IBM Processors

[arn]

Jeremy West wrote in with the following:

IBM just announced (http://story.news.yahoo.com/news?tmpl=story&u=/nm/20020225/bs_nm/tech_ibm_dc_21) they have functional 110 Gigahertz chips that will be ready for market at the end of this year. No specific mention of PPC, but very interesing indeed. Imagine a 100 Gigahertz Mac by MWSF 2003.

Apparently the new chip is based on "silicon germanium chip-making":

Silicon germanium also includes layers of germanium, a substance that is like silicon but which makes a transistor that can switch faster and perform better.

We'll see....

[Choppaface]

damn
:eek:

[ThorPrime]

as much as this seem fabulous but I have a feeling this technology will spark no big improvements for anyone anytime soon.

[mischief]

Lemme see:

110,000,000,000 Hz chip on a 333,000,000 Hz Bus. Bus speed is 3% of processor speed at that hypothetical, which is quite fast for a bus. Which means that 97% of the time the processor is blue-balled. Why bother? :rolleyes:

Sure it's fast, but how do you keep up with it? Give you machine Meth?

[Biggles]

yeah! let's all pay $50,000 for a computer that is almost "too" powerful

:)

[rekras]

wonder what the price will be on this? i hope, but it's doubtful that this chip will ship on macs...

[Quark]

Before everyone get's too far ahead of themselves, understand that they are not talking about producing ANY CPUs this year.

They stated that they are making some microchips and say nothing about making any sort of Central Processing Units (CPU) for this year.

The technology is mind boggling. We are getting closer to the Star Trek: TNG era.

If they can mass produce these babies, there is no reason to think that the price will be that high by the end of next year.

I wouldn't be surprised to see a PPC system, close to 100GHz under $5,000, in Q4 2003.

Take Care!

[jaykk]

Its not a CPU, its the fasterst communications chip
http://story.news.yahoo.com/news?tmpl=story&cid=581&u=/nm/20020225/tc_nm/tech_ibm_dc_22

[Gelfin]

"Too powerful?" I'm recognizing words here, but you're losing me somewhere on the idea.

Eh, it's a good thing IBM's raising the bar like this. You'll probably need a 100GHz CPU for the next version of Windows anyway.

[Yamson]

Don't expect to see this in a Mac or any computer any time soon. This is a COMMUNICATIONS chip designed for use in high-speed network routers. The circuitry is much different than that of a desktop computer CPU. There are many more issues involved with getting a CPU up to those speeds simply because the wiring is that much more complicated.

Also, if you read the story you'll notice that IBM developed this chip to replace their 80 GHz version also built on the same silicon-germanium technology... you don't see that in desktops, now do you? And that chip has been out for a while (I'm assuming...). Be realistic. According to Moore's Law -- to which the last 30 years have conformed to fairly well -- we should be seeing 110 GHz chips in desktops in MAYBE 2010 or so. We're not going to suddenly jump 3000% in clock speeds in one year.

What is worth noting, however, is that this is a new manufacturing technology that has the POTENTIAL to take desktop computers to faster speeds. Silicon is beginning to reach its limits, so engineers are and have been rapidly developing new ways to extend Moore's law. This is an excellent possibilty, but right now there are simply too many issues to overcome to make this a consumer-level desktop computer CPU in the several years.

[Biggles]

Originally posted by Gelfin
"Too powerful?" I'm recognizing words here, but you're losing me somewhere on the idea.

eh, I'm just exaggerating a bit. I just don't see the need for something with that much power yet. I mean it would be great if it was a reasonable price, but i doubt that'll happen for a very long time.

Maybe if they used the same technology to produce a 25-50 ghz machine i'd get a little more hopeful of it ever coming to market anytime soon

[dualburn001]

These chips aren't going to be in home computers for at least 5 years. The demand for a chip like this now is limited. It cost them millions to produce it and theyre gonna charge a good amount for it to.
Who's gonna buy it? IBM will put it in it's main servers and then ship it to Microsoft and Sun.
Doesn't mkae me feel very happy.

