Slashdot Mirror
Tiny New Chips Win ChipCenter Award
phi1o writes "Ever wonder why a burrito-sized pentium is also called a "micro" chip? Me too... This page @ ChipCenter shows a new line of chips from National Semiconducter that are *really* small. They're all analog and of course nowhere near as complex as a P3, but you gotta wonder if processors are headed down this road in the future, to where they are mostly silicon rather than ceramic and plastic. On the one hand, this will make for smaller devices. On the other, it's gonna make building anything without a giant factory (ie in your bedroom.... hardware geeks with me?) much more difficult. I don't know what to think. "
I sure hope DIPs remain popular. I am an electronic hobbiest and I like the idea of being able to solder things to PCBs :)
Oh great, now you've done it! I suppose it's only a matter of time before someone tries to plug "VIC Avenger" into his new BX board!
Hmmm...
I must admit that 'burrito' isn't the first thing that comes to mind when I see a P2/3 cartridge.
Actually, they're very similar in size and shape to the game cartriges that the old Commodore VIC-20 used.
But you're right... Not 'micro', by any means. I thought 'micro' was supposed to mean that you can't see it.
>Ever wonder why a burrito-sized pentium is also >called a "micro" chip? Me too... Might have something to do with the size of computers *before* the IC arrived...
Comment of the year
This looks like a very good thing. By focusing on shrinking the package as opposed to the trace size, they avoid the costs associated with building new fabs to support the shrunk designs. Additionally, by incorporating the package on the wafer and eliminating manufacturing steps, it should be very attractive to implement.
Provided that the tech can scale up, and that the market adopts it, it could be a big step towards super-small devices embedded in just about anything.
Let's see...a microcontroller, a temp sensor, and a d-a converter in a 3-pin-head size. Stick that in your shirt and you've got something that beeps when you get hot under the collar!
Of course, when they come up with a CCD and a couple megs of RAM on the head of a pin, that's when I'm going to worry.
------
I agree with you that maybe, its the way things are going to turn out..
November issue of PC PLUS (UK)did a piece on how Keele university is looking to make Harddrives that are Bigger that all the competitions, yet the size of a credit card and with no moving parts.
These look to be released early 2000... And are to retail at about £30 - Bigger hard drives (space wise) smaller physically yet nothing compared to the price of most hard drives.
If this product does come out will it mean a rat race from all the companies to try and come up with similar technology?
It would be interested to see what does happen when it comes out (if it does)
Sometimes the only way to stay sane is to go a little crazy!
These are "smart" sensors, each of which has a unique 64-bit address, up to 254 of which can reside on a single pair of conductors that can be up to 300 meters long.
How many transistors are in a couple of op amps? I'm guessing on the order of 10-100.
How many transistors in a PIII? Millions? Tens of millions?
If processor manufacturers used this same process we would expect the the PIII to be 10,000x to 1,000,000x bigger than this 'little' device, giving it a size of 200mm x 210mm to 2000mm x 2100mm (given that this device is 2mm x 2.1mm).
The only thing interesting here is the packaging and that's not something that is going to scale well when you try it with millions of transistors. You need the heat dissipation mechanisms that the P3 has.
/*
FBI CIA NSA IRS ATF BATF DOD WACO RUBY RIDGE OKC OKLAHOMA CITY MILITIA GUN HANDGUN MILGOV ASSAULT RIFLE TERRORISM BOMB DRUG HORIUCHI KORESH DAVIDIAN KAHL POSSE COMITATUS RANDY WEAVER VICKIE WEAVER SPECIAL FORCES LINDA THOMPSON SPECIAL OPERATIONS GROUP SOG SOF DELTA FORCE CONSTITUTION BILL OF RIGHTS WHITEWATER POM PARK ON METER ARKANSIDE IRAN CONTRAS OLIVER NORTH VINCE FOSTER PROMIS MOSSAD NASA MI5 ONI CID AK47 M16 C4 MALCOLM X REVOLUTION CHEROKEE HILLARY BILL CLINTON GORE GEORGE BUSH WACKENHUT TERRORIST TASK FORCE 160 SPECIAL OPS 12TH GROUP 5TH GROUP SF
*/
Burrito sized you say? Now we get the best of both worlds... small, efficient micro SMD's AND your choice of cheese, beans, and hot sauce!
