device -- long -- it period-applied, developed and large-sized manufacture investment was carried out And the competitive device has been raised in the company.
However, I think that a scenario began to be out of order from the time of becoming digital and AV and IT uniting just for a moment. Since it was going to do rapidly in the place which know-how accumulates, a manufacture infrastructure accumulates, and a platform stacks, and does not have a raising when it was called [ or ] why. Since there was no technology in the AV side, it became using the solution of IT. When doing so, it became the structure supplied from outside (other company) in horizontal division of labor like IT.
However, if it becomes so, added value all flows out, and even if it makes the product regarded as this being the point aimed at of good fortune, it will have structure which the other company tends to run after. This is not being what happened only to Sony and having happened to all home electronics makers. Since that get it blocked and the profit ratio of a home electronics maker is falling is doing what South Korea, Taiwan, China, and the included wide use take place, and can be made in every area of other. since it is standardized for compatibility -- [untranslated phrase] [ LSI ] -- being alike -- [untranslated phrase] -- a thing can be made anywhere Therefore, even if it takes some out, it is the price which does not profit in the twinkling of an eye.
- The Sony group unites.
[Q] It is to have become the same industrial structure as PC.
[Mr. Kutaragi] PC can take out PC also with Mr. NEC, Mr. FUJITSU, or Sony anywhere, although the copy tip of Intel does not come out. There is nothing without whose making a profit, and Intel and Microsoft are profitable perfectly there.
For the reason, Intel threw up even DRAM. It is not "regrettably not doing DRAM, although it profits", and only the place which profits most is done. If two or more makers fight all together and the back should just become a commodity. I think that it is the business model completed by brilliant calculation.
However, in order to succeed by the system here, you have to do the same thing as Dell Computer. Although the supply chain management of Dell is uncanny, neither Dell nor Amazon.com can think that no less than 100 companies are realized all over the world. Only the company restricted very much remains.
[Q] Intel will become painful if a time moves from PC to CE.
[Mr. Kutaragi] Supposing Intel stops at PC, it may become painful. However, after all, Hurd and software sublimate a mainframe computer, can do it and the present computing is not independently their invention. Therefore, if a new direction can be seen, naturally they will also go there. However, a direction will not only be seen.
Ten years ago, when PS (PlayStation) was taken out, it could not see (them), and it was not visible even when several year before PS 2 (PlayStation 2) was taken out. Now, it may be beginning to feel at last.
Of course, PC is very convenient at PC, it is the tool to which man's capability is expanded, and it is uncanny at it. However, in PC (CE), the purpose is not the same. We go to the entertainment and home instead of a tool which an individual enjoys. It is the direction of Sony.
And it is bad, if the selection and concentration of investment for development which are turned previously are considered exactly and its manufacture infrastructure is not done by its investment not only Sony but after this.
[Q] If it does so, the flow which goes to CE will be foreseen and plant-and-equipment investment and technical accumulation will be performed. If the Sony group adopts Cell computing as the next-generation architecture of the core.
[Mr. Kutaragi] It is handled by all SONI.
It is the scenario of sharing each other's mutual strong points, uniting completely, and repulling once runaway No. 1 after this until now although a game group and other groups stimulate each other within a group.
[Q] Was the talk of the neighborhood done with the Sony head office for some time?
[Mr. Kutaragi] Now [ all ], consensus is reached.
- Put silicone photograph NIKUSU into a view.
[Q] SCEI seems to adhere to the inclusion memory very much.
[Mr. Kutaragi] The biggest problem is supply of a memory.
For the present PC architecture, the footprint (the amount of loading) of a memory is a life. And it is decided by the quantity and the type of a memory in which PC industry can produce which can stack a main memory by it with all (memory industry) investment. Since that has little volume in the ability of the high-end graphics of PC to adopt DDR2 and DDR3.
However, if TV results in tens of millions of sets and a cellular phone, as for consumer, the unit of 100 million comes out. When it does so, it is difficult for me to have about the same footprint as PC supplied. It will be restrained.
Since that I do Embedded DRAM wants to cut off bondage (DRAM industry) in the consumer of a huge shipment. Although the footprint of a memory decreases, I want to do my best there with a device as the side effects.
[Q] It was said that the topology of a computer was changed by optical interchange connection. For that purpose, it is [ that it is necessary to conquer the latency of a network, and ].
[Mr. Kutaragi] Even if there is a latency, to be able to predict it and what is necessary is to be able to be just adapted. It tests on speculation (an object is performed) but perfectly (latency).
I want you to think. A surprising latency is eaten by length of about 10mm on a tip. Although there is technology which makes it high-speed (copper wiring on a board), such as Redwood, the speed which can be attained does not have a great thing compared with light. The speed of the photon passing through the inside of an optical fiber is completely different. Therefore, it goes there fundamentally.
