
Intel Looks to Billion-Transistor Processors 136
Weedstock writes: "EE Times has an article about Intel's next decade roadmap. It explains what are the current issues with the actual "plastic bumped organic land grid array" packaging technology and how it will be modified into a "bumpless package with built-up layers" to accomodate billion-transistor processors."
Re:This is offtopic... (Score:1)
Re:This is offtopic... (Score:1)
Dude, who stole my house? (Score:2, Funny)
OT: magnetic refrigerator article - dupe (Score:1)
Re:OT: magnetic refrigerator article - dupe (Score:1)
Unbelievable (Score:2)
Re:Unbelievable (Score:1)
Re:Unbelievable (Score:1)
Re:Unbelievable (Score:2)
Hell, long before the Amiga, you had a seperate computer that did nothing but handle the display (e.g Pluto, Pixar Image Computer, Ikonas), and people thought it was pretty cool when you could integrate graphics into your main computer (not the the CPU, but the same box).
Re:Unbelievable (Score:2)
accelerate popular functions in the HARDWARE, instead of relying on software to carry out the functions.
With that in mind...one word:
Winmodems
But I agree. I've said to many people: "Never replace hardware with software".
.
Re:Unbelievable (Score:1)
Re:Unbelievable (Score:1)
by Anonymous Coward on Friday January 04, @09:45PM (#2789471)
I think KDE has something to say about Gnome. KDE is like a mechanized Panzer assault. They will blitz right through Gnome
I agree with you. QT uses C++ while Gnome and GTK are stuck in the Dark Ages of Coding. Maybe GTK2 will resolve this issue...ack.
Re:Unbelievable (Score:1)
Re:Unbelievable (Score:1)
Of course! (Score:2)
Re:Of course! (Score:2)
From Microsoft's point of view the ultra fast and powerful processor will allow them to write the 2012 version of Windows in Visual Basic 13.0; they will be able to hire beggars off the streets of Bombay at $0.40 a day to write their OS - no more expensive college grads to hire. Here is the real reason (from the Microsoft perspective) for more powerful computers. Naturally their PR people will tell everyone that the new version of Windows cost almost a trillion dollars to write, and everyone in the press will solemnly repeat that claim.
If you project down that path a little more you can arrive at true artificial intelligence so that Microsoft can have computers writing the next generation of Windows without the need of human intervention. That way they can cut out their largest expense - programmers - and jump their gross profit margins from 90% of sales to 99.9%. Once this occurs you will actually start to see faster versions of Windows as machines won't need dumbed down languages to program in.
Old news but learn more about it.... (Score:4, Informative)
bottleneck (Score:3, Interesting)
The thing that i'm curious about is whether or not these changes in chip packaging will result in a disorganized series of changes in chip/board interface standards. socket 7, slot a, socket 370, etc.
Will the various companies(most notably intel and AMD) all be independently trying to solve the same problem in different ways? And will this mean that not only will we have rapid interface generations within the same company but that we will have to deal with even further incompatability between chips of competing companies?
-
Re:Says absolutely nothing (Score:2, Informative)
Bumped to bumpless is important because the solder balls were starting to dominate layout and packaging considerations.
Having better lithography processes and smaller traces is important because even though everyone expects them to continue to improve, someone still has to discover the technology required to do so first.
The new gate oxide has been shown off and does exist. It's only really improvement is for current leakage so it's not stop the world kind of news. Cost per wafer and manufacturability with existing machinery are the real things to worry about.
This is a different SOI approach than previous chipmakers have been using. They're also doing variations of xOI whereas other approaches use a standard silicon substrate.
The whole slashdot crew needs to learn more than just technology before posting articles.
