My rig is a beast in x64

[DrPepper]

Originally Posted by Solaris17

its ok pepper. I hate it when people corre t me personally. zomg bro x86-32bit. ya i know. but everyone else in the universe understands when I say x32 or x64

Well at least everyone kind of understood what I meant when I say x32. I should say 32-bit or x86-32. Anyway I don't mind being corrected as long as people don't be an ass about it, even more so for spelling and grammar.

[Solaris17]

Originally Posted by InnocentCriminal

I honestly still don't understand the point of the thread, are we asking for proof here other than 3DMark scores?



But don't you want to be technically correct? People that don't know that x32 is actually x86 are misinformed.

nope dont need to be. Its like a type of short hand. I know the difference. As such I perfectly understand what people are doing. People that absolutely need the differentiation are wannabe programmers. Or super obnoxious real programmers. i refuse to type out

X86-32bit

or

x86-64bit

simply to appease those people. IMO if its that important to them they should either

A get a life
or
B check themselves. If they want to use proper terminology visiting this forum is the wrong place to be.

[Fourstaff]

Sorry, feeling a bit touchy due to this thread

[Solaris17]

Originally Posted by Fourstaff

Sorry, feeling a bit touchy due to this thread

not your fault you werent a dbag or assumed we were retarted.

[DrPepper]

Originally Posted by Fourstaff

Sorry, feeling a bit touchy due to this thread

I didn't notice anything wrong. Infact I was confused at first.

[qubit]

Originally Posted by FordGT90Concept

Your benchmarking clearly shows there is.

There's some games that perform better in XP (TrackMania being an example as well as ARMA II).

There's others that show virtually no preference like Call of Duty 4, Left 4 Dead, Fallout 3, and Dawn of War II.

And some that perform better in Vista/7 like Crysis, Far Cry 2, S.T.A.L.K.E.R.: Clear Sky, and Resident Evil V.


If you're looking for a gaming performance upgrade, the answer isn't a different OS (except in the case of tesselation on DX11). A better GPU will generally increase the FPS no matter the OS.

Well hey, out of that lot, I've only got cod4 & L4D, so I can't comment on the others. I'll have a crack at benching those two and see what results I get. I'm fairly confident they'll be similar to TrackMania, but I'm happy to be proved wrong.

[Robert-The-Rambler]

I still don't get what the excitement is totally about. It looks like the differences are just that when you switched to 64 bit you didn't have the Video RAM shadowed into the System RAM and had a little more memory to work with and got a little more performance. I guess......

[DrPepper]

Originally Posted by Robert-The-Rambler

I still don't get what the excitement is totally about. It looks like the differences are just that when you switched to 64 bit you didn't have the Video RAM shadowed into the System RAM and had a little more memory to work with and got a little more performance. I guess......

Isn't that a DX10 feature that it doesn't shadow Vmem to the RAM.

[FordGT90Concept]

Originally Posted by DrPepper

Well at least everyone kind of understood what I meant when I say x32. I should say 32-bit or x86-32. Anyway I don't mind being corrected as long as people don't be an ass about it, even more so for spelling and grammar.

x86-32 doesn't exist either.

Windows (the desktop and server variety) only runs on three architectures:

-x86 (32-bit with 16-bit legacy support). Sometimes refered to as IA-32 (Intel Architecture 32-bit).

-x86-64 (64-bit with 32-bit and 16-bit legacy support). It has two subtypes: AMD64 and EM64T. They are generally cross compatible. Technically, the processor is still 32-bit but has been extended to be able to address >32-bits worth or RAM. Some processors, for example, are 36-bit capable and others are 48-bit. No exisiting x86-64 processor is a true 64-bit processor.

-IA-64 (64-bit with 32-bit legacy support). Not compatible with x86 or x86-64. Only found in datacenter or super computer server systems. It is slated to be discontinued by Intel and Microsoft soon.

[DrPepper]

Ah so it's just x86. I assumed x86-32 based on x86-64.

[Aleksander]

The meaning of this thread:
There was the beauty and the beast.
Beauty was in love with the beast.
The beast was a beast.
My rig is a beast, but only in x64, cuz there is the beauty.
Thank you all for missunderstanding the phrases

[qubit]

Originally Posted by FordGT90Concept

x86-32 doesn't exist either.

Windows (the desktop and server variety) only runs on three architectures:

-x86 (32-bit with 16-bit legacy support). Sometimes refered to as IA-32 (Intel Architecture 32-bit).

-x86-64 (64-bit with 32-bit and 16-bit legacy support). It has two subtypes: AMD64 and EM64T. They are generally cross compatible. Technically, the processor is still 32-bit but has been extended to be able to address >32-bits worth or RAM. Some processors, for example, are 36-bit capable and others are 48-bit. No exisiting x86-64 processor is a true 64-bit processor.

-IA-64 (64-bit with 32-bit legacy support). Not compatible with x86 or x86-64. Only found in datacenter or super computer server systems. It is slated to be discontinued by Intel and Microsoft soon.

