• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

64 bit?

N

netsider

Limp Gawd
Joined
Oct 12, 2004
Messages
466
What's the big deal about 64 bit processors and OS's? I mean, what's the reason for them? My Pentium D supports EMT64 bit...should I switch to a 64-bit os? Thanks.
 
I doubt you're going to notice any difference with a CPU that old. As for advantages of 64bit OSes / processors, the main one these days is the ability to address more RAM. A 32bit OS can only handle around 4GB of addressable memory -- that being RAM, GPU memory, soundcard memory, and addresses allocated by the motherboard components themselves (NB/SB, etc). This leads to as little as 3GB of RAM being usable. A 64bit OS can handle a lot more than that; namely, 16 Terrabytes in current 64bit versions of Windows (and a far larger amount possible).

Machines with 4, 6, or even 12GB are not uncommon today, and more RAM is a given in the future. That, and operations with large data sets like un-RARing/un-ZIPing are a good bit faster.
 
Another difference is with the speed of using 64-bit values in programs. Take the same exact program that uses 64-bit values, compile it for both 32-bit and 64-bit use, and the one running in 64-bit will be way faster in the areas it uses the 64-bit values... if it is coded properly that is.
 
There is also 3 times the registers when running in 64 bit mode and the SIMD registers are also larger. Although none of this is a big advantage on the Pentium D. Most of the 64 bitness of that cpu is emulated using 32 bit hardware.
 
cyclone3d said:
Another difference is with the speed of using 64-bit values in programs. Take the same exact program that uses 64-bit values, compile it for both 32-bit and 64-bit use, and the one running in 64-bit will be way faster in the areas it uses the 64-bit values... if it is coded properly that is.
Click to expand...
"way faster" That's quite arguable. An increase? Possibly but definitely not "way" faster.

Anyways, a lot of native 64 bit apps I don't believe benefit much from 'being' 64 bit compared to their 32 bit counterparts. Simply compiling it as 64 bit and actually using true 64 bit word value programming (using the entire 64 integer as code rather than zeros) in the coding are completely different. You could have a 32 bit word value with 32 zeros and be 64 bit and there won't be any benefit to performance.

Anyways, OP, 64-bit OSes are used to allow over 4GB of ram as the second poster suggested. With a modern computer and GPU you're going to have over 4GB of ram, especially with SLI and CrossFireX.
 
Last edited:
bigdogchris said:
"way faster" That's quite arguable. An increase? Possibly but definitely not "way" faster.

Anyways, a lot of native 64 bit apps I don't believe benefit much from 'being' 64 bit compared to their 32 bit counterparts. Simply compiling it as 64 bit and actually using true 64 bit word value programming (using the entire 64 integer as code rather than zeros) in the coding are completely different. You could have a 32 bit word value with 32 zeros and be 64 bit and there won't be any benefit to performance.

Anyways, OP, 64-bit OSes are used to allow over 4GB of ram as the second poster suggested. With a modern computer and GPU you're going to have over 4GB of ram, especially with SLI and CrossFireX.
Click to expand...

Hense why I said "coded properly".
 
Part of the problem with 64 bit especially on the desktop is the lack of native x64 software but that's starting to change albeit still slowly. Office 2010 now has an x64 version and there are some clear benefits for people dealing with REALLY large spreadsheets, Excel 64 bit is limited only by memory. On a powerful PC the analytical power of Excel x64 is fantastic.

However even Microsoft recommends sticking with 32 bit Office for now as most 32 bit add-ins won't work on x64.
 
silent-circuit said:
I doubt you're going to notice any difference with a CPU that old. As for advantages of 64bit OSes / processors, the main one these days is the ability to address more RAM. A 32bit OS can only handle around 4GB of addressable memory -- that being RAM, GPU memory, soundcard memory, and addresses allocated by the motherboard components themselves (NB/SB, etc). This leads to as little as 3GB of RAM being usable. A 64bit OS can handle a lot more than that; namely, 16 Terrabytes in current 64bit versions of Windows (and a far larger amount possible).

Machines with 4, 6, or even 12GB are not uncommon today, and more RAM is a given in the future. That, and operations with large data sets like un-RARing/un-ZIPing are a good bit faster.
Click to expand...

Pentium Pros could address a max of 64GB of memory in 1995, This is a limitation of the OS only because the OS limited it, not because the OS is 32bit.
 
