# What does 8-bit 16-bit and 32-bit mean?

What are bits? They do something with graphics I think. Eg. in video game consoles some systems have 16-bits or 64-bits…what does that mean? What does it do?

1 bit is equal to a 1 or a 0. 1 byte is 8 bits. We get our ASCII charachter set from the 8 bit ASCII system because you have 256 combinations (0-255). 8-bit color deals with 256 colors. 16 bit is 65,536 combinations (or in the case of color, 65,536 colors). 32 bit is MILLIONS of combinations/colors. Now you see why 64-bit is so awesome?

Sketch

it depends on what “thing” you’re talking about – bits mean different things. for graphics it’s like Sketch said, different combinations of colors. in video game systems and your computer’s CPU, it means how many bits the processor can process per clock cycle. (current PC CPUs are 32-bit BTW).

so bottom line is, though, a bit is binary…a 1 or a 0. bytes are 8 bits. 32 bit represent millions of bits, whatever those bits do is up to the device–CPU, Video card, sound card…whatever.
Sketch

Originally posted by DR_LaRRY_PEpPeR
(current PC CPUs are 32-bit BTW).
Until we get the new 64-bit Intel Itaniums!

an 8-bit microprocessor (CPU) would work on words of size 8 bits… a 32-bit microp on 32-bits… generally the larger the bit-word value the more efficient the cpu is in computation processes because it needs less time to work on more data…

Originally posted by dominique
Until we get the new 64-bit Intel Itaniums!
That is what I am talking about!

BAH! who want’s an Intel processor?! No one in their right mind.

Sketch

I think currently Itanium is something around 800 mhz, the new McKinley should be something around 2GHz.

I think by 32 bits mean that a single instruction can have an operator and operand (combined) of upto 32 bits. Higher bits should mean lesser processor cycles. That is the reason that 800 Mhz 64-bit processor will outperform a 32-bit 2+ Ghz processor.

I might be wrong about these things …

Originally posted by Sketch
BAH! who want’s an Intel processor?! No one in their right mind.
I’d rather have a 64-bit Intel processor over a 32-bit processor from anyone for the reasons Umair.ms mentionned! They are MUCH faster!!! Nevermind the fact that no software is written for the new 64-bit processors!

LMAO! Well…you could…hmmmm…I don’t know. Software??? BAH! Who needs software?!

Sketch

Originally posted by Umair.ms
[B]I think by 32 bits mean that a single instruction can have an operator and operand (combined) of upto 32 bits. Higher bits should mean lesser processor cycles. That is the reason that 800 Mhz 64-bit processor will outperform a 32-bit 2+ Ghz processor.

I might be wrong about these things … [/B]

i took a course on Microprocessor Architecture and i can safely say that you’re right!

Actually, I don’t think an 800 mhz 64 could outprocess a 2 ghz 32 simply due to the fact that with the increased processing capacity there will be some addition overhead involved.

So, even though a 64 bit processor is theoretically equivalent to a 32 bit processor twice its speed, it won’t be able to perform quite as well due to the additional overhead involved.

Originally posted by dominique
Nevermind the fact that no software is written for the new 64-bit processors!

lol. No kidding.

Go ahead, get a 64bit chip the day they come out. You won’t be able to do a THING until software makers recompile their apps on the 64bit chip.

having a hard time getting your apps/hw/drivers working for XP? lol, you aint seen nothin yet

Originally posted by Umair.ms
I think by 32 bits mean that a single instruction can have an operator and operand (combined) of upto 32 bits.

Absolutely! But in agreement to AES, remember, “smaller is faster”. More registers also means an increase of clock cycle time. The current ISA uses 32 registers based on this principle.

On a more general note and as an aside, bit patterns are ambiguous. For example, 10011 could mean 35, -13 in 2’s complement representation, or the opcode to declare an R-type MIPS instruction. In other words, bit patterns are arbitrary. It’s the implementation that gives them meaning.

Originally posted by Sketch
[B]BAH! who want’s an Intel processor?! No one in their right mind.

Sketch [/B]

Too right! Intel buyers are balmy!

The problem with all IBM PC clones (x86) is that they’re all designed to be backward compatible with 16-bit mode of the 8086 (which was a jump forward over the 8-bit mode of the 8088), and even most operating systems do much of their stuff in 16-bit mode.

In fact, extra overhead is used by modern chips in order to convert new instruction sets to old instruction sets, etc etc, … Sometimes I just think we’d all be better off with a G4. Almost.

Intel uses the term macroprocessing for the new 64-bit processors. And I think that Windows XP 64-bit should be out soon and a few other operating systems. Intel says that existing programs will have to be recompiled on the newer 64-bit systems. And the existing 32-bit software will not run as effeciently on the 64-bit machines.