Re:Says absolutely nothing (Score:2)
Yet somehow, I managed to get OVER_RATED in a previous post for saying the same thing?
that's great and all, but what about the chipset? (Score:2, Insightful)
Re:that's great and all, but what about the chipse (Score:1, Interesting)
also, a large important factor is the use of those billion transistors. it could be used as a large onboard cache, or a massivly parallel adder, or something completely useless. and the something completely useless part is probably what intel will produce, not because their produces are crap necessarily, but because they continually use that pathetic x86 architecure. no matter how many clever tricks you use to decode, how many stages you make a pipeline, and how risc-like your core is, the external instruction set is still a severe limiting factor. it becomes uneconomical (in theory) compared to simpler alternatives. At least, with ia32 it is awful (excited me in middle school, then i realized how toy-like it was compared to something useful, like a mips or an ibm ppc or something). im not as sure about the ia64 architecture. if they're going to make something that sophisticated, i'd hate to see it blown by lousy implimentation. "yay, my cpu has
Re:that's great and all, but what about the chipse (Score:1)
Wow! (Score:1)
Wow! This is pretty amazing. Just makes you wonder when traditional computing ( i.e. not quantum computing ) will reach its limits. I remember reading that this could occur around 2010, but then again that is barring new advances in physics.
Even then we do know that there are limits, for example there is a minimum limit to the amount of heat produced in a computation ( this is a result of the Second Law of Thermodynamics ). So there is a limit to the number of transistors that can be fit into any given area, otherwise the processor would be putting out too much heat energy.
Well anyway, this is very interesting and will make running simulations of real life scientific phenomena better, and as a result our understanding of the universe around us will be enhanced.
Re:Wow! (Score:2)
But we are nowhere near that limit.
A billion here, a billion there... (Score:2, Funny)
So What Good are a Billion Transistors? (Score:1)
Cool! (Score:2)
Re:Cool! (Score:1)
Heating a problem? (Score:4, Interesting)
But the article claims that the new technology will allow them to *embed* the
processor(s) inside the casing material, unlike today where the core actually
sticks out above the packaging.
But the advantage, as I see it, to having the core *above* the packaging, is
that heatsinks, thermal grease, etc... all have direct (or extremely close
to direct) contact with the core - which is what generates the heat. Mabye
in reducing voltage, heat output will drop significantly, but I digress.
With the core embedded in the casing, it would seem hard to help cool the core
when a heatsink doesn't have direct contact.
I may be wrong, and in that case just ignore this comment, but I don't know
how Intel would plan on dealing with that as a problem (if it in fact is one).
Re:Heating a problem? (Score:3, Informative)
In other words, your heatsink will have more or less direct contact with the core, but there will be other material around which will make sure that you don't accidentally crush the core when you push down on the heatsink.
Re:Heating a problem? (probably less) (Score:4, Informative)
As for embedding the core in the packaging - it's probably a great bonus. As has been pointed out this means that the top of your chip will be completely flush so you'll hopefully get better thermal transfer since you have a bigger surface area.
On a current intel chip the space between the packaging and the heatsink is currently acting as an insulator (since air does that best when it's not moving).
In addition to this, I would speculate that if the core is embedded into the packaging it might allow for small heat pipes to run directly into the core, allowing particularly hot areas of the chip to have additional passive cooling.
That said, given fabrication facilities i'd struggle to make even a single pnp transistor and whilst i could probably remember how to build simple mos (and hence cmos) gates - i'd struggle to replicate what intel was doing in the 70s... so dont take me as any sort of authority on this one.
obligatory comment ... (Score:3, Funny)
Re:obligatory comment ... (Score:2)
Re:obligatory comment ... (Score:1)
Power, signal integrity & materials HUGE issue (Score:1, Insightful)
Then you have the issue of signal integrity, particularly for high-speed analog and differential pair signals, which smaller traces only aggravate. The most advanced flip chip packages in the world currently push around 2000 connections for power and signals. This will only get more aggravated and congested at the board and package level as the level of integration increases due to the feature size decreases on the silicon. Small lines increase impedances, and merely cutting layers away will NOT help. Differential signals are supposed to be pushing 40Gb/s per pair in a year or so using modulation on top of differential signaling, so what are they expecting that these packages will be supporting when they have such strict routing requirements both in the signal and redistribution layer routing AND through the package? Not to mention the fact that they still have to attach these monsters using a substitute to lead solder to avoid alpha particles causing false switching in already small noise margins.