Dualburn001

[Gelfin]

Yeah, I know it's not realistic anytime soon, but I do get this silly mental image of Steve standing on the stage at the MacWorld keynote, saying, "Now if you look here, we have a top-of-the-line Windows machine, and we'll start rendering this poster-sized FraxFlame image using KPT5 in Photoshop."

[Windows machine begins grinding away. You see a smidge of progress bar appear at the left end of the meter.]

"Now let's see it on the new PowerMac. Here we've got Photoshop set up."
*click* *bloop!*
"Oh, sorry, did you blink? Wanna see it again?"
*click* *bloop!*

[Repeats two or three more times, grinning like a fiend.]

[Windows machine displays another two pixel-widths of progress bar.]

[buffsldr]

And just like that an Apple buyout seems more likely.

If IBM does buy Apple, I guess the pee-cee world can hire Apples "MegaHertz Myth" Public Relations Department.

[Xapplimatic]

Just the view of a tech here..

Perhaps a bad comparison.. but if a Germanium version of a transistor (which have been out for years and years!) costs several fold what the standard silicon version costs, that probably means IF they could make a PPC CPU out of Germanium, the cost would be several fold over the cost of silicon ones.. too expensive for average joes. The advantage of Germanium is that yes, it's faster, but more than that, it has a much wider thermal range meaning that it's perfect for laptops! (= less cooling needed).. and yes, we won't see germanium CPUs any time soon.. I don't recall any manufacturers mentioning that germanium was on their to-do list for CPUs... but hey, they can make a lightning fast 555 timing chip!

[rubikcube]

The problem with having a 110 ghz chip is that you have to do a lot more than increase the clock rate. The memory, north and southbridge units, and pretty much everything else on the board would have to be sped up to incredibly high speeds as well. Heck, even a titanium geforce4 would slow that thing down. Every part on the computer would have to be made with the new manufacturing process in order to benefit. That isn't going to happen any time soon. I would guess maybe 5 years down the road we will start seeing this technology emerge, but not in 110 ghz models. I would say that we might be seeing 10-20ghz. The reason for this is that the different architectures are going to run at different speeds with the new process and no one really knows what is going on in that area.

RubikCube

In the meantime, i hope to get my new dual 1ghz within the next 6 weeks. Hopefully it won't be overshadowed by 110ghz powerpc's by that time.

[littlejim]

Originally posted by buffsldr
And just like that an Apple buyout seems more likely.

If IBM does buy Apple, I guess the pee-cee world can hire Apples "MegaHertz Myth" Public Relations Department.
But surely that'll be renamed the GigaHertz Myth department?


LittleJim
----------

He's fallen in the water

[ibjoshua]

where do some of you guys find the time to write such gar-bo-ley!
dream on.

wouldn't you just like a 2 gig mac?

[mymemory]

At that speed I would loose what is left of my hair...:eek:

[mischief]

It's prohibitively expensive as a whole chip just in materials cost. Is there a way to do a Germanium-on-Silicon hybrid chip? Say you just fused a section of Germanium on the wafer per chip area. Cut the Altivec on the Germanium, cut the bulk of the processor on Silicon.

If nothing else it WOULD be terribly quick.:cool:

[motive8]

I always find that my Germaniums grow best in a light topsoil and compost mix. Of course, the best compost is made from a combination of shredded monochrome CRT's mixed with all the RAM i have that no longer works in the machines i keep upgrading to.
Companion planting old 520c laptops are recommended. The flies are attracted to them instead of the Germaniums.

[StuPid QPid]

Originally posted by motive8
I always find that my Germaniums grow best in a light topsoil and compost mix. Of course, the best compost is made from a combination of shredded monochrome CRT's mixed with all the RAM i have that no longer works in the machines i keep upgrading to.
Companion planting old 520c laptops are recommended. The flies are attracted to them instead of the Germaniums.

