http://www.overclockers.com/tips00198/
just wondering does anyone have a different oppinion to the 64-bit.
just wondering does anyone have a different oppinion to the 64-bit.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
64-bit
- Thread starter CPL.Luke
- Start date
- Joined
- Jan 2, 2004
I disagree, because extending the general purpose registers to 64 bits also gives some unanticipated benefits.
Example: Stuff two 32-bit integers into the same register and perform the same op. (See also what Intel has been doing with MMX, SSE, but on a grander scale and in the arithmetic unit(s) themselves.)
Scaling MHz-wise is not an issue when architectural improvements are made with each progressive revision. And Newbie's Law No. 3 states:
Larger clock speeds do not necessarily indicate higher performance; and comparisons of clock speeds should ONLY be done when the processors are of the same core design and to be really picky, of the same step code and batch.
Example: Stuff two 32-bit integers into the same register and perform the same op. (See also what Intel has been doing with MMX, SSE, but on a grander scale and in the arithmetic unit(s) themselves.)
Scaling MHz-wise is not an issue when architectural improvements are made with each progressive revision. And Newbie's Law No. 3 states:
Larger clock speeds do not necessarily indicate higher performance; and comparisons of clock speeds should ONLY be done when the processors are of the same core design and to be really picky, of the same step code and batch.
- Joined
- Jun 28, 2002
- Location
- Phoenix, AZ USA
Its a very interesting topic as more bits does not necessarily equate to better performance either. I dug up a great article someone posted a while back.
"A few final words about performance
Note that I attributed the CS performance increase to x86-64's larger number of registers, and not the increased register width. On applications that do not require the extended dynamic range afforded by larger integers (and this covers the vast majority of applications, including games), the only kind of performance increase that you can expect from a straight 64-bit port is whatever additional performance you get from having more memory available. As I said earlier, 64-bitness, by itself, doesn't really improve performance for anything but the rare 64-bit integer application. In the case of x86-64, it's the added registers and other changes that actually account for better performance on normal apps like games."
http://arstechnica.com/cpu/03q1/x86-64/x86-64-1.html
ps that is a great site for CPU info. Take some time and check it out if you can.
"A few final words about performance
Note that I attributed the CS performance increase to x86-64's larger number of registers, and not the increased register width. On applications that do not require the extended dynamic range afforded by larger integers (and this covers the vast majority of applications, including games), the only kind of performance increase that you can expect from a straight 64-bit port is whatever additional performance you get from having more memory available. As I said earlier, 64-bitness, by itself, doesn't really improve performance for anything but the rare 64-bit integer application. In the case of x86-64, it's the added registers and other changes that actually account for better performance on normal apps like games."
http://arstechnica.com/cpu/03q1/x86-64/x86-64-1.html
ps that is a great site for CPU info. Take some time and check it out if you can.
- Joined
- Aug 27, 2002
- Location
- The OCFORUMS
Somebody fire him. Seriously.
Sure, not everyone needs 64bit, but lets dupe them into thinking they do so we can reap the benefits. Someone has to push the envelope, and its people like him that slow everyone else down.
Sure, not everyone needs 64bit, but lets dupe them into thinking they do so we can reap the benefits. Someone has to push the envelope, and its people like him that slow everyone else down.
- Joined
- Jan 2, 2004
Unless the application is heavily renaming registers, then there is no support for the extended registers unless the instruction fetch/decode unit is automatically assigning different arguments to different registers. 
Dude, forget this x86 bullcrap. Let's all go RISC. Right now!
Dude, forget this x86 bullcrap. Let's all go RISC. Right now!
- Joined
- Jun 28, 2002
- Location
- Phoenix, AZ USA
Captain Newbie said:
You can say that again!
- Joined
- Jan 27, 2004
- Location
- Wasilla, Alaska
I Third that motion!! It's time to move on. Everything else in the PC Industry moves FORWARD with technology....why move laterly with processors?!
- Joined
- Jan 2, 2004
XWRed1 said:
I shall check up on the instruction-register thing. Assembly wise, being careful for overflow, I don't think it would be that difficult to do. MMX does it in the 80-bit FPU registers anyway, why not speed it up and move it into the GPRs?
But it's compatibility that has shackled x86 into the hodgepodge mess that it is now.
