AMD Performance rating

[Crepa]

is there a mathmatical formula to calculate the performance in AMD's Chips in comparison to Intel's?.... b/c i was messing around with SiSoftware Sandra and it said that i have a 2200+ with a performance rating of 2652..?and if there is a formula would you please present it to me so that i could maybe make a program for it to distribute to anyone who doesnt want to work out the formula, unless it's far too easy to understand.

[KingJackal]

Kind of - that performance rating is actually a benchmark score taken from a large suite of benchmarks which AMD feel represent th overall performance that a CPU gives.

I'm not sure if AMD have released the exact suite of benchmarks, though I would suspect they have.


The fact that Sandra or whatever gives a rating is kind of irrelevant. Just because a program can give a rating doesn't mean it'll give it correctly - who knows, they may just be back-tracking from the MHz ratings ( which wouldn't work with new 166MHz fsb chips, or chips running an overclocked fsb ). :confused:

[_N_]

Si Soft Sandra has their own PR rating system which nobody really follows anymore, :)

[Chaos Archon]

http://www.overclockers.com.au/techstuff/a_qs_calc/

I use that, but remember fsb comes into play

[Crepa]

Originally posted by Chaos Archon
http://www.overclockers.com.au/techstuff/a_qs_calc/

there are 10 type of people in this world, those who understand binary and those who dont..

it would actually be 0010, and there are really 0011 types of people in this world, those who know it, those who dont and those who dont care.

[_N_]

>>those who dont care.

but those either know it or not, so therefore they are one of the 0010 types of people.

[fatsanchez]

Originally posted by Crepa
it would actually be 0010,

how exactly, are 10 and 0010 different numbers?

[Grrr!!]

Originally posted by fatsanchez
how exactly, are 10 and 0010 different numbers?

They aren't.

[Crepa]

Originally posted by fatsanchez
how exactly, are 10 and 0010 different numbers?
b/c in binary everything is represented in a set of four numbers ..who's to say that 10 couldnt be 1000 and be 8? How do we know what you think we expect?

[Deviant]

Originally posted by Crepa
b/c in binary everything is represented in a set of four numbers ..who's to say that 10 couldnt be 1000 and be 8? How do we know what you think we expect?

bull, it is common to see it in sets of 4 digits because of hex, but it doesn't have to be.

Oh whats this number:8

or is it 80, or maybe 800.

[KingJackal]

Originally posted by Crepa
b/c in binary everything is represented in a set of four numbers ..who's to say that 10 couldnt be 1000 and be 8? How do we know what you think we expect?

Whoever gave you that idea? And I suppose you think decimal digits only come in blocks of <ARBITRARYNUMBERFOLLOWS>6</ARBITRARYNUMBERFOLLOWS> or something? :confused:

That QS-MHz calculator is interesting, but no longer applies when your fsb isn't 133MHz. Even AMD know that - which is why their new 166MHz fsb chips don't have the same XP ratings per MHz ;).

[NinJa_CookiE]

Originally posted by Crepa
it would actually be 0010, and there are really 0011 types of people in this world, those who know it, those who dont and those who dont care.

that would be BCD (Binary Coded Decimal) system, ever so slightly different than binary :p


BCD:______________Binary:________Decimal:
0001................................1............. .................1
0010................................10............ ................2
0011................................11............ ................3
0100................................100........... ...............4
0101................................101........... ...............5
0110................................110........... ...............6
0111................................111........... .............. 7
1000................................1000.......... ..............8
1001................................1001.......... ..............9
0001 0000.......................1010................... .....10
etc

theres some useless information for yas


*EDIT: since the forums doesnt like spaces or lines or underscores etc my like table i wrote up hasnt come out to pretty, but you get the idea ;)

[_N_]

how is this ten? 0001 0000?

[Deviant]

Originally posted by _N_
how is this ten? 0001 0000?

BCD is always represented like that. Each decimal digit has four binary digits associated with it. For example 1234 would look like this:

0001 0010 0011 0100

Normal binary is converted to BCD for display on 7 segments displays like those in calculators, or watches etc.

[Geek4Life]

So you put the individual digits into a 4 bit binary instead of putting the whole number in binary... weird.

