Slashdot Mirror
Intel Promises A Cool Billion (Transistors)
NevDull writes: "CNN is reporting that Intel has announced new semiconductor packaging which will lead to CPUs with a billion transistors running at 20GHz within 6 years. Yummy!" The advance here is removing the balls of solder between the chip's packaging and the microprocessor core, which leaves room for more transistors (or a thinner package). Like it says, though, this is years away from your pocket Cray.
And as usual, Motorola and IBM will develop this technology first and promise chips as fast as 25 GHz. But in reality, the first IBM/Moto chips will come at 12 GHz at about the time that Intel releases their 20 GHz chips.
Apple will introduce the chips in it's new iHyperMac which is the size of a quarter with a holographic display but they will be running downclocked to 10 GHz for marketing reasons.
Intel already removed the balls from a processor... it was called "Celeron".
to create an OS so bloated that you need a 20 Ghz chip to run it. ::Sigh::
If I can't see it in Lynx I'm not interested.
Except that clock speed is becoming a useless benchmark. At what point do we realize that Intel's 20 Ghz machine and AMD's 12 Ghz machine have an unnoticable speed difference? If they were talking about a pocket cray as suggested, yes, I guess there is a use for it. They're not talking about supercomputing, they're talking about Pentium 4's!!! At 20 Ghz you'd have to slow the thing down to play Diablo!!
There is no reasonable defense against an idiot with an agenda
:wq
...or are you just happy to see me?
With that many transistors running at X GHz, will Intel be providing a fusion plant to run this thing? With some small duct works, you could even use it to heat your house!
Seriously, though, who knows what other kind of breakthroughs are going to be made that may obsolete this? There are advances being made in optical and even quantum computing all the time. Someone is even working on a biological hard drive using DNA strands!
My 1/50 $ (US)
Vote monkeys into Congress. They are cheaper and more trustworthy.
If only it were so - but looking back on the development later new cpu generations I'd bet it's going to be a HOT billion...
requirements for cooling of new cpus are becoming ever more demanding, just the cpu can burn in excess of 50W in existing cpus.
So, for my own requirements I'm more interested in getting an (energy) efficient system that can run with as few fans and noise as possible - it's practically impossible nowerdays to get a box where CPU power is NOT sufficient for even the most demanding tasks. The downside is that most modern boxes seem to be best suited for running flight simulators - at least they sound like jet engines.
Also if you're working in an office with a lot of computers, the heat output of computers and monitors can be VERY noticyble, esp. in summer. (No, there's no aircondition in my office).
Hopefully the new technology will not only be used to reate overpowerd energy hogs but also find its way into (mobile?) processors - same cpu core as existing cpu, but smaller layout, lower core voltage and correspondlingly much cooler/more silent.
you have moved your mouse, please reboot to make this change take effect
A cray in your pocket? It would better have some good cooling then, or you'd get some nice pick up lines.
Him: Hey, baby, you make me feel *hot*.
Her: Just take that cray out of your pants, geek boy.
I intend to live forever, so far so good.
At the least, we have *Crays* on our desks...
"I'm sure we'd be able to walk around a cray in our pocket :-)"
But if Intel's current crop is any indication, your bits would burst into flames unless you had Freon cooled undershorts.
âoeWho knew something as harmless as willful ignorance could end up having real consequences?â
Now we will have a 20GHz processor which will tell us that 4+4 is 7.9999999999 approximately :-D
Well, it would be a Pocket SGI, wouldn't it? Palm OS with 1024 bit math! I can just imagine the sales pitch:
"How many times have you been sitting on the bus, in need of some quick supercomputing? You're 20 minutes from the office and you just need to solve Planck's Theorem RIGHT NOW. That's why you need the Palm 2.5e+12!"
The Dopester
"Yes, I'm a Karma Whore, but I'm doing it to pay my way through school."
I have to commend Intel for trying to tackle a problem that is daunting at best. But there are enough problems with existing IC packages that need to be taken care of between now and then. These include:
1. High-speed signal isolation - two wires switching at enormous speeds like 10GHz are going to have effects on other signals in the package. There's enough trouble with this on high-speed multi-gigabit-per-second interfaces and even Rambus' crap TODAY. With signals packed in so close, how are they going to manage this tomorrow when the current memory bus is already at 3.2Gb/s? At 10GHz+, how hungry will the processor be for memory bandwidth? It's a fight between lower-speed highly-parallel signaling for density and higher-speed low-density serial signaling for signal integrity. A smaller package isn't going to help this. A larger package, even with fewer layers, will only aggravate signal coupling.
2. Power delivery and consumption - on some packages, up to 30% of the total connections are for I/O and core power delivery. Making these smaller as Intel proposes will not help matter, considering that switching at 10GHz is going to make power consumption skyrocket. How do they expect to get the power to the chip? People have enough problems today trying to bump their processor voltages up when they attempt to overclock. This is only going to get exponentially more difficult.