Instead, you need different package materials than simple organic laminate subtrate and different silicon process materials than silicon dioxide and tungsten vias. When it gets to this, they have to rely on material science, which is the gating factor in a lot of science right now. I don't believe that Intel's core competency includes material science per se, so they'll be relying on outside companies and research labs for a good chunk of the new materials. Since this is out of their direct control, I don't see how they can deterministically schedule their packaging roadmap - not without forming clear strategic alliances with companies whose core competencies lie in material science related to the above-listed materials. I wish them all the luck and blessings in getting there though.
Why?! (Score:1, Troll)
Who needs a BILLION transistors in a processor, for crying out loud?! Let me tell you something. A slow 4- or 8-bit processor can execute amazing things when coded correctly. Embedded developers have interfaced these processors to memory, hard drives, CD-ROMs, the ISA and PCI busses, and just about every kind of peripheral out there. I'm beginning to think that a fully functional and FAST computer can be built with NO x86 processor, but with about $20.00 (US) worth of these cheap, slow and small processors. It's the software that needs to be engineered correctly, and I'm afraid that nearly all software out there isn't.
What happened to the good ol' days when programmers--real programmers--wrote very clever, small and fast programs? When it had to be written correctly or it didn't work?
Try explaining to me why nearly all hardware needs to be engineered correctly, for a minimum of components and a maximum of performance, yet nearly all software is slopped together, taking up tens or hundreds of megs and running noticeably slow on today's powerhouse machines. You know what? There's no excuse.
I've seen a hard real time operating system coded in 700 words. I've seen processors with 128 bytes of RAM control industrial robotics. Speaking of industrial stuff, I've seen an automation system that packs a real time operating system, high speed communication, interactive user interface (including full control of the display hardware), and all the automation software... in 20 kilobytes. Seeing this, I cannot understand why something simple like a word processor program should be several megs in size (and why it should hog a ton of memory).
So back to the billion transistors question... why? Why should the processor have to predict the next mess of instructions, load them into a cache, find out it predicted incorrectly, dump the cache, find the correct location, load the instructions... Why are processors marketed by their internal clock speed when they spend most of their time waiting for data? And above all, why does software suck so badly?
OH WELL.
The Lord of the Rings. The book rocks. The movie sucks. Yeah, it SUCKS! I left the theater halfway through it. It SUCKS! But the book is awesome.
OH WELL.
Re:Why?! (Score:3, Interesting)
#1 - Larger memory sizes. Terabyte databases require terabytes of RAM. Current 32-bit processors can't touch that with a 10-bit pole. Even the most elegant 4- and 8-bit processors can't do anything about their memory addressing limitations without huge kludges.
#2 - Engineering/Scientific problems. Ever try to model the fluid/thermal dynamics of a star? You need ungodly amounts of processor power to do this properly, or ungodly numbers of processors. Preferrably both.
#3 - 3D multimedia and design. This is my area of work. I've got five (count 'em, five) dual Athlons right this moment rendering like mad, churning through a 1 hour 3D animated sequence with lots of volumetric lights, NURBS, and tons of polygons. 3D eats cycles like they're going out of style, and in my business if you can cut your render time in half, you've just doubled your production capability. You can never buy enough render power.
#4 - Gaming. Yes, games. Doom. Quake. Doom II. Quake 2. Quake 3. Unreal Tournament. Every game pushes the triangle count, texture resolution, and framerate to higher highs. Photorealism is the holy grail, and it's going to take absurd amounts of transistors running at an unheard of clockrate to do this.
You'll note that business apps are anywhere in there, and they shouldn't be. Your average desktop processor spends about 99% of its time idle waiting on the operator between keystrokes. Nobody needs a 2Ghz P4 or a 1.6Ghz Athlon for these tasks, despite Intel's propaganda to the contrary.
I know you long for fast, tight code, but that isn't being taught in college anymore (heck, it wasn't even when I went through in 1990). Profs are encouraging rapid design and quick-to-market code over elegant design. It's unfortunate, but the market itself is rewarding this philosophy. I don't agree with it, but the fact is that the company that produces a "good enough" piece of software quickly will generally steamroller a company that produces "elegant" software but comes out later.
After all, beta means alpha, and 1.0 is really an extended beta. Kick it out the door, the marketing campaign is scheduled to start! Who cares if it works, we can always patch it later or put the bugfixes in version 2.0!
