47 million parts,in this tiny die

[Pootie]

:O Maybe it would make a nice keychain or some shit :/ I have several,maybe I will make a clock :)

[XCNuse]

wow thats a tiny chip

lol maybe i should get out my pentium chip, that thing is fairly large :p can fit the whole base of it in the palm of my hand

[Pootie]

lol,ya the old intel chip,Belive I have one of those as well :P

[Forbin]

A friend of mine and I cracked open a 486 once some years ago, around the time the Athlon (the Slot A version) first came out. The 486 die was about the same size. :)

[Shotglass]

I've still got an old Slot 1 Pentium II computer lying around here. That thing is just ridiculous, it's like a Nintendo cartridge or something.

the actual processor isnt any larger than a modern one though

[Pootie]

A friend of mine and I cracked open a 486 once some years ago, around the time the Athlon (the Slot A version) first came out. The 486 die was about the same size. :) Ahh,so there was just a plate over top of it?

[Forbin]

Yep, a little Pentium history.

First was the Pentium. It had a small L1 cache and the L2 cache was located on the motherboard in the form of SRAM chips and ran at the same clock frequency as the FSB.

Then came the Pentium Pro. With this one, Intel took the L2 cache and put it directly on the CPU die where it ran at full CPU clock frequency. This made it rather large and expensive to produce, so it ended up being used mostly in servers.

Next was the Pentium II, which used a slot connector instead of a ZIF (Zero Insertion Force) socket. The die itself was mounted on a PCB with the L2 cache located on the PCB but separate from the die in the form of SRAM chips once again. This allowed for cache clock frequencies much higher than they could run if mounted on the motherboard. However, it was also a step back from the PPro.

Pentium 3 then saw the reintroduction of the on-die cache, if not with the first ones then at least shortly before the P3 moved back to a socket format. x86 CPU's from Intel have had on-die caches ever since.

Hopefully that illustrates why Slot 1 (and AMD's Slot A) were introduced.

[Forbin]

Ahh,so there was just a plate over top of it?
Yeah, as I recall, there was a small metal plate on the underside of the CPU, surrounded by the pins. However, we were unable to remove it until we took a screwdriver (or maybe a chisel, I'm not sure), placed the tip of it on the ceramic packaging, and smacked the end of the screwdriver with a hammer. This broke off a large chunk of the packaging (without damaging the core) and we were able to pry off the metal plate by hand. You'd be surprised just how fine the wires in there were, too, especially for such relatively old technology.

[Pootie]

Yeah, as I recall, there was a small metal plate on the underside of the CPU, surrounded by the pins. However, we were unable to remove it until we took a screwdriver (or maybe a chisel, I'm not sure), placed the tip of it on the ceramic packaging, and smacked the end of the screwdriver with a hammer. This broke off a large chunk of the packaging (without damaging the core) and we were able to remove the metal plate. You'd be surprised just how fine the wires in there were, too, especially for such relatively old technology. Well it just plain amaze's me that they can put all those parts in that tiny thing... Maybe I'll go smack one of my old intels to see what I can see :P

[Forbin]

Well it just plain amaze's me that they can put all those parts in that tiny thing... Maybe I'll go smack one of my old intels to see what I can see :P
My understanding is that they don't actually put them in, rather they carve them out using light and some nasty chemicals. A wafer of chips starts out as a large disc of several layers of material and nothing more.

[P5YcHoM4N]

Well it just plain amaze's me that they can put all those parts in that tiny thing... Maybe I'll go smack one of my old intels to see what I can see :P
The parts aren't put in, as persay. But cast in. As part of one of my courses at uni we have to look at the architecture of chips, and how silecon is bruned. And for some reason the write of the book (W. Stallings) has an Intel fetish, and slight Intel Fanboyism as he quotes the Intel P4 3.2ghz as being the fastest CPU ever. Now unless he means clock speed (and I so hope he is) and not preformance. I'd never buy books by this guy if I was any of you.

[ColeusRattus]

My understanding is that they don't actually put them in, rather they carve them out using light and some nasty chemicals. A wafer of chips starts out as a large disc of several layers of material and nothing more.

That is about right... As I am working in semiconductor business (oh, I love the sound of that :D), I can roughly describe the process. It is basically a lithography (in fact, it is quite a similar process like used in modern printing plants). The silicon is covered in a photosensitive "paint" which is exposed to the circuit diagrams. The paint which was hit by light gets hard, and the rest is washed off. Then acids do the rest...

