More on the PPC 970 MP, 970GX, and mobile G5
Thinksecret reports that Antares will be at 3 Ghz. That apple has recieved a number of protypes of both the 970 MP and GX. They plan to end testing at the end of January which we might see the 970 MP in the Powermacs around spring. TS also says the MP might find its way to Pbooks but, most likely and reports still hold strong that the G5 mobile will go in there instead and top out at 2 Ghz.
http://www.thinksecret.com/news/0412powerpc.html
Quote:
Confidential documents pertaining to the development of IBM's future single-core PowerPC 970GX and dual-core 970MP processors, code-named Antares, obtained by Think Secret, reveal that Apple has already received a number of prototypes of the processor. Apple is also expected to wrap up Process Verification Testing of the 970MP around the end of January, meaning systems featuring the processor could be ready as early as the Spring including, possibly, the long-awaited PowerBook G5. However, sources continue to report that IBM is working to develop other low-power 970-series processors topping out around 2GHz that could find their way into future PowerBooks instead. Antares will debut at 3GHz.
Confidential documents pertaining to the development of IBM's future single-core PowerPC 970GX and dual-core 970MP processors, code-named Antares, obtained by Think Secret, reveal that Apple has already received a number of prototypes of the processor. Apple is also expected to wrap up Process Verification Testing of the 970MP around the end of January, meaning systems featuring the processor could be ready as early as the Spring including, possibly, the long-awaited PowerBook G5. However, sources continue to report that IBM is working to develop other low-power 970-series processors topping out around 2GHz that could find their way into future PowerBooks instead. Antares will debut at 3GHz.
http://www.thinksecret.com/news/0412powerpc.html
Comments
If you go to the Apple online store today and go through the ordering/configuring process for a 12" powerbook and a 12" iBook, you can get the following results:
12" PowerBook
- 1.33GHz G4
- 80GB drive (max)
- 1.2GB RAM (max; 333MHz)
- NVIDIA GeForce (64MB video RAM)
- with DVI output port
- Airport extremen built-in
- Bluetooth built-in
- Combo drive (you can get a SuperDrive)
- PRICE: $2,499
12" iBook
- 1.2GHz G4
- 80GB drive (max)
- 1.2GB RAM (max; 266MHz)
- ATI Mobility Radeon 9200 (32MB video RAM)
- no DVI output port (only VGA)
- Airport extreme built-in
- Bluetooth ($50 option)
- Combo drive (SuperDrive not available in 12")
- PRICE:$1,700
While some of the PB spect are better (processor & memory speed and graphics), the basic box is the same and I can't see why anyone considering a 12" PowerBook would not be nearly equally satisfied with a 12" iBook.
And while a speed bump would be nice for the 12" PB, I personally think that that won't cut it.
It will have been a full two years since the basic PB as we now know it came to market. It's time for some WOW! factor in the PB sphere once again. The iMac got it this year. The G5 PowerMacs are solid, steady and still relatively new where speed-bumps are what matters.
I just took the 12" as an example and with th exception of the 17", the other 2 PB vs. iB models are just too close. I'm way happy to see where the iBooks have gone as they are emerging as a really solid platform for students and teachers at all levels. But...the PowerBook line needs to make a move and my money says it's to the G5 space.
Just a guess from a happy 12" PB user.
Originally posted by onlooker
What are the chances that they have had these, and they may show them at MWSF?
Have them, pretty high; show them at MWSF, pretty slim. That's at least what I think.
Originally posted by PB
Have them, pretty high; show them at MWSF, pretty slim. That's at least what I think.
I agree. I'd expect another g4 update, 1.7-1.8ghz. Cross fingers for 2ghz
IBM, AMD claim a better way to strain silicon
... Called "Dual Stress Liners," or DSL, the technique will ideally eliminate much of the complexity involved with strained silicon. Chips containing DSL are quietly already being sold, but both companies will start to use it in chips built on their 90-nanometer production processes in the first quarter, which will lead to greater proliferation.
According to early data, DSL improves transistor performance in its chips by 24 percent, but incorporating it does not decrease the number of good chips that come out of a wafer, meaning that it should be relatively inexpensive to adopt.
or
IBM, AMD boost chip speed
Companies use technology called strained silicon to improve transistor performance by 12%.