[Q] Then, does it enter and go to silicone photograph NIKUSU (integration of the opto-electronics to a silicon chip top)?
[Mr. Kutaragi] They are right [ that ] and silicone photograph NIKUSU. It is our research domain and is studied also in Mr. Intel. Fundamentally, finally it goes to the world of laser or a diode from the semiconductor of a MOS base.
And light is a certain meaning and is the elated work of the Sony group. For example, CD and DVD (technology) have been developed using laser. The new application different from an optical disc is shortly considered using it. Say that media consist of a disk there (optical network). The topology of a computer is changed by it.
- A transistor is to the world of a high polymer.
[Q] Does the change of computing come?
[Mr. Kutaragi] After man gains a computer, time is still shallow. 32 years after what 50 year from ENIAC, and a microprocessor. If a computer comes out at last, it thinks that it is interesting at i4004, it thinks that it is interesting at VAX11 and that is regarded as wanting personally a long time ago when I was a student, everybody has a more highly efficient computer now. Therefore, a surprising thing happens only by maintaining the speed for ten more years.
[Q] Therefore, exponential increase of a transistor requires progress of a computer. Does this think that it still continues?
[Mr. Kutaragi] Since a transistor count increases linearly and goes by log (logarithm), it changes and this world goes by uncanny vigor. This pace will continue.
Probably it will go to the world of quantum mechanics at least in ten more years. Now, even an atomic film and a single molecule are visible and the tunnel effect is beginning to become a problem.
I think that the point becomes a high polymer from there. Of course, suddenly, it becomes a high polymer like protein, it does not divide and appear in it, and there is a step. In order to carry out the simulation of the protein and to compound it, a supercomputer is required for interesting one, and it is accelerating protein (utilization) with silicone after all. The essence of a computer is the compressor of a time-axis. A computer accelerates progress of a computer. For example, although the semiconductor of 200 million transistors is not made in the time of a paper streamer a long time ago, if it is the present computer, the design of such a semiconductor can be performed.
I think that a still uncannier supercomputer can be done from now on. Still, although it goes by grid computing etc. so so (performance), it is a man machine interface like a game machine, and the power of a supercomputer can be used in the future. If it does so, it comes to be able to perform more creative research and a more creative design. It is exciting when it thinks.
(September 2, 2003)
Kutaragi Ken: "The Semiconductor Strategy Of Sony"
- Thread starter ChryZ
- Start date
Not Intel, but Mr Intel.
....
....
Joe DeFuria
Legend
Paul said:
Exactly because PS3 is not around the corner, and they need a new round of marketing fluff to try and sustain some form of hype machine until the day when they can actually talk about PS3.
They are talking about PS3. With it's chips going into production 2004 it's litterally is around the corner.(Around a year and a half) The marketing hype will come when actual PS3 specs come out not unlike PS2, I suspect everyone will be talking for atelast 18months prior to a USA launch. Just like with PS2. AKA around mid 2004, which is when Cell should be in limited production with a mass production around late summer early fall.
It was one big interview and watch.impress.co.jp split it up, releasing a segment once a day. Unfortunately there is no english equivalent of that site, so the chances to get those interview segments in a proper english translation are very slim. Too bad, that there are not even well translated quotes on 3rd party news sites. After this serie is done, I am sure it's back to radio silence ...
Re: Kutaragi Ken: "The Semiconductor Strategy Of Sony&a
I mean, if you're fluent in Japanese, then for all it's worth provide us with a meaningful summary. If you aren't, a link to the Japanese original would be fine.
cu
incurable
And who forces you to post an obviously flawed translation?ChryZ said:Nobody is forcing you to read it, stop bitching ...incurable said:Stop doing this to my brain!ChryZ said:
I mean, if you're fluent in Japanese, then for all it's worth provide us with a meaningful summary. If you aren't, a link to the Japanese original would be fine.
cu
incurable
Who "forces" anyone to find anything? Some people might like a marginal translation, and some won't. Since you know what the thread is about in one second, and can choose to read it or ignore it afterward (or choose to discuss parts of it or ignore it forever), it's not like this is different from ANY OTHER information being posted.
Re: Kutaragi Ken: "The Semiconductor Strategy Of Sony&a
Come on, the translation is not perfect, but it works. Provide something better or mute yourself
I am just a nice guy, who tries to provide some "still readable" information to the forum, he likes, in order to provide certain people with something to bitch aboutincurable said:
Nope, I am not fluent in Japanese and there is a link pointing to the original japanese article at the end of "obviously flawed translation".incurable said:
Come on, the translation is not perfect, but it works. Provide something better or mute yourself
Joe DeFuria
Legend
Paul said:
Blah...blah...blah...