And concerning the sizes, it is quite impressive. The layouts of chips which are about half the size of a thumbnail are about 2 squarefeet in size, just that so that yo can make out single parts with the naked eye...

[AndroidXP]

Ahh, the memories of building a micromaster in school :irked: :D

I'm so glad I don't have to deal with this anymore, even though I was one of the few who actually understood Assembler to a degree and could program it. I'll stick to my trusty C#, though.

[Shotglass]

Then came the Pentium Pro. With this one, Intel took the L2 cache and put it directly on the CPU die where it ran at full CPU clock frequency. This made it rather large and expensive to produce, so it ended up being used mostly in servers.

not quite the cache was actually on a seperate die on the processor case
and another reason why it never really hit the desktop market was because it wasnt a pentium but completely different core with noncompetitive 16 bit performance (which at the time was stil the selling point for desktop users

Pentium 3 then saw the reintroduction of the on-die cache, if not with the first ones then at least shortly before the P3 moved back to a socket format. x86 CPU's from Intel have had on-die caches ever since.

if im not mistaken the first "real" pentium series chip to feature on die cache was the legendary celeron 300a

Well, presumably not, but you try finding it. In fact, try finding it in my P2-266 which has run (and is still running) just about every day since 1998 and hasn't been cleaned in about five years. Take the lid off that computer and you'd think I'd found it in a swamp.

i took the case of my old p2 233 (now replaced by a nifty 333 which is still my internet-surfing machine) apart and it was pretty clean except for a lump of dried out thermal paste (couldnt get the case completely open so it ran for a few year without any cooling and the front metal plate hanging off the pcb only held by the last 2 bolts)
was a fun time though since you cant secure the cpu in the slot (unless you somehow find a celeron rentention mechanism) so it fell down a few times only stoped by the graphics card when a whacked my pc cause the psu fan was rattling again
blowing air into the case with a compressor works wonders btw ... my pc looks almost brand new

btw does anybody know what kind of capacitor c33 on a p2 board is ? it somehow went awol when i took mine apart

[Anarchi-H]

:O Maybe it would make a nice keychain or some shit :/ I have several,maybe I will make a clock :)

I made a keyring once.
In case you can't make it out, it was a 2100 athlon.
And before anyone starts crying, it was dead before I got my mits on it :p

[Pootie]

Wow,that thing was trashed out eh ??:P wtf happened to it? get hot and explode :P

[Anarchi-H]

lol, no... It was mint (except for having only stubs for pins) when I first 'converted' it.

Thats what it looks like after a year of jangling about in my pocket with my wallet, change & keys, not to mention having my boney arse sit on it several thousand times :p

The die is very delicate , and I couldn't find anything hard enough to coat it with that wouldn't make it look lame

I want an 800-1.2 socket 370 celly / P3 next cos they have this really cool blue die. I'm not sure which core it is, but I've a few dead ones somewhere; I'll see if I can find em :p

[P5YcHoM4N]

lol, no... It was mint (except for having only stubs for pins) when I first 'converted' it.

Thats what it looks like after a year of jangling about in my pocket with my wallet, change & keys, not to mention having my boney arse sit on it several thousand times :p

The die is very delicate , and I couldn't find anything hard enough to coat it with that wouldn't make it look lame

I want an 800-1.2 socket 370 celly / P3 next cos they have this really cool blue die. I'm not sure which core it is, but I've a few dead ones somewhere; I'll see if I can find em :p

I have both. My lappy had a 766 Cel when I got it. So recently I dropped in a 866mhz PIII. I was sent a 1ghz PIII-M, but it was too small, and before that I was sent a 1.4ghz Cel. The mobo only support 1ghz. So I had to send it back to my mate. What sucks is one guy still owes me a 900mhz PIII D:

[steve]

you know your a nerd when..

[tristancliffe]

...you have a hamster shagging a computer mouse as an avatar on a racing simulation forum?

:D

[Anarchi-H]

...you have a hamster shagging a computer mouse as an avatar on a racing simulation forum?

:D

lol; y'know, I reckon steve modded that mouse with a glory hole or something cos last time I checked, the intellimice didn't have a little hole for mister hamsters tool :D