Your computer!

[Captain Picard's Hair]

Vista comes in both 32 and 64 bit versions.

LOL Rochey. Hell, I'm only a mechanical engineer. Imagi e the horror if a REAL computer expert laid down the gory details!

As to the -86 stuff - a while ago Intel chips were designated by numbers; 286, 386, 486, The chip which was to be called '586' was renamed 'Pentium' due to a copyright dispute of some sort (5 =='penta-'). Actually Intel held onto the "pentium" monicker even after it moved past 586 architecture, so pentium 4 was really 786.

I put 'i' before the numbers to signify 'Intel.'

[Grundig]

THAT's what happened! I always got confused as to why that naming convention broke down.

Thanks Ian, that tip fixed my macbook camera!

[Captain Picard's Hair]

Intel actually modified tge older 686 architecture of the pentium 3 series after it retired pentium 4. Hence the latest (Intel) chips are 686-based. The biggest change is hat they are multi-core, ie they really have more than one processor in one physical part installed as if it were a single CPU. For example, core 2 duo chips have 2 CPUs and core 2 quad chips 4 CPUs. This gives you better performance for much less (elecrtrical) power usage than it would take to run a single core of equivalent computational power, and creates far less waste heat as well. There's that word again -efficiency. P4 chips ran headlong into a power/heat wall. I've never seen either of my cores go hotter than 35C or so - downright chilly in the electronic s world.

[IanKennedy]

My work machine is a 2GHz CoreDuo iMac with a 250GB HD and 3GB RAM. My laptop is a 2.4GHz Core2Duo with 200GB HD and 4GB RAM and I've a little Mac Mini under the tv with a 2.0GHz Core2Duo 500GB HD and 2GB RAM. I also have a 340GB external HD.

[Captain Picard's Hair]

I'll try to explain the 32- vs 64- bit stuff simply (wish me luck)

The CPU is just a little machine that performs lots of little calculations. Whether you're browsing DITL, looking at a picture, editing pictures or multimedia, using office software, watching porn, etc, to the computer it's nothing but mathematical operations on binary digits (0's and 1's). In software, the number of bits refers to the size of each instruction (each operation to be executed by the execution engine inside the CPU), in bits. In hardware, it refers to the size of the registers used to store the instruction.

A register is a little electronic element which stores a set of bits. It can be visualized as a set of boxes, each containing a 0 or a 1 (since the number system is binary). A 32 bit register has 32 boxes and a 64 bit register 64 boxes. You need a bigger resister to hold bigger binary numbers, hence 64 bit processors are needed to run 64 bit insstructions. All modern processors are 64 bit, though they are still backwards-compatible with 32 bit software.

The most important advantage gained with 64 bits vs 32 is the amount of RAM you can use.

To access RAM, the CPU needs to know where in RAM it needs to go to get the data it wants. So, RAM is broken down into lots of "boxes" which each contain a set number of bits of data, each of which has an address. The address has to be stored somewhere, too (in a register). When the computer goes to RAM, it first looks up the address of the box that stores the data it needs to find, and then gets the data from that box. Now, the number of bits in the address determines the number of addresses it can store.

With 1 digit, you can store a number from 0-9. WIth 2, a number from 0-99. WIth 3, 0-999, with 4, 0-9999, etc. The number of digits determines the largest number you can hold.

In RAM, the size of the address is fixed, limiting the number of boxes that can be addressed, and hence, the amount of data you can hold in RAM. It turns out that with 32 bit addresses, you can store up to 4 GB natively (without some kind of jury-rigged system, which wouldn' t work as well as natively supporting a bigger address).

RAM is an important contributor to system performance. Running more programs, or opening more tabs in more windows, uses more RAM. More sophisticated programs need more RAM to run. Already high end computers are going over 4 GB of ram, necessitating more than 32 bits, so they went to 64 bits. With 64 bits, you can hold a sh!tload of RAM, so we won't run out of RAM for a while yet.

REALLY old computers ran on 16 bits. They were limited to so little RAM in today's terms that people laugh at it.

[Captain Picard's Hair]

Right now I'm on the computers on my college's (uni for yous across the pond) campus. The machines themselves are fine, quite up to spec. The peripherals, however, are the worst crap ever. They use the cheapest Dell stock keyboards/mice they could find (a stiff breeze might shatter either one into millions of little pieces of plastic, they're noisy, small, and uncomfortable) and the monitors, aside from being tiny, are horribly worn. There are visible dark spots on many of the monitors where the LCD panel has been damaged, often in the corners. Just awful.

Well, off to class now!

[IanKennedy]

Well, you almost got it correct, however, addressing in computers is done in binary rather than decimal. In binary there are only two states for a number 0 and 1. This is the smallest item in a computer and is called a bit. Now you can join bits together like you can join decimal digits together and assign place value to them. For example the decimal number 12 is 1 ten and 2 units. In binary you can have the same two digits. That gives the following patterns 00, 01, 10 and 11 (which in decimal would be 0, 1, 2 and 3), because you can only use 0 and 1.

Now in decimal the largest number you can store in n digits is given by (10^n)-1 (where ^ is raise to the power of). For example the largest value you can represent with 3 digits is 999, that is (10^3)-1. The same rules apply in the binary world, except instead of 10 there are only two options available so it's (2^n)-1. So really old computers, ie those that where common when I first learned about computers, where 8 bit. That would mean you had a range of numbers from 0 to (2^8)-1 or 255. Now when we say that they where 8 bit that was a bit of a lie, their registers where 8 bit but their address range was 16 bit. The address range allows you too access memory, so the largest number you can store in it's bit size tells you the maximum amount of RAM your computer could have. With 16 bits that would be (2^16)-1, or 65535 bytes or 64KB of RAM.

