Oita 300mm fab complete

[nondescript]

This might be beating a dead horse, but, I thought at least a few people would be interested in this.

http://www.eetimes.com/semi/news/OEG20040220S0025

Toshiba ready for 65 nm fabrication in first half 2005

By Yoshiko Hara

EE Times
February 20, 2004 (10:23 a.m. ET)

TOKYO — Toshiba Corp. has completed its 300mm wafer fab in Oita, Kyushu, and intends to lead the industry with fabrication of chips on a 65 nm process line in the first half of next year.
The 65-nm process will be for volume production. Part of the new fab's capacity will be allocated to the virtual joint venture fab with Sony Corp. Toshiba and Sony respectively invested 42 billion yen (about $389 million) for the virtual fab.

Toshiba and Sony have been jointly developing process technology, starting with the CMOS4 90 nm node. The CMOS5 65 nm node development is scheduled to end by the end of this fiscal year (March 31).

The virtual fab arrangement is the same as the current for fabricating the Emotion Engine for PlayStation2. That joint venture, Oita TS Semiconductor, between Toshiba and Sony is a virtual fab that utilizes Toshiba's 200 mm Oita fab facility.

Early this month, Toshiba and Sony had agreed to extend their process development collaboration to the CMOS6 45 nm node, which will begin immediately after CMOS5 development is completed next month. CMOS6 development is being targeted for completion by the end of 2005 and will also use the new 300-mm facility, a Toshiba spokesman said.

Toshiba's process migration started with volume production using its 90 nm process named CMOS4 this fall followed by fabrication using the 65 nm CMOS5 process next year.

Overall during five years Toshiba will have invested a total of 200 billion yen (about $1.9 billion) into the new 300mm wafer fab. Its pojected capacity is expected to be 12,500 wafers a month in 2007.

To summarize:
- CMOS5 (65nm) development will be completed by March 31st.
- Volume Fabrication of CMOS5 in early 2005
- CMOS6 (45nm) development targeted for completion at end of 2005.

(*Waiting for the obligatory "well what about PSX2OAC blah blah..."*)

[Edit: completing => completion. Typo on my part. Apologies everyone]

 

[jvd]

CMOS6 (45nm) development targeted for completing at end of 2005

What do they mean by this ?


Anyway all the rest of the info we have known. Earliest they can launch is middle of 2005 to the end of 2005 depending on what the volume actually is of the launch.

 

[Deadmeat]

..

OK, pack 300 million transistors into a die size of 86 mm2 and we will believe this press release.

Or else this is just another 130 nm chip fabricated on what is supposed to be a 65 nm process.

 

[Megadrive1988]

Cell OS development and PS3 game development (no doubt very complex) will probably ensure PS3 does not launch until March 2006, in Japan, at the soonest. methinks anyway. 8)

 

[cthellis42]

CMOS6 (45nm) development targeted for completing at end of 2005

What do they mean by this ?

I would assume testing the first bits of tech off the line to know that the equipment is set up fine and functioning. Same as they mean regarding CMOS5 on March 31st. If it's something different, then I guess we find out on March 31st, eh? ^_^

 

[Qroach]

Megadrive,

Cell OS development and PS3 game development (no doubt very complex) will probably ensure PS3 does not launch until March 2006, in Japan, at the soonest. methinks anyway.

Yes i agree... and it echos what I was saying about 2006 launches previously.

 

[ChryZ]

Megadrive,

Cell OS development and PS3 game development (no doubt very complex) will probably ensure PS3 does not launch until March 2006, in Japan, at the soonest. methinks anyway.

Yes i agree... and it echos what I was saying about 2006 launches previously.

Quick question: on what is that assumption based? Wouldn't it be possible, that the CELL OS or platform support (compiler, libs, etc) are developed along the hardware design/implementation. There is a lot you could do with emulators and even if the project is ending in 2006, it doesn't have to mean, that software will be finished at that point. Software could be finished way before 2006 and there would be still alot to do until 2006 like, dev support, compiler optimization, etc ... no?

 

[Qroach]

Chryz,

Quick question: on what is that assumption based?

It's actually not an assumption. This is information I was given by devs woking on the next gen platfomrs.

Wouldn't it be possible, that the CELL OS or platform support (compiler, libs, etc) are developed along the hardware design/implementation. There is a lot you could do with emulators...

Emulators? Emulators are typically complete crap to work with. There was even PS2 emulators provided to developers at one time and those were simply horrible and incomplete. Even the xbox emulators were useful to an extent, but you really needed time with the real hardware to do anythign good.

On top of that there's always things you just can't emulate properly (like bandwidth). From my experience, the real development can't commence until the developers have had the real hardware for decent amount of time.

... and even if the project is ending in 2006, it doesn't have to mean, that software will be finished at that point. Software could be finished way before 2006 and there would be still alot to do until 2006 like, dev support, compiler optimization, etc ... no?

That's right, even if they start mass producing the hardware by mid 2005, I still doubt the games will be ready by that point, and you just can't lauch a console these days with no game support ( well Nintendo tried with the N64 and I don't think that was a good situation. I can recall waiting for a long time before a number of new first party titles started showing up.).

 

[Vince]

Quick question: on what is that assumption based?

Absolutly nothing. There have been mention of software development happening concurrently with hardware development comming out of STI Austin since early 2002.

Word is that the open-source Eclipse IDE has been utilized to an extent with the Cell architecture since 2002, with a proof of concept shown in late 2002 to IBM executives and Kunitake Ando. And I'm sure this is a tip-of-the-iceberg scenario where there's much that we don't hear about due to being on the outside.

