is my buddy brainwashed?

[plonkersaurus]

Not necessarily true.

For example, the differences between RISC and CISC, to cite a trivial example.

The difference between RISC and CISC is the way in which they choose to implement the emulation of the turing machine.

There is absolutely nothing a RISC machine can do, that a CISC machine cannot and vice versa. It may need more steps to do something (ie, PERFORMANCE) but it will be able to do it.

Thats how you can get MIPS (a RISC processor!!!) emulated on a x86 (CISC) platform.

 

[fraserhutch]

The difference between RISC and CISC is the way in which they choose to implement the emulation of the turing machine.

There is absolutely nothing a RISC machine can do, that a CISC machine cannot and vice versa. It may need more steps to do something (ie, PERFORMANCE) but it will be able to do it.

Thats how you can get MIPS (a RISC processor!!!) emulated on a x86 (CISC) platform.

Not true. Anyways, this is off topic. If you want, we can discuss this in PM.

 

[bennychico11]

Dude, there is no such thing as a processor being more intelligent than another. IT IS ALL ABOUT SPEED. As all a processor does is basically mimic a turing machine.

please tell me you're joking.
that's the whole deal with AMDs...they have been known to process tasks more efficiently.

If speed was everything, then the Celeron processors in my friends Dell machine would be able to kill my AMD 64 machine.
http://www.tomshardware.com/cpu/20041221/cpu_charts-13.html

make sure you read down the list and notice how the speeds DO NOT go in decending order. In fact, there is even an instance where a Pentium 4 1.3GHz does better than a Celeron 2.6!!! HOW IS THAT POSSIBLE!!! THE CELERON IS SUPPOSEDLY TWICE AS FAST IN YOUR MIND!!!!!

http://www.zdnetindia.com/print.html?iElementId=109342
be sure to click on the image links and see how many times over the slower AMD 64s out perform the faster Pentium 4s

and last but not least, ever heard of a front side bus?? do you know the speed different between the 64s "theoretical" FSB and the pentiums??

 

[plonkersaurus]

This topic doesnt really have a topic anyway anymore. I also believe its a fairly informative thread for anyone that follows it.

But I dont see how what i have said is not true.

 

[plonkersaurus]

please tell me you're joking.
that's the whole deal with AMDs...they have been known to process tasks more efficiently.

If speed was everything, then the Celeron processors in my friends Dell machine would be able to kill my AMD 64 machine.
http://www.tomshardware.com/cpu/20041221/cpu_charts-13.html

make sure you read down the list and notice how the speeds DO NOT go in decending order. In fact, there is even an instance where a Pentium 4 1.3GHz does better than a Celeron 2.6!!! HOW IS THAT POSSIBLE!!! THE CELERON IS SUPPOSEDLY TWICE AS FAST IN YOUR MIND!!!!!

http://www.zdnetindia.com/print.html?iElementId=109342
be sure to click on the image links and see how many times over the slower AMD 64s out perform the faster Pentium 4s

and last but not least, ever heard of a front side bus?? do you know the speed different between the 64s "theoretical" FSB and the pentiums??

You have mis interpreted what i have said. I havent said anything about the amount any processor processes per clock. You said that the processors mentioned have different "intelligence" which isnt true. The pentium 4's have a heavily pipelined core so run at a higher frequency due to doing less work per cycle. The AMD's run at a lower frequency but do more work per cycle.

HOWEVER the work that they will ultimately carry out would be identical, neither is more powerful than the other as neither can do more than each other. Being powerful is a measure of the tasks it can achieve, not the speed in which they are accomplished.

When i say speed I dont mean the clock speed, i mean the speed in which it will complete tasks.

 

[bennychico11]

Being powerful is a measure of the tasks it can achieve, not the speed in which they are accomplished.

When i say speed I dont mean the clock speed, i mean the speed in which it will complete tasks.

which is what i was saying!!!
sorry the word intelligence meant something totally separate in your book, but i meant it as "processing a task more efficiently". doing more work per cycle is pretty damn intelligent to me.