Oh, and I strongly disagree with your assessment of Lord of the Rings. I found it a very good adaptation of such a sprawling book. What did you dislike about it so much that you descend to profanity to describe it?
Re:Why?! (Score:2)
Re:Why?! (Score:4, Insightful)
We decided we wanted to do more with our computers. It's all very well to long for the days of very clever, small and fast programs but it's entirely another thing to create software which does all the things we have come to expect today while still keeping the software incredibly small and fast. It's even harder when you want to stay within a tight schedule and budget.
Lets look at something near and dear to our hearts, something that many of us here have contributed to and something that isn't affected by budgets or timelines (well, mostly) - the Linux kernel. The Linux kernel is undoubtably a very good piece of software development, arguably the best that's currently available and it has been created by a wide range of people many of who come from the days when RAM and CPU time was expensive. Despite this, the linux kernel is certainly not small, and it shouldn't be. It has a wide range of devices to support, it has to be able to handle multiple users simultaneously and it provides a bunch of services that previously would never have been provided in an OS, let alone in a kernel.
It could be argued that the Linux kernel is clever, and with my lack of knowledge of the kernel source I can't really comment. I think it is safe to assume that it's not as clever as it could be though - it doesn't use every trick in the book to reduce file size and increase efficiency because it's no longer small enough to make that kind of thing feasible. It's also modularised so that things can be loaded and unloaded as needed, there's extra code and overhead required to provide that. Finally, it supports a range of architectures now and is more portable. Going back to the old ways of doing things gives up all those benefits.
Finally, the linux kernel is not fast - it is comparably fast for all the things it does, but it is not as fast on a per-cycle basis as OS's were back when every cycle mattered. It does however provide more features (like loadable modules), more portability and a faster release schedule for fewer man hours.
So when you really sit down and think about it, while programs these days take up more RAM and CPU power there are a range of benefits that come from this. You should also note that comparatively the overall experience of using a computer has become radically faster then it previously used to. You may think that a program feels slow when you run it on a 3 year old machine, but what you fail to realise is that you've just gotten used to how much faster your new machine is. Having said that, some software is just plain crap, but so are some cars and bridges so the bad apples don't just come from software engineering.
Why should the processor have to predict the next mess of instructions, load them into a cache, find out it predicted incorrectly, dump the cache, find the correct location, load the instructions...
Incredibly poor chip design actually. This problem really only becomes significant when pipelines are made too long (such as in the P4). The pipelines are extended to make it possible to use a higher Mhz rating - though because of the extended pipeline and the problems caused by having to guess ahead so far the CPU doesn't actually function anywhere near as fast as the Mhz would indicate it should. This is why people talk about the Megahertz Myth - there's a ton of information on it around the web.
Why are processors marketed by their internal clock speed when they spend most of their time waiting for data?
Because consumers don't understand computers well enough to know this and Mhz has been used as a rating mechanism for so long (and previously it had been reasonably accurate). Marketers will jump at any opportunity to make their product sound better than the competition.
And above all, why does software suck so badly?
It doesn't. There is and always has been poorly written software but to say that all software sucks is unjustified. There are cars that break down due to manufacturing defects, bridges that collapse, constructions which go over time and budget and a myriad of failures from all types of engineering so of course not all software is perfect but it is improving whether or not you like the way it is improving is another matter.
Re:Why?! (Score:2)
Actually, pipelines aren't made longer to get higher MHz rating only, but to increase throughput [in optimal case]. Current crop of CPUs do more per clock than older ones (well, not counting P4, usually). You can nowadays add more than two numbers in one clock cycle and possibly do additinal multiplication in the same time. Even P4 should be really fast if all you do is basic operations without loops. P4 has 3+GHz ALU unit for this! Unfortunately, we really don't need that much computing power but logic power partly because we have additional processors on our sound and graphics cards where the computing power really counts. If you really need to emulate DSP in software, then P4 is what you need, otherwise deep pipeline is going to hurt badly.
Perhaps it's just you didn't expect that much from computers a couple of years ago. I remember using 75MHz Pentium with sucky graphics adapter for not too many years ago and it felt plenty fast. I'd hate to have to use that kind of crap anymore - no matter what software I used. And that's because I know about better.