I thought that the "Microchip" description applied to the size of the piece of silicon - not to the whole package of plastic and ceramics and what not...?
I know the Pentium series isn't exactly small still - but it's still a lot smaller than the whole pacakge of heatsink and fan would lead you to believe...
Gentlemen, start your penguins
There are obviously many different kinds of chips but for processors and ASICs the trend is towards larger and larger die sizes. The three main factors in this are 1) need for more I/O and power pins, 2) Advances in semiconductor processing facilities and equipment have reduced defect density over the years to make larger chips practical, and 3) For processors especially, an increasing fraction of the chip is memory which can be protected to a large extent by redundancy.
The eight bit micros like the Z80, 6800, and 8080A had die sizes of around 20 - 22 mm2 and 40 pins. In the 16 bit era the die sizes were around 40 to 80 mm2 and pin count ranged from 40 to 68. The early 32 micros like the 68020 and 386 were about 85 to 100 mm2 in size and used 114 to 132 pins.
In the start of the modern PC era the emphasis was on performance and features and die size was pushed hard with the anticipation of multiple process shrinks to allow the part to enter the manufacturing sweet spot of around 100 to 150 mm2 for inexpensive and high volume production.
The Pentium and Pentium Pro (P6) were both introduced around 300 mm2 but didn't reach high volumes until shrunk into more advanced processes.
The Pentium shrank from 306 mm2 in 0.8 um, to 163 mm2 in 0.5 um, to 91 mm2 in 0.35 um (with caches doubled in size and MMX added).
The largest processor today is the PA-8500 which checks in at a whopping 475 mm2 in 0.25 um. But remember this chip is nearly 3/4 cache SRAM (1.5 Mbyte). The 8500 is probably the largest chip that can fit in the field of exposure with the current generation of optical steppers.
We can expect mainstream processors to get larger in the future because of the trend towards integrating level 2 and even level 3 caches on-chip. Memory redundancy makes these larger chips practical while the growing disparity between fast wide issue CPUs and main memory performance makes large on-chip caches mandatory. For embedded processors and CPU cores in system on a chip type applications the die size is typically kept from shrinking much below 50 mm2 by bonding pad spacing requirements ("I/O limited").
Strictly speaking, microchip indicated the relative size of the device (at the time when the average Mainframe took up an entire building), but now indicates a class of device
As for happy hacking with microcontrollers/microprocessor: as long as there is demand, they will make devices in plastic and ceramic packaging
some karma... and kinda lukewarm about it.
-Why are you searching the carpet with a magnifyng glass?
-I dropped my laptop.
You are the same decaying organic matter as the rest of us.
"Whatever else you are told to believe,
do not accept that size is unimportant"
Yeah, we hear *that* one all the time. So what are you saying?
Oh, smaller is *better*? Yeah, those nerds use *that* excuse all the time.
Any female slashdot readers out there want to confirm or deny these
accusations? And what about that bit with the standard mounting process,
do you liberated women agree with that too? And... hey, is this segfault?
---
pb Reply rather than vaguely moderate me.
pb Reply or e-mail; don't vaguely moderate.
"Teacher, I couldn't finish my homework because the cat swallowed my computer."
-=-=-=-=-
-=-=-=-=-
My mom's going to kick you in the face!
At one time, I had thought that smaller, more specialized processors would be needed to handle different devices. In stead of having a huge processor (PII/PIII) that can do everything, we would have many different chips performing single tasks. This may be so when we get down to intelligent appliances. However, with the age of the pda/cellphone/etc etc upon us I have to wonder just how small we can get and still have good functionality. You had an article about IBM's wearable PC coming out next year. With that and the pdQ and more devices to come, I tend to think the smaller the better. If we have a "factory" in our house, it's our own fault. It will be almost impossible to specialize in "everything" and therefore it will be unneccessary to have spare chips laying around. While I'm in favor of smaller chips, I would much rather have a smaller GOOD chip that didn't have all the built-in video and sound additions to it. Sure I want color in my pda/cellphone and I want sound, but it's not neccessary for me to play EverQuest or Half-Life on this device when I'm board. And even if I could, I could handle not having high-res graphics on it... I can save that for the machine at home!