3. Die attach and reliability - I know they want to have solderless connections to the package. This is good - currently alpha particles from solder will occasionally cause false switching in memory elements. But with lots of heat cycles from power cycling up and down and questionable assembly yields that are usually tolerant to less than 0.5% from raw die to package. We take for granted the fact that the die will stay attached to the package today. How they will get the reliability to that point is beyond me, even if they've made a "major" stride. How do they account for field failures or age-related failures in a test lab?
4. Substrate material - the package material itself is critical to thermal matching on the board as well as to signal integrity inside the package. At the speeds they propose will the current substrates be sufficient for reducing signal coupling? As usual, material science is again lagging behind the rest, and we need far more research into exotic materials to be able to get fast packages going.
So, to me I think there's going to have to be larger packages with advanced cooling. I'm not going to get too excited. I certainly don't think that Intel will be able to take this course alone. What I forsee happening is to have new committees set up specifically for packaging as there are for IC process technology today. It's too capital and research intensive to be able to get away from having to use committees.
Actually, Moore's law isn't about clock speeds, it's about the number of transistors.
42million x 16 (four doublings) = 672 million.
They're planning on slightly outpacing Moore's law, not lagging behind it.
"Removing the balls of solder between the chip's packaging and the microprocessor core"...
Well, sure, that'd give you lots more room for transistors... It'd also give you a lot more room for defects to creep in. This is functionally no different from expanding the die size to the point where the CPU size is the same. While it might be less expensive than cutting fewer chips per wafer, it does nothing to increase the reliability of the process.
I think this is more of a pricing advance, and you'll see this lowering the cost of existing processor layouts, since you can decrease the die size without affecting the CPU design. But CPU size increases will still result in lower yield.
Imagine.
Just today, Intel announced the release of the Pentium 5 processor. The new processor runs at 50 Ghz. It features a 300-stage pipeline. It will take 2 minutes for each instruction to complete on average. But to optomize, programmers can send 1800 intructions at a time, as long as they have no dependence on each other at all.
This will be key in driving down the cost of computing, as custom logic will always be more expensive than commodity logic.
While I would expect these developments to also obviously drive down the cost of developing custom logic, volume production will always make commodity logic more cost effective.
Eventually the masses will interact with computers by speech and video. Text and keybords will be secondary. Current computers cant quite do this yet, but how much is software versus hardware?
Intel today announced its new 1024-bit (1 kilobit) microprocessor architecture technology. Named the Shiitakeum, Intel's new processor core boasts powerful new technologies which will enable content providers to deliver compelling enterprise solutions.
The Shiitakeum incorporates the following new features:
* SingleAtom technology squeezes one thousand transistors into a single atom.
* The processing pipeline has been broken down into 299,792,458 discreet steps, enabling Intel to remove the internal clock altogether and run the processor at the speed of light. One "cycle" represents the absolute cosmic measure unit of time, and all operations occur in one cycle. (Compete with that, AMD! Bwahaahahahaha!!)
* 24,856 new instructions have been added since the previous model, bringing the new total to over 72 trillion instructions. The entire UNIX operating system can be programmed in one instruction!
* RAM has been depreciated. 4 terabytes of internal general-use registers allow software to make more efficient data access, providing a more compelling Internet experience.
* Intel (r) AnswerNow (tm) Technology bends the space-time continuum, allowing the results of branch instructions and mathematical operations to be used before they are computed. The computations take place during idle cycles at some future time.
* Intel (r) CodeSpirit (tm) Technology processes machine code by its spirit, rather than its letter, completely eliminating software bugs and preventing malicious code, such as a virus, from executing.
* Intel (r) AlienCode (tm) Technology, based on CodeSpirit, allows users to execute programs written for any other processor, without previous knowledge of that processor's instruction set. The technology examines and "decyphers" the instructions and data in much the same way that scientists decypher written languages used by past civilizations. Via AnswerNow and CodeSpirit technologies, programs written for other processors actually run faster and better on Intel platforms than on their native processor. As a side effect, the processor now directly executes programs and scripts written in Java or any P-code or text-based language. In fact, even instructions spelled out in English are understood and executed by the processor.
* Intel (r) BrainWaves (tm) Technology allows the processor to read and write information in the user's mind. The processor is given away for free, and based on the user's thoughts, targetted advertisements are inserted directly into the user's mind. The process is painless, and simply feels like a song stuck in your head. A nominal (i.e., expensive) fee can be paid daily to eliminate the advertisements.
The Intel Shiitakeum Processor. Mushrooms Inside.