Your bit in bold is wrong, GT. The part of the CPU that gives it its size is the word size. That is, the amount of data it can hold in its main data processing registers in one go, not the address bus size. So, a 32-bit CPU can hold 32-bits and a 64-bit CPU 64-bits at once. This means that a 64-bit CPU can chomp through twice as much data in one go than a 32-bit one. They are therefore literally twice as fast as 32-bit CPUs in a straight run, keeping all other factors such as clock speed, instruction cycle time etc the same. One often doesn't see a speedup in the real world, because they are either working in 32-bit mode, or they are not working on large data sets where you would notice the difference.

They would be twice as fast for example, in copying large blocks of memory to other large blocks of memory, as they will grab twice as much in go.

The size of the address bus is often confused as defining the size of the CPU. It doesn't. The x86/x64 architecture specifies a data bus that has the same width as the address bus, but there's no reason why this must be so and many other processors don't (think of the old 8-bit 6502 & Z80 CPUs that had 16-bit address buses for example). What you've seen is that the full 64-bit address bus isn't physically implemented, so you get 36 bits, 48 bits or whatever. However, it is logically implemented in all x64 CPUs. x64 CPUs all have a 64-bit data bus. If you don't believe me, have a look at the spec sheets on Intel's or AMD's websites, or even look this up on Wikipedia; it'll tell you the same thing.

Technically, you could have a 64-bit CPU with a 1-bit address bus and it would still be a true 64-bit CPU ie it's registers would hold 64-bits in one go. You'd have to be nuts to make one of course... lol Just to confuse things a little further, the physical data bus doesn't have to be the same size as the word size of the registers. Making it narrower would simply cause the CPU to take extra clock cycles to store and load data to memory and therefore significantly hurt its performance. However, it would make it cheaper to manufacture and cheaper to manufactur the motherboard as it would need less parallel circuit traces.

The fact that they will handle smaller word sizes such as 16 & 32 bits for compatibility is also incidental and not relevant.

This is stuff that I know like the back of my hand, so if you would like me to explain anythging else, please ask away!

[Aleksander]

+1 for qubit. I think this was the reason of the thread and the answer to my first question :P

[FordGT90Concept]

Originally Posted by qubit

Your bit in bold is wrong, GT. The part of the CPU that gives it its size is the word size. That is, the amount of data it can hold in its main data processing registers in one go, not the address bus size. So, a 32-bit CPU can hold 32-bits and a 64-bit CPU 64-bits at once.

Not necessarily. For example, all recent x86 CPUs (e.g Intel Core, AMD Athlon XP) can handle 64-bit calculations although they are not capable of addressing more than 32-bit memory.

Originally Posted by qubit

This means that a 64-bit CPU can chomp through twice as much data in one go than a 32-bit one. They are therefore literally twice as fast as 32-bit CPUs in a straight run, keeping all other factors such as clock speed, instruction cycle time etc the same. One often doesn't see a speedup in the real world, because they are either working in 32-bit mode, or they are not working on large data sets where you would notice the difference.

They would be twice as fast for example, in copying large blocks of memory to other large blocks of memory, as they will grab twice as much in go.

...

This is not the case with x86-64. 64-bit is treated the same as 32-bit in a 64-bit environment. 64-bit is not faster than 32-bit (where it is in undeniably 64-bit architectures like IA-64).

Edit: double checking this... might have different benchmarks turned around.

[qubit]

Originally Posted by Aleksander Dishnica

+1 for qubit. I think this was the reason of the thread and the answer to my first question :P

Awsome dude, thankyou.

[qubit]

Originally Posted by FordGT90Concept

Not necessarily. For example, all recent x86 CPUs (e.g Intel Core, AMD Athlon XP) can handle 64-bit calculations although they are not capable of addressing more than 32-bit memory.

This is not the case with x86-64. 64-bit is treated the same as 32-bit in a 64-bit environment. 64-bit is not faster than 32-bit (where it is in undeniably 64-bit architectures like IA-64).

You're still confusing the address bus with the word size. 64-bit is indeed inherently faster in example I gave you and other similar types of tasks. It's like painting a wall with a brush and then comparing it's performance with one that's twice as wide. The wider one will complete the paint job in half the brush strokes and therefore half the time.

The width of the memory doesn't matter either: all that counts is that the CPU can grab 64-bits of data off it to fill its registers. If the data bus is narrower, because of RAM config for example, then it just hurts performance. It's still working in a full 64-bit mode.

I strongly recommend you look this up on Wikipedia etc if you disagree with me and you'll see I've explained this subject exactly. I know this stuff, because it's been my hobby for years and I also learned it formally in higher education too. No need to go back and forth on this thread about it.

[FordGT90Concept]

Note how 64-bit is actually 50% faster than 32-bit:

UInt averages:
  [0]: 2164491716
ULong averages:
  [0]: 3025861192

If what you said was true, 32-bit should be double that of 64-bit.