SL33ZE, MCSD
em: joedipshit@hotmail.com
SL33ZE - Artificial Intelligence is No Match For Natural Stupidity -
First off, while I don't think that all discrete logic and analog ICs will move towards this type of packaging, it definitely is the trend for newer chips of every type.
I've been a home hardware hacker for seven years or so, and I've seen things get a lot better, and a lot worse in different respects.
While it's easier than ever to get information about chips online now, with almost every electronics company offering datasheets in
As for actual hardware hacking, personally, I don't think it's too bad. It's just a question of what you're trying to do. Microcontrollers, discrete logic, and analog ICs are all still available in DIP (pin) packages and can be thrown together on a breadboard.
But as for high-speed digital design, this is simply out of the range of most home experimenters. There is a lot more too it than just finding a free/cheap PCB cad program (which do exist, try www.cadsoftusa.com), and a socket for that BGA processor (www.emulation.com). Not only is it next to impossible to make the multi-layer PCBs which are needed for these designs, but only the most expensive PCB software packages can help you keep track of the timing requirements for these systems.
The hardware hacker isn't dead, he just is shut out from moving upwards with the rest of technology.
I'm hoping to start a web page soon to help the hardware hacker who wants to move on to higher-speed and more complex systems.
Dr Wonky
Hello! Have you been living under a rock for the past fifty years? Or do you just think that the entire state of computer science is summarized by your IA32 Linux box running Quake?
Amazingly enough, the fundamental process dimensions of a chip are not precisely reflected by the packaging it ships in. Never mind that they're cramming components that used to fill huge rooms onto a small wafer; now it has to come in a small yet attractive package.
Please, Slashdot people, think before you type!
you gotta wonder if processors are headed down this road in the future, to where they are mostly silicon rather than ceramic and
plastic
I don't quite see the comparison with a P III. The peripheral components for CPUs are mainly for heat loss optimization (the fan itself is as big as the CPU). As far as packing more transistors/mm goes, CPU designers are already fearing the imminent era of hitting the size barrier of a few molecules/transistor.
I know that on my own PC with a celery 300 overclocked to 450, the little fan and packaging gets pretty hot to the touch, and if the transistors are already packed as tightly as they can be, I'm tempted to assume the National chips are of a different breed altogether, which doesn't make sense to compare to pentiums.
But then, I couldn't understand much of the electronics in the article, so maybe I'm wrong.
L.
>Hardware hacker - newer machine, DSP or FPGAs, prototyping boards, software (hard to have GPLed tools here) - $5000-$25000-How deep are your pockets...
well if youre a ee, and really interested in that kind of thing, well i know a couple of guys whove written motorola and asked for a chip to do projects with and gotten it(one a 68k and the other the 56k dsp chip). you should only ask if you really are interested though. sure youre not building your own desktop pc but you can still have fun with things you can get for free or cheap
Most of the size of the PII/III comes from the PCB and the external L2 cache mounted there. The Celeron had its L2 moved internal and became the socket 370 chips (about the size of ole pentium classic). The new flip-chip design for forthcoming Pentiums will also be about this size. Although the pins will be in the middle of the chip instead of around the outer-edges. Hence flip-chip. Sorry for AC dont remember password right now.
cool - an ickle cute beowulf cluster!!! yum yum :O)
"On the other, it's gonna make building anything without a giantf actory (ie in your bedroom....hardware geeks with me?) much more difficult. I don't know what to think. "
You don't HAVE to be an EE to get free electronic samples. Most of the time, you CAN get them for free.
You *CAN* do some hacking, be it hardware OR software on the cheap. But to become Woz, and create a company out of your garage (HP, Apple etc la) just isn't gonna happen. And soon, the same will be said for software, when software moves past the artist stage its in now to mass production stage its heading for.
The big push these days is SOC (System On a Chip). This involves buying IP (intellectual property) from the vendor and making said project. None of this is bedroom material.
If it was said on slashdot, it MUST be true!
(ie in your bedroom.... hardware geeks with me?)