Haha, LOL, but oh dear!
That's definitely worth the groan of the week. Or should that be grown of the week ;)

[Rocketman]

Originally posted by Yamson
Don't expect to see this in a Mac or any computer any time soon. This is a COMMUNICATIONS chip designed for use in high-speed network routers. The circuitry is much different than that of a desktop computer CPU. . . .

What I find possible say a year after they introduce this chip to the Cisco market, is a coprocessor. The whole purpose of this type of chip is wide bandwidth and alot of little transactions.

A CISC or even a RISC chip is more monotonic.

However as a CPU upgrade it would replace several MP systems in software and hardware and make an outstanding server CPU.

Mac G6 server 4U 4 processor = 44 G5's

Now one valid point was what one even do with that much computer capacity?

Image storage and serving including complete movies, interactive backgrounds, and systems monitoring services, navigation and remote, wireless networking. Wideband.

We're not all that far away from reality.

Rocketman

[stoid]

Yeah, windows can now crash in .00032 seconds!!! Mabye its .000319999, I'm not sure :D :D

[littlejim]

Originally posted by stoid
Yeah, windows can now crash in .00032 seconds!!! Mabye its .000319999, I'm not sure :D :D
Like it ... like it ...!!!

[AmbitiousLemon]

ok im getting way ahead of myself here, but where do we go next? now that we are using Ge we have run out of metaloids in the carbon family.

btw we have been using Ge in semiconductos for awhile now and it hasnt found its way into desktops yet.

[motive8]

Squibb Labs has just introduced a new hyperconductive "antacid chip" that essentially clears up render wait states and upset stomach.
Chairman at the company, in conjunction with heavy input from Johnson and Johnson said that the chip comes in fruit and chalk flavors and should be taken with meals for maximum throughput. Katie Couric is slated for spokesperson, according to the company.

[Scottgfx]

I used to work at a TV station that had an old RCA video tape machine that dated back to the mid 1960's. I always heard the engineers make remarks about the germanium transitors that were used, that they were unreliable. Here we are 40 years later, and everything old is new again.

If you want to see a picture of the RCA VTR, it's here...
http://www.lionlmb.org/quad/tr70b_1.jpg

[Mike Gagne]

who needs it? and for what? web browsing? Photoshop?......oh wait.....faster email..... yeah, thats it!:D

[StuPid QPid]

Originally posted by AmbitiousLemon
ok im getting way ahead of myself here, but where do we go next? now that we are using Ge we have run out of metaloids in the carbon family.

btw we have been using Ge in semiconductos for awhile now and it hasnt found its way into desktops yet.

If you want to know where computing is heading then read this article from the Scientific American...

http://www.sciam.com/1998/0698issue/0698gershenfeld.html

Looks like transistors will eventually go out of fashion...may be...

[unclepain]

Originally posted by Mike Gagne
who needs it? and for what? web browsing? Photoshop?......oh wait.....faster email..... yeah, thats it!:D

Mike, you obviously do not do any multimedia creation on a Mac, and you certainly have never tried to do any video compression. On my G4 400, It can take HOURS to compress a 6 minute video using Cleaner. Multimedia producers wil NEVER have enough power to satisfy them. Even if some of this chip technology is possible in the PowerPC chips, it could take the speeds into areas we only dream about, but even then I will still have to wait on my Mac to finish rendering my Final Cut Pro project. So just for the record, most of the people who use their macs for anything creative or multimedia centric will gobble up every bit of power they can get and still leave the table hungry.