[Sydog]

Originally posted by Crepa
..who's to say that 10 couldnt be 1000 and be 8? How do we know what you think we expect?

please explain to us how 10 could ever be equivalent to 1000, even in BCD binary 10 (binary) would be 0010.

[_N_]

yeah, I don't get it, how can it be 0001 0000 for ten? :rolleyes: someone, please explain for the number ten, why is it 0001 0000

[Deviant]

Originally posted by _N_
yeah, I don't get it, how can it be 0001 0000 for ten? :rolleyes: someone, please explain for the number ten, why is it 0001 0000

/* slap with a big wet fish */

how can you not understand the last few posts explaining it?

[_N_]

oh, I see, it's like in thousands, hundreds, tens, and ones,
:eek: :confused: :)

[O'Neill]

Originally posted by Deviant
/* slap with a big wet fish */
Seconds Slapping

[-=DEI[]v[]OS=-]

This is what I love about this place....we're such geeks that we sit and discuss binary.....I mean really....who throws a shoe anyway:p

[Tiggerz]

Gee I am sure they should be teaching this kind of basic stuff at school shouldnt they ?


Binary by convention is always represented as two 4 bit nibbles in big endian format. Thus 000 0001 would be decimal 1.

01 refers to an octal notation number where the leading character is always a zero. So 010 would refer to 8 decimal.

Hex is always represented as byte pairs so AE 10 would represent 44560 decimal.

There are a couple of other conventions that I need point out. Octal is always assumed when mixing number conventions. The suffic 'b' is appended to binary to indicate that its binary not octal and 'h' is appended to indicate hex.

Thus:
01 would be decimal 1 in octal
01b would be decimal 1 in binary
01h would be decimal 1 in hex.

0101 would be decimal 65 in octal
0101b would be decimal 5 in binary
01 01h would be decimal 257 in hex

The over-riding convention governing this is that we always work in octets that is 8 bits unsigned. So all the above numbers would be correctly represented as:

01 in octal
0000 0001 in binary
01h

01 01 in octal
0000 0101 in binary
01 01h in hex.

To make the assumption that there are two types of person (represented by binary 01). In this example you are using two bit positions so there are technically 4 types of person since there are possibly four different possible states 00, 01, 10 and 11.

The actual inferal to the joke is that you are using binary in which there are two states 0 and 1, therefore 2 types of person. However, this is also technically wrong, because in binary 0 is not a state. Its the absense of state. So technically there can only be one type of person or an unary person to use the technical phrase and is represented by 1b since the absence of state 0b would actually infer a null-person.

However, no-state is a valid condition to be in, so again the joke infers that there are two types of person - an unary person and a null-person. Hence the joke is considered funny :)

[Tiggerz]

Also for those that dont know how the numbering systems work. Here is how its all done.

Binary is read from right to left where each bit that is set is a multiple of 2. The best way of working it out is by writing the following on a piece of paper.

128 64 32 16 8 4 2 1

Then underneath it, write your binary number.

128 64 32 16 8 4 2 1
0 0 0 0 1 0 1 0

For each bit that is set, add up the numbers above it. So decimal 10 is represented as 01010 binary.

Hex is read from left to right using the digits 0 to 9 and letters A to F (always in upper case).

0 1 2 3 4 5 6 7 8 9 A B C D E F

This equates to the numbers, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 ,10, 11, 12, 13, 14, 15.

It gets a bit hairy after that as the number 16 is represented by 10 (pronounced one-zero), then 12 (pron one-two).

So 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 1A, 1B, 1C, 1D, 1E, 1F, 20, 21 etc.

Would represent decimal 16, 17, 18, 19, 20, 21, 22, 23, 24 etc.

Octal is fairly simple as it works in the same way as hex, but only using the digits 0 to 7. Thus 0, 1, 2, 3, 4, 5, 6, 7, 10, 11, 12, 13, 14, 15, 16, 17 represents the decimal values 0 to 15.

[Deviant]

Originally posted by Tiggerz
Binary is read from right to left

Nah, left to right.

This is 8 in binary read left to right:
1000

If you try reading it from right to left you would get Decimal 1.

[NinJa_CookiE]

Originally posted by Geek4Life
So you put the individual digits into a 4 bit binary instead of putting the whole number in binary... weird.


its binary, ofcourse its weird :p

[mird-OC]

*looks at topic*

i cannae hold her captain, she's broken loose!