Re:Why?! (Score:1)
Re:Why?! (Score:1)
Intel needs new products coming all the time to stay in business. This is the exact same business model that Microsoft follows. In the near future, when the saturation of technology hits some level, you will see some truly stupid product ideas to get people to buy even more crap.
obligatory amiga comment... (Score:1)
Go out and buy an Amiga
Anyhoo - the way I justify software bloat is that hardware is so cheap these days does it really matter? I mean on a desktop level...
My favorite comment about memory and the Amiga - was an issue of Amiga Format that had a full (older) copy of Real 3D - which was one of the first programs to ever do particle kenimatics. Anyhoo - the label said "warning requires at least 4 megs of ram" - I probably have the disk around here somewhere if someone doesn't believe me.
Anyhoo - software is bloated sure, but does it make any difference when hardware is so cheap?
Re:Why?! (Score:1)
As to massively parrallel chips idea, its good in theory but horrible in practice. Most code can't be broken up into bite size chunks to be handled independantly. You rely on previous data, and have to be sure its completed. You can't access the same segment of memory or because you may be reading/writing the wrong information. And to solve this it takes more code, not less. So you may have a slower implementation and one that's harder to design and maintain.
The ATM machine problem shows this, where you have a husband&wife both taking money out at the same time (eg. emptying it). If both can access the data attribute, they both check to see it has cash and are allowed to withdraw. The bank goes into the red. You say, make the line repetitious so one dollar at a time, but you still need a check.
If (x>0)
x--;
The if goes through, but the husband withdraws at the same time. The bank still loses in the worst case $1. Dealing with parallelized code is a pain with shared resources and since its bigger, more is cranked out. Hopefully you get a speedup by having more code, but running in parallel on many chips, yet often less code is faster. Maintance is hell, so often the work to do this on important aspects, not every little thing.
And you'd ask us all to write in assembly, ugh. Think about writing 8 or so lines for a simple switch statement, keeping track of jumps/labels, dealing with a small # of registers, dealing with memory. A simple program is horrible to write in assembly, since simple if-else code takes branches, jumps, labels, etc - longer code. Put this on something more then a few lines in C, and its hard to debug since its all addi's and beq's for ever little command.. stacks to return from a function. Its messy! Try dealing with your parallel goodness in assembly.. insane. And compilers know all the tricks, so often a modern compiler is better then a skilled assembly writer, since they can do far more tricks easier. You need to know which instructions are slow and not to use, optimize registers load/stores to reduce stalls, etc. Sure a perfect assembly writer may know it all, but its insane on chips with huge numbers of instructions, registers, and a big project. The myth that good assembly is faster then a compiler is just that, a myth. Ideally is true, in practice time is more important and a compiler often wins.
We do parallize code like crazy, but in smart ways. Up at the cpu level is okay and done a lot, but not much more efficent. Go down a level. We use pipelining to parallize the cpu stages, so its not stuck computing one instruction through the whole process, but each stage can work on one. 1 in, 1 out every cycle (different instruction) or just the same in 5 (multiplier waiting for the decoder). Look at SMT which fills in the bubbles (stalls) when a stage must wait by simulating another CPU so other data can fill it. Think ILP and EPIC with prediction to replace branch prediction, by using more hardware to do the task in less time. Instead of picking the result for an if statement when waiting for memory to respond and being wrong, you do both simultaniously and throw out the incorrect data. Sure its brute force rather then trying to be 'smart' but its faster. That 10% of the time your wrong is gone, so your better off.
I could go on, but I spent so much freaking time writing this for no reason. Don't need nor will likely get mod points, doubt you would care enough to learn. If you would like to know more, ask though. I'll leave you with this:
What happened to the good ol' days when programmers--real programmers--wrote very clever, small and fast programs? When it had to be written correctly or it didn't work?
Programmers have to write big programs, smart and clever in radically different and innovative ways. Design is no longer about size, but modularity, cleanness, reducing debugging and maintance time, and adding features. Larger code is acceptable if its better code - its easier to fix, and only slightly slower. Today's real programmers deal with designing massive, complex projects - not optimizing to hell for some platform or language. We leave the platform designers to optimize their end and the compiler to optimize to the hardware. Most real programmers have more important things to spend their time on.