A 3.0GHz Pentium 4 versus a 3.0GHz Athlon 64. According to you the AMD will do more work per cycle than the Pentium 4. And thus will be completing more tasks...faster. And thus, is acting more intelligent than the Pentium design. Yes, I know the chips do not have intelligence...I'm not a moron. But the way they were created to do work is smarter, IMO.

 

[plonkersaurus]

Its actually just the way the engineers split up the instructions, so the processor wouldnt have to wait for 1 piece of data to go all the way through before starting on the next piece. Intel chose to split it up more, AMD less. It allows Intel to crank up the CPU frequency more by using the approach. Its not really more intelligent design, its just a different approach to it.

 

[fraserhutch]

Well, I changed my mind. I don't really have time for this, but I will say this:

Stating that because two GP processors implement a turing machine really says nothing. All algorithms can be said to be turing systems. Granted, it's been 15 years since I've been in university and have studied turing systems, to the best of my knowledge this is true. I don't think they've changed turing systems since then.

The APIs provided by RISC and CISC chips are nowhere near identical. The differences between them need to be made up in software, not hardware. That is the whole point of RISC.

I can state this as a categorical fact, have developed in assembly code on both systems. The point is NOT about turing systems, it is what services the chip provides to the OS. They are NOT the same.

That is how the statement is not true.

 

[plonkersaurus]

I have also done assembly code on both MIPS and 386. The apis are not identical at all. That much is correct. Im not saying that software that will run on MIPS will run on 386 as is. Im saying that the computations that the software implements could also be implemented on either platform, regardless.

Youve managed to change the point away from the computations a computer does, to the instructions it provides which isnt the point at all. The point is, there arent algorithms a RISC processor can compute that a CISC cannot and vice versa. It has everything to do with turing machines.

The only differences you have in todays systems are for external devices (not really relevant to the power of the processor its more the domain on the chipset) and the amount of memory available.
Having bigger registers and more instructions means you get better performance when they are used, but only performance.

 

[fraserhutch]

I have also done assembly code on both MIPS and 386. The apis are not identical at all. That much is correct. Im not saying that software that will run on MIPS will run on 386 as is. Im saying that the computations that the software implements could also be implemented on either platform, regardless.

Youve managed to change the point away from the computations a computer does, to the instructions it provides which isnt the point at all. The point is, there arent algorithms a RISC processor can compute that a CISC cannot and vice versa. It has everything to do with turing machines.

The only differences you have in todays systems are for external devices (not really relevant to the power of the processor its more the domain on the chipset) and the amount of memory available.
Having bigger registers and more instructions means you get better performance when they are used, but only performance.

Yes, ultimately the ame thing can be done on both systems. Hence the term GENERAL PURPOSE processor.
But to say that both cheaps do the same thing is what I quibble with. They implement diufferent instruction sets, and ultimately what is missing from the RISC set needs to get implemented in software in the OS.

So you are the one who is missing the point. You are talking about what gets implemented in both software and the processor, whereas I'm talking about the processor itself. There is no doubt that the CISC process has more "intelligence" built in. Then the "C" is CISC. Can you get there with the RICS via software?. Certainly. VIA Software.

Note that *I* am not talking about efficiency here. I am talking PURELY about what's built into the chip.

You seem to be lumping everything together into the computer as a whole. I'm not. I'm talking PURELY the chip. Hence my statement.

Now, as to which is the better chip - my money actually lies on the RISC chip. But "better" is not what we're talking about here.

And as for giving negative rep when you disagree with someone, I will at least give you props for having the balls to sign it, as I did when I gave you yours.

 

[plonkersaurus]

Yes, ultimately the ame thing can be done on both systems. Hence the term GENERAL PURPOSE processor.
But to say that both cheaps do the same thing is what I quibble with. They implement diufferent instruction sets, and ultimately what is missing from the RISC set needs to get implemented in software in the OS.

So you are the one who is missing the point. You are talking about what gets implemented in both software and the processor, whereas I'm talking about the processor itself. There is no doubt that the CISC process has more "intelligence" built in. Then the "C" is CISC. Can you get there with the RICS via software?. Certainly. VIA Software.