[Kela]

Yo Freaks, umm this is like a hoax thread incase you havent noticed. Look up germanium and youll see what i mean. ( no not only the fact tht its a flower)
:rolleyes: :rolleyes: :rolleyes:

[Pants]

Originally posted by Kela
Yo Freaks, umm this is like a hoax thread incase you havent noticed. Look up germanium and youll see what i mean. ( no not only the fact tht its a flower)
:rolleyes: :rolleyes: :rolleyes:


lol - plonker - the flower is gerAnium, whilst the element is greManium (element number 32 - discovered in 1886 by Clemens Winkler (a forefather of the Fonz no doubt) - and named after his place of birth, Germany).

it is however a bit of a fake thread though - i mean, this is a logic gate not a processor. Physics demands that 'traditional' processors wont/cant run at these speeds.....

[MacNut]

xan u imagine how many fans that thing would need too keep cool

[lewdvig]

This post was taken completely out of context. This is communication technology that has almost no immediate relevance to PC semiconductors. Every other cut and paste site I read at elast mentioned this fact.

IBM previous best was 80ghz. Do you see any 80ghz PC chips?

A little context next time. You don't want all the dumb sheep who read your site for valuable hardware rumours to freak do you?

How much harder is it to cut and paste all the relevant information?

Stupid monkey.

[Xapplimatic]

I am guessing that semi conductors will really take off when they move to advanced composited carbon-synthesized diamond lattice chips. Something along the lines of what NUKE Scientific was working on...

[StuPid QPid]

Originally posted by Pants


lol - plonker - the flower is gerAnium, whilst the element is greManium (element number 32 - discovered in 1886 by Clemens Winkler (a forefather of the Fonz no doubt) - and named after his place of birth, Germany).

it is however a bit of a fake thread though - i mean, this is a logic gate not a processor. Physics demands that 'traditional' processors wont/cant run at these speeds.....

:eek: Oh double Pants on a stick...
Check your spelling when trying to correct others'...
I presume you meant 'germanium' not 'gremanium'.

Unless a gremlin has got into the works, or should that be a 'germlin'?

[pglaskowsky]

(Some of you may remember me as Dennis Sellers' primary analyst resource for his series of articles entitled "The Myth of Megahertz" at http://maccentral.macworld.com/news/0104/09.megahertz.shtml )

IBM's recent announcement involves a simple "ring oscillator", a circuit consisting of just a few transistors that generate a clock signal. This original high-frequency signal is then divided down into a stable lower-frequency signal that can be used by other chips.

The base region of the transistors used in this circuit are made from a silicon-germanium (SiGe) alloy with a very low percentage of germanium. The germanium increases electron mobility in the silicon, allowing circuits to operate substantially more quickly.

The SiGe layer in the chip is first made from pure silicon, exactly as would be done for any other chip. Germanium atoms are then implanted into the layer. The electrical properties of this alloy depend on the exact proportion of germanium, and it is very, very difficult for IBM to get this proportion right for very many transistors on a chip.

As I wrote for our newsletter, Microprocessor Report, in May of 1996:

The new technique is not as useful for large digital components such as microprocessors, because it improves only bipolar transistors, not the FETs used in CMOS devices. SiGe transistors could be used to extend the life of BiCMOS design, which most vendors have turned away from due to the minimal gains for bipolar devices at geometries below 0.35 microns. IBM Microelectronics has no immediate plans to develop SiGe microprocessors, but the new technology could allow faster, less power hungry versions of current bipolar components.

IBM does another round of publicity for its SiGe process every three years or so. Every time, news services and end users leap to the conclusion that faster microprocessors are on the way. I don't know if this is what IBM wants people to think, but I wish they'd be more careful.

There are other silicon vendors, such as Motorola, using similar techniques to build high-speed transistors for communications devices, but they all have the same limitations. None of these methods can be used to create 100GHz microprocessors.

. png

Peter N. Glaskowsky
Senior Editor, Microprocessor Report
Columnist, Electronic Business magazine

[rastalin94]

The big deal with germanium is the simple fact that electrons travel faster in it than in Silicon. So you basically get a faster chip just by switching substrate. Of course it is more expensive and harder to work with, but that is why it is usually used in military applications.