But when am i... (Score:1)
Non-computer applications (Score:3, Interesting)
Thought for today: why do HTDV receivers cost so much? A GeForce 3 board has 35 million transistors in the CPU, 64MB of RAM, and costs under $200 at retail. The radio part of a cell phone, which is more elaborate than the radio receiver for HDTV, has a parts cost of about $10. $600 will buy a pretty good computer, monitor and all. Why do HDTV receivers cost upwards of $500 without a display device?
Re:Non-computer applications (Score:1)
This will teach them for dragging their feet on High Definition Television!
Re:It does not matter. (Score:2)
For example, if your processor was fast enough you could do accurate physics modelling in real-time. The amount of data is quite small (not memory bound) but the number of equations you have to solve are enormous.
This would be fantastic for 3D games - objects would behave like like do in the real world rather then being forced to follow strict animated sequences. You could model chains, pulleys, swaying bridges, tumbling crates, bouncing balls, ramps etc.
Thanks to the GeForce and other high-end graphics cards we have the power to render scenes with this level of complexity, it's just a shame that we don't yet have the processor horsepower to accurately model them.
Re:Thank god for Intel (Score:5, Insightful)
AMD is not copying Intel's IP. If they were, Intel would be winning the suits against AMD, not losing them as they have been. AMD has reverse-engineered the x86 instruction set that has been around for quite some time, but implemented in silicon differently. The end result is greater performance, as evidenced by any benchmark you care to run. Like it or not, the fastest x86 processor on the planet right now, even according to Intel's own benchmark suite, is the Athlon XP 2000+.
And to further add insult to your injury, AMD doesn't stand for "American Micro Devices", it stands for Advanced Micro Devices. If you'd done the slightest bit of reading, researching, or thinking before you posted your previous comment, you'd know that.
Re:Thank god for Intel (Score:1)
Since I do a lot of 3D rendering work, the Athlon has somewhat of an advantage, even moreso than normal, because of its incredible FPU. I don't think Northwood has a revamped FPU, so I really don't think even a 2.2Ghz Northwood is going to beat an Athlon 2000+, much less a dual setup.
Still, newer processors mean lower prices for everything else. What's there to lose? I love it!
Re:Thank god for Intel (Score:1)
Any slight hope of resale of your old kit?
Not that that is a new thing, I still have more or less every PC I've owned, just because they are more use to me that the £50-300 I'd get by selling them.
Re:Thank god for Intel (Score:1)
Intel is so full of shit it ain't even funny, saying the P4 is better than the P-III. If Intel had spent time shrinking the P-III die instead of going megahertz-happy with the P4, they'd still be producing good stuff. 3D folks are avoiding the P4 like the plague.
Re:Thank god for Intel (Score:1)
Re:Thank god for Intel (Score:3, Interesting)
And you tout Intel for get thermal protection working?! Howabout some chips that don't need thermal protection (like the recent iMacs which can cool themselves with ONLY air convection [meaning no fans for you lesser literates])?
Next thing you know, you'll be crowing about MS's innovation of the GUI over Xerox's... ah.... GUI?!
Re:Thank god for Intel (Score:2)
Re:Thank god for Intel (Score:2)
I don't think "American Made" is a significant point when deciding on a processor...
Where does one get ATX-format boards and processors to be able to build these ultra-quiet, ultra-fast marvels? (Seriously - I like quiet PCs, but I don't much like Macs, so that would be a nice LinuxPPC box, potentially).
Re:Thank god for Intel (Score:1)
AMD is headquartered in Sunnyvale, California:
One AMD Place
P.O. Box 3453
Sunnyvale CA 94088
800-538-8450
Re:hypocrisy on slashdot (Score:1, Offtopic)
Just because they don't say anything about security means they are going the MS route of security through obscurity. Remember, the holes found in MS's software is found by 3rd parties, documented, and THEN MS denies it, doesn't patch it for days/weeks, or calls it a feature.
Do you have any evidence what-so-ever that this upgrade has anything to do with security issues? No? Then shut the fuck up, troll.