This is information I was given by devs woking on the next gen platfomrs.

This is from a member of IBM's Extreme Blue program who formerly worked at STI-Austin.

 

[Qroach]

Absolutly nothing. There have been mention of software development happening concurrently with hardware development comming out of STI Austin since early 2002.

Excuse me? just what does this hve to do with the date internal SCEA develoeprs are targetting? You can keep denying 2006, but that's a date I've been hearing for a few months now. yet all you do is come back with information about IBM executives, fab information and things like that. What about the actual game development?

This is from a member of IBM's Extreme Blue program who formerly worked at STI-Austin.

Once again, just what does this have to do with the games being developed on the platform vince? Unless sony plans to launch without new game software, we're not going to see a launch in the fall of 2005 (I'm talking outside of japan btw, just like the last time I brought up this date.).

 

[Vince]

Excuse me? just what does this hve to do with the date internal SCEA develoeprs are targetting?

Whoa, where did SCEA developers enter this conversation? I saw the following:

Wouldn't it be possible, that the CELL OS or platform support (compiler, libs, etc) are developed along the hardware design/implementation."

I'm responding to the fact that there is concurrent software development ongoing and has been since 2002. And that your assumption as per the level of concurrent software development is hindged on a game developer... I mean, don't make me use the eye roll. Nothing more or nothing less was intended.

You can keep denying 2006, but that's a date I've been hearing for a few months now. yet all you do is come back with information about IBM executives, fab information and things like that. What about the actual game development?

Care to point me in the direction of said comment? I came back saying there has been work on software development since 2002. Where did I state the phrase "2006" or "2005" or any launch date in this thread? I purposly didn't state anything about an actual launch date.

Once again, just what does this have to do with the games being developed on the platform vince?

Well, I assume the development tools for a given platform are for eventual... development? And that a comment based in information from a member of the team who worked on the IDE would affirm that work has been underway for some time, lending credence to ChryZ comments.

Furthermore, you don't know what level STI's development has progressed to and how it compares with Microsoft's. I mean, where's Microsoft's dev kits? How many are out there? And when do they intend to launch by?

 

[Qroach]

SCEA developers entered this conversation where? I see the following

When I said devs (as in game developers working on next gen platforms). just like the last time you got all up in arms about my saying we wouldn't see the new consoles until 2006.

I'm responding to the fact that there is concurrent software development ongoing and has been since 2002. And that your assumption as per the level of concurrent software development is hindged on a game developer... I mean, don't make me use the eye roll. Nothing more or nothing less.

Like I said before, it's not an assumption, it's a date SCEA and MS are supposedly targetting. I didn't say it was hinged on a game developer, I said devs, as in more than one developer. One doing an internal MS game for launch, and they othe rdoing an internal PS3 game for launch. I also didn't argue that the above was possible. You can keep you're rolling eyes face, as I think you're arguing this from the wrong point of view.

Care to point me in the direction of said comment? I came back saying there has been work on software development since 2002. Where did I state the phrase "2006" or "2005" or any launch date in this thread?

Perhaps I took what you wrote wrong, were you in some way agreeing with me when I said 2006? Or do you still stand by the fall of 2005 launch comments you made the last time I brought up that date? Please do explain...

Furthermore, you don't know what level STI's development has progressed to and how it compares with Microsoft's. I mean, where's Microsoft's dev kits? How many are out there? And when do they intend to launch by?

Once again, you keep saying the level STI's development has progressed. I'm not arguing about that. All I know is that Sony hasn't released ANY dev kits just yet, which is one factor in gettnig games ready for the fall of 2005. MS also hasn't released any dev kits as of yet. I'm basing my comments off the date SCEA internal developers are looking at for launch. I could care less about any PR talk or annoucments about fabs being ready as it will all come down to the games being ready to be launched. Until I hear differently from these people working on both platforms, then that's the date I'll believe. It's possible it could change, but right now I haven't heard that.

 

[Vince]

Q, stop arguing. We're talking about different things and I refuse to just argue for the sake of arguing again. You're using the current state of tools in developers hands as a metric of sorts to gauge when the consoles can be logically introduced.

I'm not debating this, I never questioned your 2006 statement in this thread and don't drag what we've previously debated into every topic it can tangetially be related to. I'm merely providing ChryZ's comment on concurrent tool/software development some credence since I know that it's been occuring internally at STI-Austin wrt the Eclipse IDE since early 2002 and that proof-of-concepts were shown and that work has continued - nothing more, nothing less.

PS. My comments on this topic concerning Eclipse are about as PR-talk as you quoting your own sources, I never question you and as for PR, when was the last time you saw a PR concerning this... um... never?

 

[Fafalada]

Once again, just what does this have to do with the games being developed on the platform vince? Unless sony plans to launch without new game software, we're not going to see a launch in the fall of 2005 (I'm talking outside of japan btw, just like the last time I brought up this date.).

PS2 launched a bit less then one year after first devkits made appearance.
And ~6months after the first DECENT devkits were out.
And ~4months after first full speed devkits were out.

I don't think it would be a good idea to compare notes on dates for That on a forum though, assuming you've been hearing those too.

Anyway, Eclipse PR should be good news (except for haters), not only in terms of early software development tests but also in hope that we won't end up with nothing but a buggy piece of crap Cygnus put together in two weeks and a command line debugger to start working with.
Although until I see things with my own eyes I'll remain skeptic I don't trust Japanese love of straight C either. :?