As for "code bloat" - deal with it, you are getting something bacl. Look at the memeory consumption for KDE2 vs. blackbox. sure, you are using ten times the memory, but in return you are getting a great deal of functionality. Your computer is there to be used, not preserved. Why not fill up that RAM? Why not saturate that CPU?
Actually, what Moore's Law essentially says is that the number of transistors on a chip will double every 18 months. The speed somewhat follows, but we have seen that simple scaling of transistor size is not sufficient to increase the speed linearly.
Take AMD for example. AMD stays with basically the same microarchitecture as when they first crossed the 1 GHz boundary, over 18 months ago. What are they at, 1400 MHz? That's a 40% increase in the past ~18 mos. Hmm...
Then you look at Intel. Intel practically abandoned the P3 to work on the P4, knowing the P3 was a dead end due to critical paths when scaling up the speed. The reason being that there are some parts of the microarchitecture that simply don't scale linearly with the rest of the process, primarily the memory system. Intel realized that the GHz race will guarantee market share, and has effectively succeeded in maintaining "Moore's Law" in the speed realm by scaling from 1GHz to 2Ghz in the same 18 mos. Sure, but it requires a reimplementation to do it.
If you scale these rates over 6 years, Intel has, yes the 2^4=16x increase you are predicting. AMD on the other hand has but a 1.4^4=3.8x improvement over the next 6 years. End result, Intel would have the 32GHz machine, and AMD would have the 1.4GHz*3.8 = 5.32 GHz Athlon that they call the Athlon 30K which actually performs as well as a 7 GHz P4, (yet still heats the small city.)
This really sounds bad for AMD, not to mention their incredibly-shrinking market share.
Intel has more info on this (both pdf's):
This backgrounder (4 pages, 17kb) has a basic diagram showing the change.
This briefing (18 pages, 2466kb) is a presentation, but actually has some nice detail. It has some photographs of the devices, better diagrams, and a picture of a naked man in the shower (really!).
I'll summerize:
PGA packaging (as used in many big processors) is basically a ceramic or fiberglass carrier board with pins on one side, wires in the middle (like a small PC board), and some method to directly attach to the chip. The chip is usually connected to the board with small solder balls, like BGAs, but on a smaller scale. The balls provide some flexibility and loose tolerances, but since they are bigger than the wires they connect they require a fairly large pad on the chip. This technology is a way eliminate these balls, allowing for smaller pads, freeing up more area on the die.
But you should check out the pictures -- they describe it better than I do.
HIV Crosses Species Barrier... into Muppets
>The advance here is removing the balls of solder between the chip's packaging and the microprocessor core
:)
Was I the only one who twitched when reading this?
oh man... I am a geek..... help!!
--- Metamoderating abusive downgraders since my 300th post.
I really feel I need 20 GHz. Anything that shaves even a few minutes off my day is very welcome. Considering just the work I do now, a 20 GHz processor might make my day 10% shorter.
If I had that speed I would do a lot of video processing. I also hope there would be good voice recognition. Long waits for compiling would disappear.
Bush's education improvements were
This is a *GREAT* comment. Please mod this up, it is worth it even though it was posted AC. It's a lot better than the standard, "Hey, look how fast I can run Diablo II now!" comments.
Anyway, by committees do you mean standards organizations similar to IEEE groups? I completely agree with that point, it would really help to get the research moving along. Unfortunately, I think many of the IC manufacturers are too worried about squeezing every last cent out of their current technologies before they put the newer technologies on the market. Really, there is no rush to market new technologies as long as they are still making money, and people are happy with the current products. That is what often cause the technology to stay behind closed doors for longer. A standards committee could help get things to market more quickly???
Maybe it's a perception thing, but I feel like my compile times stay constant no matter how much I upgrade my machine. Perhaps it's memory bandwidth or hard drive access, or perhaps it just that I've moved from ASM to C to C++ to (god help me) C#...
If you were blocking sigs, you wouldn't have to read this.
While this will be cool, it's not amazing. (Neither is the fact that that computer will come with about 10GB of RAM.)
--Ben
- raw processing power is way oversold
- machines are too damn noisy
I wonder, though if the CPU is the main culprit. A small, 50 watt gizmo doesn't generate that much heat. It's true that CPUs often have heat dissipation problems, but only because so much heat is generated in such a small space.On the other hand, we still use the basic IBM layout for PCs, where a huge transformer is mounted inside the box. That so-noisy fan is there primarily to cool the transformer. Even with hotter and hotter CPUs, the cooling needs of the rest of the computer have actually decreased over the years, because systems use fewer and fewer chips.
This design was obsolete two decades ago, when it was first introduced. Manufacturers at that time were moving to external power supplies, which can dissipate heat through radiation. Unfortunately, any computer not profoundly compatible with IBM's original quick-and-dirty design is now commercially nonviable.