... Both IBM (Research) and AMD (Research) said they will begin shipping microprocessors that use the technology, called dual-stress strained silicon, early next year. The advance costs little to install in factories and boosts transistor performance by about 12 percent, compared with current straining technologies, they said.
So, doing some speculation, take a 1.6 GHz 970fx, manufacture it on this new process, out comes a 1.8 to 2 GHz 970fx with the same power consumption as a 1.6 GHz 970fx. The probability of a Powerbook G5 in Q1 05 has just increased a bit.
I guess this also means that IBM is abandoning low-k dielectrics?
Originally posted by THT
...<snip>
I guess this also means that IBM is abandoning low-k dielectrics?
Why?
Would implementing DSL/dual-stress strained silcon prevent the use of low-k dielectrics?
If I'm reading these articles right, why would one say,"According to early data, DSL improves transistor performance in its chips by 24 percent." while the other says,"Companies use technology called strained silicon to improve transistor performance by 12%."
How does a 12% improvement in transistor performance equate to a 24% increase in chip performance? Unless of course I completely misunderstood and these articles are discussing 2 different technologies.
Originally posted by rickag
Why?
Would implementing DSL/dual-stress strained silcon prevent the use of low-k dielectrics?
If I'm reading these articles right, why would one say,"According to early data, DSL improves transistor performance in its chips by 24 percent." while the other says,"Companies use technology called strained silicon to improve transistor performance by 12%."
Isn't strained silicon what is available today (SSOI)? I haven't read the article yet (I have to leave to go take a final in like 10 minutes), but perhaps they were comparing todays tech with tomorrows (12% vs 24%)?
i know they're very unreliable, but if that has any truth to it... be interesting to read the update once they post it at least.
Originally posted by drsuse
i know they're very unreliable...
So unreliable that it's not funny sometimes.
Originally posted by rickag
Why? Would implementing DSL/dual-stress strained silcon prevent the use of low-k dielectrics?
I don't know. There has been no news for low-k at 90 nm for awhile now. With this new technique giving about the same improvement, the investment in low-k may not be worth it to IBM anymore.
If I'm reading these articles right, why would one say,"According to early data, DSL improves transistor performance in its chips by 24 percent." while the other says,"Companies use technology called strained silicon to improve transistor performance by 12%."
Who knows what the two performance numbers are comparing to.
Originally posted by PB
So unreliable that it's not funny sometimes.
as long as you take it as pure speculation, it's no different from surfing through rumours forums. i agree that they much of what they say is totally out there, but at least sometimes it's interesting.
Originally posted by THT
I don't know. There has been no news for low-k at 90 nm for awhile now. With this new technique giving about the same improvement, the investment in low-k may not be worth it to IBM anymore.
I guess Chipworks may be waiting a little longer than "soon". Didn't they say they expected to see a version of the 970FX using a low-k dialectric "soon" and that IBM was qualifying the low-k dialectric process at the time of their report back in September?
Just wondering if the 970 would reach 3.0GHz using both this new DSL and a low-k dialectric.
2.5 x 1.24(24% for DSL) = 3.1GHz
3.1 x 1.20(20% for low-k) = 3.72GHz
or maybe in order of introduction(if the Chipworks crystal ball is to be believed):
2.5 x 1.2(20% for low-k) = 3.0GHz
3.0 x 1.24(24% for DSL) = 3.72GHz
either way I'd be pleased.
Originally posted by THT
Who knows what the two performance numbers are comparing to.
Certainly not me.
Originally posted by drsuse
mosr is saying: "Next Update: Monday (In Progress Now: "IBM to open second G5 fab for Apple")"
i know they're very unreliable, but if that has any truth to it... be interesting to read the update once they post it at least.
Actually I find that they are remarkably reliable... at being unreliable.
check out the register article. it's pretty informative.
Anyway, IBM says it'll be shipping these chips in the first half of 2005, which, to me, means PBG5. I'm guessing they've had the technology, or at least its specs, around for long enough for Apple to design a powerbook around it, and Apple's ready to ship as soon as they get the chips.
But, of course, I know nothing about this, and am merely speculating to fulfil my desire for a new Powerbook.
Anybody have an eye on quantities in the field? Apple likes to let these fall to near-zero before releasing new models of powerbooks.