Later on the PC CPUs came out and they sort of worked in 16 bits, the 8086 for example allowed you to have offset addressing, where one register would select the 'page' you wanted to work with and another would then allow you to access the 64K in that page. This lead to the then famous thought that you could only have 640KB of RAM in a PC, because the original designers couldn't imagine that you would need more than that, so they put all the graphics memory, add in cards and other such stuff just about 640KB and only allowed programs to use the lower part of the address range. The maximum address range with these half and half CPUs was 1024KB or 1MB. The 286 (or really the 80286 was it acually was) was also a fiddle and didn't really improve things that much. It was only when the 386 came around that you had the ability to access larger amounts of RAM. There was a mode on the 386 where you could get true 32 bit memory addressing, which gives you an address range of 0 - (2^32)-1, which is 4294967295 a range of 4294967296 bytes or 4194304KB or 4096MB or 4GB.

Now there have been lots of 64 bit CPUs around for quite some time which should allow you to address 18446744073709551615 bytes or 17179869183GB or about 16exabytes, however, once again the CPU design does not actually meet the description on the box. For example the AMD Athlon X2 has only got a 40bit address bus so can only access (2^40) addresses or 1024GB of RAM.

Now, if you are still with me there's another problem. That is your operating system. It stores memory addresses in locations of a given size, for example most versions of windows from 95 through to XP where 32 bit and thus are stuck with the 4GB limit, even if your computer and CPU can cope with more. Since XP came out there have been versions of Windows that are 64 bit but they are very rare and expensive to buy. Vista is the first to bring true 64 bit availability to the windows world.

[Mark]

Wow. Computer jargon. My eyes just crossed. Ian, how can you possibly remember all of that stuff???

[IanKennedy]

It's my job. It's like asking a car mechanic how they remember all those engine parts.

[Captain Picard's Hair]

Well, you almost got it correct, however, addressing in computers is done in binary rather than decimal. In binary there are only two states for a number 0 and 1. This is the smallest item in a computer and is called a bit. Now you can join bits together like you can join decimal digits together and assign place value to them. For example the decimal number 12 is 1 ten and 2 units. In binary you can have the same two digits. That gives the following patterns 00, 01, 10 and 11 (which in decimal would be 0, 1, 2 and 3), because you can only use 0 and 1.

Now in decimal the largest number you can store in n digits is given by (10^n)-1 (where ^ is raise to the power of). For example the largest value you can represent with 3 digits is 999, that is (10^3)-1. The same rules apply in the binary world, except instead of 10 there are only two options available so it's (2^n)-1. So really old computers, ie those that where common when I first learned about computers, where 8 bit. That would mean you had a range of numbers from 0 to (2^8)-1 or 255. Now when we say that they where 8 bit that was a bit of a lie, their registers where 8 bit but their address range was 16 bit. The address range allows you too access memory, so the largest number you can store in it's bit size tells you the maximum amount of RAM your computer could have. With 16 bits that would be (2^16)-1, or 65535 bytes or 64KB of RAM.

Later on the PC CPUs came out and they sort of worked in 16 bits, the 8086 for example allowed you to have offset addressing, where one register would select the 'page' you wanted to work with and another would then allow you to access the 64K in that page. This lead to the then famous thought that you could only have 640KB of RAM in a PC, because the original designers couldn't imagine that you would need more than that, so they put all the graphics memory, add in cards and other such stuff just about 640KB and only allowed programs to use the lower part of the address range. The maximum address range with these half and half CPUs was 1024KB or 1MB. The 286 (or really the 80286 was it acually was) was also a fiddle and didn't really improve things that much. It was only when the 386 came around that you had the ability to access larger amounts of RAM. There was a mode on the 386 where you could get true 32 bit memory addressing, which gives you an address range of 0 - (2^32)-1, which is 4294967295 a range of 4294967296 bytes or 4194304KB or 4096MB or 4GB.

Now there have been lots of 64 bit CPUs around for quite some time which should allow you to address 18446744073709551615 bytes or 17179869183GB or about 16exabytes, however, once again the CPU design does not actually meet the description on the box. For example the AMD Athlon X2 has only got a 40bit address bus so can only access (2^40) addresses or 1024GB of RAM.

Now, if you are still with me there's another problem. That is your operating system. It stores memory addresses in locations of a given size, for example most versions of windows from 95 through to XP where 32 bit and thus are stuck with the 4GB limit, even if your computer and CPU can cope with more. Since XP came out there have been versions of Windows that are 64 bit but they are very rare and expensive to buy. Vista is the first to bring true 64 bit availability to the windows world.

Thanks. I was in a bit of a rush between classes and purposefully simplified it as much as I could; I know binary to decimal conversions but didn't want to get into that.

[Sionnach Glic]

Ow, my brain.

[Captain Picard's Hair]

Ow, my brain.

Don't you mean, "Ow, my CPU. " Just download the info into your memory banks

[Sionnach Glic]

I would, but they're almost full at this point already.

[Tsukiyumi]

AAAAAARRRRRRRRRRRGGGGGHHHHHHH!!!!!!!!!!!!

It looks like one of my hard drives was also damaged by the humidity after the hurricane. This is getting out of hand here.

[Mikey]

Sorry, dude.