I look forward to the day that all software is finally caught up, although somehow I get the feeling they will always be behind :eek:
 
Sgraffite said:
Pentium Pros could address a max of 64GB of memory in 1995, This is a limitation of the OS only because the OS limited it, not because the OS is 32bit.
Click to expand...

address space has nothing to do with cpu being 32 or 64bit even the old 8086s could address 1Mbyte (up 16 if you decoded the status pins) and they were 16bit processors. The maximum address space for a 32bit value is 4GB f you do not use software tricks like PAE
 
Sgraffite said:
Pentium Pros could address a max of 64GB of memory in 1995, This is a limitation of the OS only because the OS limited it, not because the OS is 32bit.
Click to expand...

Convoluted PAE would've probably sent your system into a tailspin more times than you can imagine. Drivers, drivers, drivers. :p It was a good engineering decision, as time told.
 
In the same step, you can also have file sizes larger then 4GB (on the harddrive, not memory).

This was an issue I ran into where I wanted to store "dd" images of harddrives as one single file. Had to get a 64 bit Linux OS.
 
westrock2000 said:
In the same step, you can also have file sizes larger then 4GB (on the harddrive, not memory).

This was an issue I ran into where I wanted to store "dd" images of harddrives as one single file. Had to get a 64 bit Linux OS.
Click to expand...

as Deacon said, this is a limitation of the harddrives file system.

FAT32 was limited to ~4GB file size, whereas NTFS max file size is 16 exabytes(?)

and you could use either file system regardless of a 32- or 64-bit OS
 
The main reason to go 64-bit is to be able to use more RAM. If you're currently running 32-bit Vista or 7 and have less than 4GB of RAM there's no pressing reason to upgrade. If you're still running XP and you have a 64-bit CPU you might as well buy a 64-bit version of 7 when you upgrade.

Sgraffite said:
Pentium Pros could address a max of 64GB of memory in 1995, This is a limitation of the OS only because the OS limited it, not because the OS is 32bit.
Click to expand...

Physical Address Extention is a kludge. 32-bit programs are still limited to 4 GB of RAM, even if the OS can see more that that using PAE. It's not like it's some sort of conspiracy to get people to buy more copies of Windows, the 32-bit and 64-bit activation keys are interchangeable.

Besides that, 64-bit applications do generally offer some performance improvement and 64-bit Windows has additional security features.
 
Last edited:
for such an old processor I'd suggest upgrading both cpu, system and to W7 x64. Practical benefits are more memory as stated, but even more importantly the system is more resistant to viruses and trojans. For example I enjoying seeing pop-under windows that come up blank. As for the whole "properly coded" thingy, I believe the x64 instruction set is available to all 32 bit apps, just not the memory addressing.
 
jiminator said:
for such an old processor I'd suggest upgrading both cpu, system and to W7 x64. Practical benefits are more memory as stated, but even more importantly the system is more resistant to viruses and trojans. For example I enjoying seeing pop-under windows that come up blank. As for the whole "properly coded" thingy, I believe the x64 instruction set is available to all 32 bit apps, just not the memory addressing.
Click to expand...

The biggest benefit to x64 is *not* increased speed (though some applications actually do show speed gains), but increased stability.

Those applications (including games) that are still x32 are isolated from impacting the x64 layer of the operating system (whether Windows or Linux); the only difference between the two is the method used. Even with that thunk layer, the impact on individual application performance (let alone overall system performance) is minimal to nil.

So far, in two-plus years of x64 OS operation, I have seen a grand total of *one* infected system come in, and in that case, it was due specifically to an infected 32-bit Java runtime (specifically, Java 6 update 11; an old version is known to be infected) that was deliberately let in. Even with that infection, the system in question still ran (which is more than can be said to an x32 system from the same household with the same infection, which would not stay up at all). In short, x64 can withstand much more abuse.
 
YeuEmMaiMai said:
address space has nothing to do with cpu being 32 or 64bit even the old 8086s could address 1Mbyte (up 16 if you decoded the status pins) and they were 16bit processors. The maximum address space for a 32bit value is 4GB f you do not use software tricks like PAE
Click to expand...

That's because the 8086/8088 used segmented addressing trick that split an address across two 16-bit registers to make a 20-bit address space (why didn't they just got for a full 32-bit address space?). If it had used a flat addressing mode, it wouldn't have been able to address more than 64k like a lot of other 16-bit CPU's in it's day. To go beyond 64k in a program, you had to switch segments which required a bit more work.

PAE was another trick to get beyond the 4GB barrier inherit to 32-bit processors that required more work to implement. In a multitasking OS, that almost always led to problems when one program wasn't made PAE aware and just assumed the address 0xDEADBEEF couldn't mean 0x1DEADBEEF rather than the 0x0DEADBEEF it thought it was addressing.
 
You must log in or register to reply here.
Back
Top