Note that *I* am not talking about efficiency here. I am talking PURELY about what's built into the chip.

You seem to be lumping everything together into the computer as a whole. I'm not. I'm talking PURELY the chip. Hence my statement.

Now, as to which is the better chip - my money actually lies on the RISC chip. But "better" is not what we're talking about here.

And as for giving negative rep when you disagree with someone, I will at least give you props for having the balls to sign it, as I did when I gave you yours.

I havent given negative rep, or any rep of any kind to anyone so i dont know what you mean. (thats completely honest) I would never give someone bad rep for disagreeing with me. It seems someone else has done it for me in my name, and youve just given me bad rep for disagreeing with me.

Nothing ive said is incorrect, and nothing you have said is incorrect, if you look back you are purely trying to prove me wrong by stating what you see from a different viewpoint BUT IN THE SAME PICTURE.

But still, you are talking from the standpoint of a software engineer, I am not, I am talking from the stand point of an end user, which is the most important i feel for the majority of people here. And in which case, the capabilities of both systems are ultimately the same, everything that can be done in 1 COULD be done in another. It doesnt matter to the end user if the instructions end up being done in software or hardware, except for the performance hits that brings. Most of the difference these days is made up in compilers anyway, so all that ends up different at the end is the performance the chip has!

Oh yeah and thanks for the "absolute doofus" remark in the negative rep you gave me.

 

[gordone]

Where did you go to college and what did you major in? I got my BS in Computer Science (minor in Mathematics) at Lafayette and we were required to take "Theory of Computation." I think the AB CS folks got out of this one. Very interesting stuff, although only useful if you want to get your PhD in Comp. Sci and go on and become an academic.

I did become fascinated with NP-Completeness (which actually showed up in a Simpson's episode!) and did my Senior Thesis on the Traveling Salesman Problem. I presented my research at the National Conference on Undergrad Research (NCUR) and was published in their Journal. Now I just write silly little Java windows. Heck, I even use Bubble Sort from time to time, maybe I didn't learn (errr retain) as much as I'd like from College!

Wow, interesting response for your first post here. I actually had to Google that to make sure it wasn't a typo. Very interesting! I wonder why they never covered this in my college computer classes?

http://plato.stanford.edu/entries/turing-machine/

 

[fraserhutch]

But still, you are talking from the standpoint of a software engineer, I am not, I am talking from the stand point of an end user, which is the most important i feel for the majority of people here. And in which case, the capabilities of both systems are ultimately the same, everything that can be done in 1 COULD be done in another. It doesnt matter to the end user if the instructions end up being done in software or hardware, except for the performance hits that brings. Most of the difference these days is made up in compilers anyway, so all that ends up different at the end is the performance the chip has!

I was responding to this statement:

Dude, there is no such thing as a processor being more intelligent than another. IT IS ALL ABOUT SPEED. As all a processor does is basically mimic a turing machine.

You're right - the end user doesn't care how a SYSTEM gets its work done.

A number of times I have stressed that I was responding about the PROCESSOR, not the SYSTEM.

God I love students.... Giver them a course or two....

Oh yeah and thanks for the "absolute doofus" remark in the negative rep you gave me.

I apologize for the rep then, and it's ironic, cause that means i just fell victim to the asswipe (and did the same thing to turnitdown) who is going around forging rep in other people's name.

I'll give you back the rep as soon as the system lets me. People who forge rep should be shot.

 

[plonkersaurus]

I was responding to this statement:

You're right - the end user doesn't care how a SYSTEM gets its work done.

A number of times I have stressed that I was responding about the PROCESSOR, not the SYSTEM.

God I love students.... Giver them a course or two....

I apologize for the rep then, and it's ironic, cause that means i just fell victim to the asswipe who did the same thing to turnitdown who is goign around forgin rep in other people's name.

I'll give you back the rep as soon as the system lets me. People who forge rep should be shot.

ok thanks about the rep, not really sure how the rep system works anyway?

see your picking on me because im a student! everyone does (tries) that!

now can we move on the fpga? just kidding!!

at least i have a post count higher than 0 now!