.5 mm apart or less, or somehow solder an entire array of contacts simultaneously while they sit between the board and the chip.
It's allready getting hard to do hardware hacking with anything much more than an 8 bit processor. I think you can still get a 68000 in a 64pin DIP, but everything else is heading to QFP, TSOP, and worst of all BGA. To work with these you either have to solder wires spaced
It's because of things like this that I have very few designs on building any sort of major computing device in the near future.
Start Running Better Polls
Microstepping an XYZ gantry to move a vacuum-needle pick-'n-place, with a single-jet droplet depositor for flux coming before and a hot-air solderer coming after, more or less on the same toolhead... definitely the kind of thing a linux box can be programmed to drive (with a couple of dirt-cheap MCUs -- 8052? -- for the critical-timing stuff such as the actual pulses and the microstepper control)..
:)
Yeah, I could see something like that in the freehold utility room
Seriously, whikle this stuff will be as pervasive as present surface-mount is now (it's just SMD taken all the way), clunky old thru-hole devices won't go away entirely, they'll just get pricy, probably, because they'll be used mainly for prototyping.
If I visualize this correctly, one downside to the naked chip-on-board technique these packages use is thermal-related fracture: the silicon has far less flex than the leadframe it replaces, so the board the package is bonded onto better have the same thermal expansion as the chip, the chip will fracture or tear loose a few bonding-blobs otherwise.
Well, okay, there are some differences, but this is basically BGA, which has been used for LSI parts for some time (the first one I recall in high volume production was one of Intel's Pentium chipsets - was that the TX?). This is, essentially, BGA mounting for parts that don't need umteen-gazillion pins.
As for signalling the end of ICs as we know 'em (or of home-built/small-shop production), that's about as novel a suggestion (and as much borne out by the facts) as the oft-repeated "death of net immininet" nonsense.
Well, the size of the enclosure makes it easier to cool the processor.
/ dp010897.htm
/ mp012599.htm
/ mp040599.htm
If you want to see some small (what YOU might define as "micro") processors, you need to look at the mobile processors:
http://www.intel.com/pressroom/archive/releases
http://www.intel.com/pressroom/archive/releases
http://www.intel.com/pressroom/archive/releases
(copy/paste away my friend)
The processor is smaller than a dime after enclosure and before mounting on the PCB. This is truely amazing.
Generally this doesn't work like that - you get a solder stencil and have solder paste (grains of solder in a flux paste) stencilled onto the right bits of the board.
:)
The pick & place then puts the components down, and it goes through a reflow oven where the flux evaporates and the joints solder themselves all at the same time.
It's very sweet to watch
It's almost impossible to prototype with through-hole stuff now (on the digital side, at least), they just don't bother making through-hole variants of the chips. Prototyping nowadays is "make a pcb"...
Hugo
The Garage/bedroom hardware hacker is, for the most part, dead.
There WAS a time when you visited 3M's surplus, bought dead boards by the pound, desoldered the chips and tested the gates individually.
Then you built your own machine.
As we moved from MSI to VLSI (and beyond), the hardware hacking oppertunities have become the thing you do if you have a simple project (and you can reduce them to a PAL or perhaps a FPGA).
Because the OLD method was:
Design on paper
Build the prototype
Look for errors
put the logic scope on to find the errors
loop to build prototype till fixed.
(Anyone wanna buy a CP/M based logic analyzer prototype? one of 10 in the world)
The *NEW* method is
Design on computer in test environment
Test on computer
Build on silicon
Do a bit of testing (cuz it should work on the 1st time)
Ship.
*IF* you are looking to be a Hacker:
Software hacker - get machines thrown out in dumpster, and old software you beg, borrow, or GPLed $0
Hardware hacker - newer machine, DSP or FPGAs, prototyping boards, software (hard to have GPLed tools here) - $5000-$25000-How deep are your pockets...
Being a bedroom/garage hardware hacker is no longer cheap, and based on simple tools, like it used to be.
Software hacking is the last low-cost (Bedroom/garage) hacking frontier. You may not be as productive with cc and make as you are in an intergrated environment, but tools are only part of hacking..the grey matter you have to manuplate the tools matter most.
If it was said on slashdot, it MUST be true!