This is true. The original Crays were roughly cylindrical with a bigger portion at the base. About half of the people out there are already sporting pocket-sized hardware of a similar nature.
If I recall correctly the original specs for the P4 stated a much larger cache and higher FPU. Then Intel found out that they would have to sell them for some insane price, like 1200 bucks, to make any kind of profit.
So, what did they do?
They clipped the FPU down to practically nothing, cut down the cache. Broke the JIT functionality and made the chip only able to really churn out specially optimized C code with any kind of speed.
Sorry, but MANY companies still use and program in COBOL, FORTRAN and PASCAL. Before any of you claim those are "dead" languages, remember that these languages run programs that have been in use on mainframes, companies spent millions/billions on, for more than 20 years. COBOL recently had some WWW extensions started or discussed a year or two ago as well.
I honestly have to question Intel's future processor roadmaps and production products when they show off things that are really to pricey for them to mass produce. It would be awesome if Intel was able to release the P4 like the original specs were. I would have one right now. The only thing is they didn't and the chip just ramps up the megaherts, but doesn't really do all that much more.
--
. sig seperator
--
If you ignore the other uses of a tool, does that make the tool less useful, or you less useful?
Before wishing for a pocket Cray, according to: http://www.dg.com/about/html/cray-1.html the Cray-1 was a 160 MFlop machine.
I'm not sure how to equate that to X86 floating point, or even what the Cray-1 clock speed was, and I realize that it was a quarter century ago. But I think that modern garden-variety PCs are in or above the Cray-1 performance realm.
The living have better things to do than to continue hating the dead.
Why buy an Intel 20GHz CPU for $n when you can get an AMD 14GHz CPU for $(n/2) which beats the Intel 20GHz CPU in almost all benchmark tests?
Just don't forget not to remove the heat sink.
#!/usr/bin/perl -w
use strict;
print "\n\nThe Magic Perl will entertain some queries now.\n\n";
my $quit = 0;
until ($quit) {
print "What is your yes/no question for the Magic Perl? \n";
my $ques = <STDIN>;
chomp $ques;
my @q_ans = (
"Yes.",
"No.",
"Maybe",
"My sources say, \"Yes.\"",
"My sources say, \"No.\"",
"These are not the droids you're looking for, move along.",
"You are not ready to hear the answer for that.",
"11.",
"The answer you seek is within you.",
"Certainly.",
"No way.",
"nowonmai...",
"Doh!",
"How the Hell should I know?",
"You must learn control.",
);
my $rand = rand @q_ans;
my $ans = $q_ans[$rand];
print qq(\nYou dared to ask "$ques":\nThe Magic Perl says, "$ans"\nThe Magic Perl has spoken.\n);
print qq(\nDo you have another question for the Magic Perl? Type "y" to ask.\n);
my $again = <STDIN>;
chomp ($again);
if ($again eq 'y'){
$quit = 0;
} elsif ($again ne 'y'){
print "The Magic Perl grows weary of your queries anyway! \n\n";
$quit = 1;
}
}
This
Easy Bake Processors*!!
Cook your favorite Goodies, and process your RC5/SETI packets fast! Purchase the Space heater for those cold nights in the Dorm/Batchlor-pad *Keep away from combustable material, do not touch Processor, Case, or desk. Intel coproration is not responcable for injury or death.
Anime : Drugs would be cheaper
The cool billion concept is cool, but it also points out that the processor paridigm is locked in for another 6 years.
It is my hope that within 6 yrs there is a greater focus on the -way- the little "ones" and "zeros" are processed, not necessarily how much faster it is done based on current standards.
I am me...I think
20GHz in 6 years? Sounds slow to me.
current speeds are at 2G 2 X 2 = 4GHz in 18 months
4 X 2 = 8GHz in 3 years. 8 X 2 = 16GHz and then 16 x 2 = 32GHz in 6 years. So why is IBM falling behind?
Ascii artist &
Isn't it time to concentrate the innovation on
these areas? How does a xGhz CPU help speed up
work on a 300Mb graphic?
At Intel, the trend seems to be for new-generation processors to be released 2-3 years later than predicted, while still benefiting from Moore's Law in the meantime.
For example, Merced was originally projected for release in 1998, at clock speeds around 300 MHz. (Source: Usenet postings from early 1995.)
My guess is that it will be 2010 before we see the gigatransistor chip this article is talking about, and that it will be at least somewhat faster than 20 GHz when it does appear.
Dahlmann tightly grips the knife, which he may have no idea how to use, and steps out into the plain.
Check out your hard drive ... my compile is disk load bound on a medium size project, I would expect any project with more than 10 meg of source code or so, particularly with a lot of small files would tend to be disk bound on most modern machines.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking