64bit question

[Indus Monk]

A 64 bit desktop CPU can handle upto 64 instances of binary digits.

 

^^How much of the above is incorrect?^^

[Xenift]

I just read that, almost gave me cancer.

 

Now you made me confused about x86, great job.

[Indus Monk]

I just read that, almost gave me cancer.

 

Now you made me confused about x86, great job.

then correct me where i am wrong m8

[Xenift]

then correct me where i am wrong m8

Now i don't even know whether i'm correct is the thing .

[Indus Monk]

Now i don't even know whether i'm correct is the thing .

lemme rephrase the post

[Sauron]

straight from wikipedia:

 

In computer architecture, 64-bit computing is the use of processors that have datapath widths, integer size, and memory address widths of 64 bits (eight octets). Also, 64-bit CPU and ALU architectures are those that are based on registers, address buses, or data buses of that size. From the software perspective, 64-bit computing means the use of code with 64-bit virtual memory addresses.

 

read the whole page if you want to know more, it's not something you can easily sum up in 3-4 lines.

 

-edit-

 

most of it comes down to being able to map out more memory addresses, meaning being able to access more ram.

[Xenift]

Hokay, i think that 32 bit is like colours. (British english :3)

 

So think of 32 different colours/alphabets.

 

The more colours/alphabets you have, the more of a sophisticated thing you can make.

 

"A 32-bit register can store 2³² different values." I am by no means an expert, but take my knowledge.

[BrightCandle]

The problem is that it doesn't mean what it used to mean. A modern CPU can do 256 bit floating calculations natively, it can do more than 64 bit in integer math, even the registers are bigger than 64 bits in some cases but the most common ones are 64 bit. Its not even as if we can say the CPU can address 64 bit memory locations from RAM, despite that being the implication a modern Intel CPU only actually has 42 address lines connected so it can't actually address 2^64 at all.

 

So as a statement its mostly garbage and made doubly worse by the fact that a 64 bit CPU has very little about it that is actually 64 bit these days.

[Indus Monk]

Hokay, i think that 32 bit is like colours. (British english :3)

 

So think of 32 different colours/alphabets.

 

The more colours/alphabets you have, the more of a sophisticated thing you can make.

 

"A 32-bit register can store 2³² different values." I am by no means an expert, but take my knowledge.

So each bit a CPU can handle can store one set of binary digits?

[Xenift]

So each bit a CPU can handle can store one set of binary digits?

A bit only has 1 binary value. 1 or 0

A 32 bit system meaning 64 states.

[Indus Monk]

The problem is that it doesn't mean what it used to mean. A modern CPU can do 256 bit floating calculations natively, it can do more than 64 bit in integer math, even the registers are bigger than 64 bits in some cases but the most common ones are 64 bit. Its not even as if we can say the CPU can address 64 bit memory locations from RAM, despite that being the implication a modern Intel CPU only actually has 42 address lines connected so it can't actually address 2^64 at all.

 

So as a statement its mostly garbage and made doubly worse by the fact that a 64 bit CPU has very little about it that is actually 64 bit these days.

you always make things more confusing or worse for me.

[AlTech]

Binary is either in 2 states. ON or OFF. 1 or 0 respectively. Therefore it is wrong since it can handle up to 128 instanced of binary digits (your starting to confuse me lol). Please can you rephrase it

[Indus Monk]

 it can handle up to 128 instanced of binary digits

When did i say that?

[AlTech]

When did i say that?

You didn't I did. an "instance" of binary is either ON or OFF. so therefore 2x 64 = 128 theoretical instances.

[mathijs727]

A bit is either a 0 or a 1.

This means it can hold 2 different values.

A 64 bit computer does it's calculations on values represented by and memory addresses are stored as 64 bits.

This means that it can do calculations with much larger numbers (integers) or represent numbers more precisely (floats).

 

32 bit computers can only represent numbers (and as such memory addresses) in the range of 0 till 2^32 - 1, which equals to about 4 bilion bytes (a byte is 4 bits, memory addresses are in bytes).

This is why a 32 bit operating system (an OS that uses 32 bits numbers for memory addresses) is limited to 4GB of memory.

[BrightCandle]

A bit is either a 0 or a 1.

This means it can hold 2 different values.

A 64 bit computer does it's calculations on values represented by and memory addresses are stored as 64 bits.

This means that it can do calculations with much larger numbers (integers) or represent numbers more precisely (floats).

 

32 bit computers can only represent numbers (and as such memory addresses) in the range of 0 till 2^32 - 1, which equals to about 4 bilion bytes (a byte is 4 bits, memory addresses are in bytes).

This is why a 32 bit operating system (an OS that uses 32 bits numbers for memory addresses) is limited to 4GB of memory.

 

Would be a nice simplification if it were actually true, but it isn't. SSE for example was available on 32 bit CPUs and it supported 128 floating point calculations. 32 bit Xeon's had up to 48 bit addressing with special OS extensions so could address much more than 4GB of RAM. Defining it simply is impossible because the bitness of the CPU just doesn't mean much, it applies to a particular subset of the supported instructions on the CPU and sort of roughly to the RAM in the sense it defines the maximum it could be but not the actual capability in practice.

 

@AmbarChakrabarti The real world of computer hardware is actually complicated, I don't believe in perpetuating ignorance.

[mathijs727]

Would be a nice simplification if it were actually true, but it isn't. SSE for example was available on 32 bit CPUs and it supported 128 floating point calculations. 32 bit Xeon's had up to 48 bit addressing with special OS extensions so could address much more than 4GB of RAM. Defining it simply is impossible because the bitness of the CPU just doesn't mean much, it applies to a particular subset of the supported instructions on the CPU and sort of roughly to the RAM in the sense it defines the maximum it could be but not the actual capability in practice.

 

@AmbarChakrabarti The real world of computer hardware is actually complicated, I don't believe in perpetuating ignorance.

Im talking about the basic architecture.

SSE is an extension to the x86 architecture.