slackercentral.com messageboard  

Go Back   slackercentral.com messageboard > Computer stuff > 3dfx Refugees and Computer Discussion

 
Thread Tools Reply
Old 02-05-02, 03:47 PM   #1
Kermit
Subscriber
 
Kermit's Avatar
 
Join Date: Dec 2000
Location: Rocklin, CA, USA
Posts: 18,241
P4 vs Athlon XP...

Ok... now I pulled this right off of the AMD site... so it's most defineatly biased... but I can't see how they can lie about hardware... it is what it is.

Hopefully the code flag will keep the spaces if not... forgive me.



Code:
                                                 Athlon XP    Pentium 4
QuantiSpeed™ Architecture        Yes            No 
 
Operations per clock cycle              9             6 
 
Integer pipelines                            3             4 
 
Floating point pipelines                  3              2 
 
Full x86 decoders                           3              1 
 
L1 cache size                               128K     12k µop (Trace cache)
                                                                  +8KB (Data Cache) 
 
L2 cache size                    256KB(on-chip)   256KB(on-chip) 

Total on-chip full-                       384KB       64KB + 12k µop 
speed cache   

Total effective on-chip      384KB(exclusive)  256KB - 12k µop
full-speed cache                                               (inclusive) 
 
System bus speed                     266MHz          400MHz

Ok..... now, looking at the clock cycles, it would seem like the AMD Athlon XP would KILL the P4 performance wise. Doing 3 more operations per tick than the P4 can do is a LOT. Considering that extra 3 operations is 50% of what the P4 is doing total.

Then, you consider the cache... HOLY CRAP! THe Athlon XP again kills the P4 in cache too.

The only thing that the P4 seems to have over the XP is the FSB speed.

SO, I have a few questions. First, is there something I am not seeing here. Is there some archetecure limit that makes those extra 3 operations null in comparison?

If not, then why is the P4 such a big deal? So what if you have a 300 hp engine if you only have 50 lb of torque. Ya know what I am saying?

Also, if real life is all exactly how it looks on paper here, then, do you think AMD is justified in putting a model number such as 1800, or 1900 even though the processor actually is clocked around 1.5 and 1.6 respectivly? I mean, for people like you and me, who know what the processor is capable of, and know that even though it's slower than a P4(mhz wise), the extra operations per tick make all the difference, and would still buy AMD over P4 because of the price for performance ratio. But... ma and pa kettle who just want a PC to surf the net, and play pong or whatever, are going to just see... 1.5 ghz, or 2 ghz, and obviously higher must be better... even if it's not.

So, are they justified in using model numbers instead. I mean... we all know quantispeed is just a marketing gimic... but in reality, maybe they need that to compete with intel. Intel is a household name, AMD really isn't. If you are not computer literate, you are not going to buy a slower AMD because you will just see... slower MHZ. My point is... is it really fair to just rate it by MHZ, when it is so much faster performance wise? It's good for the Computer gerues(Sp) but not for your average user who dosen't know jack about computers.

So, what do you guys think?

EDIT: OK.. the code didn't work too good... but you can get the point.

EDIT: Added pics for your enjoyment.

Last edited by Kermit; 02-05-02 at 03:54 PM..
Kermit is offline   Reply With Quote
Old 02-05-02, 04:02 PM   #2
Huber
Registered User
 
Huber's Avatar
 
Join Date: Jan 2001
Location: Buffalo, NY
Posts: 2,550
the model numbering system is perfectly justified considering artificial benchmarks showing that a AMD XP clocked at 1.53ghz outperforms a 2ghz P4 in integer operations and comes close at FP operations. thats a 500 mhz difference. so apparently those extra operations are working and they are working correctly

also it is up to resellers to explain to the cojnsumer this fact. i know that when someone comes into my store to buy a computer i tell them these facts. they may not fully understand it but they do understand the phrase "equal performance for almost half the cost."

Last edited by Huber; 02-05-02 at 04:04 PM..
Huber is offline   Reply With Quote
Old 02-05-02, 04:16 PM   #3
Coop 1.04
Registered User
 
Join Date: Dec 2000
Location: Planet MF
Posts: 1,601
Quote:
Originally posted by Kermit


The only thing that the P4 seems to have over the XP is the FSB speed.
Actually it's a misonomer. Athlon XP is 133 Mhz but it does two operations per clock cylcle like DDR-RAM. P4 can do 4 operation at 100 MHz like RDRAM. Athlon though is more efficient because it can do in one operation when P4 requires 2.

Quote:

SO, I have a few questions. First, is there something I am not seeing here. Is there some archetecure limit that makes those extra 3 operations null in comparison?
You are comparing an apple to an orange architechturally. What you should be asking is how many clock cylces does it take to do a 32 bit Multiply or some other complex operation. In other words looks at the results .

Quote:

If not, then why is the P4 such a big deal? So what if you have a 300 hp engine if you only have 50 lb of torque. Ya know what I am saying?
The P4 is a marketing gimick. It's designed to run fast and that's about it. Certain elements such as the FPU are designed to underclock IN NORMAL CONDITIONS under heavy load. BTW the number of pipelines doesn't look right

Quote:

Also, if real life is all exactly how it looks on paper here, then, do you think AMD is justified in putting a model number such as 1800, or 1900 even though the processor actually is clocked around 1.5 and 1.6 respectivly? I mean, for people like you and me, who know what the processor is capable of, and know that even though it's slower than a P4(mhz wise), the extra operations per tick make all the difference, and would still buy AMD over P4 because of the price for performance ratio.
It's not fair but so is building a stripped down chip that runs fast.


Quote:

But... ma and pa kettle who just want a PC to surf the net, and play pong or whatever, are going to just see... 1.5 ghz, or 2 ghz, and obviously higher must be better... even if it's not.
Ma and Pa kettle would probably buy the 1.5Ghz because it's cheaper unless Pa Kettle is a hardcore Q3 player.

Quote:

So, are they justified in using model numbers instead. I mean... we all know quantispeed is just a marketing gimic... but in reality, maybe they need that to compete with intel. Intel is a household name, AMD really isn't. If you are not computer literate, you are not going to buy a slower AMD because you will just see... slower MHZ. My point is... is it really fair to just rate it by MHZ, when it is so much faster performance wise? It's good for the Computer gerues(Sp) but not for your average user who dosen't know jack about computers.

So, what do you guys think?
Computer are so fast that performance is non-issue compared to price. Only ethusist like gamer or video editors would care about price performance. These people will do their homework. I think Dell is evil because if it wasn't for their hardcore support of Intel (which they are richly rewarded for) AMD would have a greater marketing share.
__________________

This Space for Rent
Coop 1.04 is offline   Reply With Quote
Old 02-05-02, 04:43 PM   #4
EG
Registered User
 
Join Date: Jan 2002
Location: Corner of Kiss My Butt Ave and No Friggin Way Ln
Posts: 46,075
Blog Entries: 10
which is why the Athlon keeps pace with the P4.....even with the Mhz difference
EG is offline   Reply With Quote
Post New Thread Post Reply


Thread Tools



All times are GMT -5. The time now is 08:07 AM.


Powered by vBulletin® Version 3.8.1
Copyright ©2000 - 2013, Jelsoft Enterprises Ltd.
Everything (at least the good stuff) posted here is © Copyright 2000-2006 slackercentral.com,
and may only be reprinted with permission or by the original author. Real slackers use The Opera Browser.
Best viewed while you're supposed to be working.