[Aleksander]

So in theory, x64 is a kind of hyperthreading. Like upgrading from core i5 to core i7 ????!!!!

[FordGT90Concept]

No, HTT has nothing to do with width.

[qubit]

Originally Posted by FordGT90Concept

Note how 64-bit is actually 50% faster than 32-bit:

UInt averages:
  [0]: 2164491716
ULong averages:
  [0]: 3025861192

If what you said was true, 32-bit should be double that of 64-bit.

It is double. I don't know what that example is; not enough detail. Once you start adding things like constant factors for timings for loops etc it alters the results, so you don't usually end up with a perfect doubling of performance.

I'd really appreciate it if you looked up the basic architectural info on the internet before coming back to me. You'll understand what I'm explaining then and we can have a more constructive discussion.

[Aleksander]

@Ford
Oh well, i still get a performance boost and that is what interests me, not the numbers. If it is 50% faster as you say, than all people would jump to 64-bit systems. Like pepper said, 250 points is not a big deal. It is not even 10% :P

[InnocentCriminal]

Originally Posted by Solaris17

A get a life
or
B check themselves. If they want to use proper terminology visiting this forum is the wrong place to be.

Why is it the wrong place? An IT forum is a great resource of information, if that information is wrong then obviously someone is going to amend that for the greater good - which is sharing and learning from accurate knowledge. Apart from people using the right terminology because it's right, using proper grammar and spelling can help other people in different ways, for example; for those that don't have English as their first language.

Anyroad, back on topic, whatever that is.

[FordGT90Concept]

Originally Posted by qubit

It is double. I don't know what that example is; not enough detail. Once you start adding things like constant factors for timings for loops etc it alters the results, so you don't usually end up with a perfect doubling of performance.

The processor is ordered to count up, nonstop, and is measured on 1 second intervals. Every value given is the average of 10 readings. I ran the test 6 times. 32-bit was more or less equal for all tests. 4 of 6 64-bit exceeded 3 billion. 2 of 6 64-bit exceeded 1.8 billion (most likely, something else loaded the core during those tests making it anomalous).


I'm not arguing that the general purpose registers weren't extended because they were. The GPR more or less defines what the processor is marketed as.

x86-64 can be faster than x86 if the 8 new registers (r8, r9, r10, r11, r12, r13, r14, r15) are put to work.

[ShiBDiB]

Originally Posted by Aleksander Dishnica

The meaning of this thread:
There was the beauty and the beast.
Beauty was in love with the beast.
The beast was a beast.
My rig is a beast, but only in x64, cuz there is the beauty.
Thank you all for missunderstanding the phrases

havent we established the only reason its faster is because u switched to xp? has nothing to do with 32/64

[Robert-The-Rambler]

Originally Posted by DrPepper

Isn't that a DX10 feature that it doesn't shadow Vmem to the RAM.

In a 64 bit OS you still have access to most of your available System RAM without Video RAM duplication. In DirectX 9 you lose address space roughly equivalent to the Video RAM of your Video Card but is not shadowed resulting in losing access to double the video card's memory in System RAM. Here is a link to the thread here at TPU.

http://forums.techpowerup.com/showthread.php?t=91260

Here is an excerpt.

Under a 32 bit OS, you have 4GB total (OS) and 2GB per application.

Under a 64 bit application, 32 bit apps are still capped at 2GB - but if its large address aware (the 2GB+ flag i mention above) then it can use upto 4GB of ram/address space per program.

Under a 32 bit operating system (XP, vista 32) you have 4GB of address space available. Address space is different to the amount of memory in your PC.

The reason a 32 bit system can only use 3GB (or 3.25GB, or whatever number you get) of system ram is because it doesnt have enough address space left. Video cards are the most important part of a PC that uses address space.

If you had 4GB of system ram and a 1GB video card under a 32 bit operating system, each individual program could only use 3GB of that system ram (due to the video card using 1GB of address space) However there is something else most people are NOT aware of.
Under DirectX 9.0C (and lower) video card ram must be duplicated into system ram. That means if you're running on the highest settings with your new shiny 1GB video card - that 1GB of video memory must be duplicated leaving you with only 2GB left for your game.

You just went from 4GB to 2GB, only considering a single 1GB video card. Things only get worse in SLI and crossfire.

Under 64 bit you wouldnt have lost that initial 1GB of ram to the address space, so you'd have 3GB of usable ram, with 1GB used in DX9.0C games. All of a sudden those modern games which border on 1.5-2GB of ram usage are playable, without your system running like a dog.

Side note: It should be noted that DX10/10.1 does not duplicate video memory into system ram. DX10 actually helps to alleviate this issue, if your system is powerful enough to run games in DX10.

Side note 2: There is more than just video card ram that affects this. System page file uses address space, as do various parts related to the BIOS (RAID cards, sound cards with onboard ram, etc) - this is why with a 512MB video card your 32 bit OS may report 3.25GB of ram - 256MB was taken away for everything else