| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Retired Forums
Forum: Member-to-Member Support [Read Only]
Thread: Compare Intel to AMD |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 7
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I've been running a single CPU AMD 2800 with 1G of memory and 10GB of disk space dedicated to the client. I'm also running a dual AMD MP workstation with 2800 CPUs, 1G of memory and 10 GB of disk space to the client, and a system with a 3.2GHz P4, 1G of memory and 10GB of disk space dedicated to the client. The Intel system has threading turn on but all three systems are running XP and I'm not overclocking. When they are running the client this is all they are doing, otherwise the client is shut down. The Workstation has the twice the number of background services running of the other two systems. My loose observations are that the AMD systems are frankly cleaning the clock of the P4 and my statistics seems to support it for the run time and number of units returned. I've been watching this for over a month. My question is, anyone else seeing the same kind of data? Or am I slipping of the pier with this?
 |
||
|
|
Alther
Former World Community Grid Tech United States of America Joined: Sep 30, 2004 Post Count: 414 Status: Offline Project Badges: 
|
You want to look at the points of each device instead of the results. Work units vary greatly in the length of time they take to run, but points are (more or less) constant.
----------------------------------------One way to measure is to use points per day of run time. You can get this from your device statistics page. POINTS/(D + (H/24)) Using the above formula, my AMD64 3000+ averages 843.9 points per 24 hours of run time. Likewise my 1.6GHz Pentium M laptop averages 751.1 points per 24 hours of run time. 11% fewer points, but at 20% slower clock speed (AMD64 3000+ runs at 2GHz). This would seem to indicate that the Intel gets better points/MHz. Take these measures with a grain of salt however. The points can vary quite a bit depending on how busy your machine normally is. The points also vary greatly depending on your RAM, network connection rating and HD allocation. Unless these are all identical, it's hard to compare devices at all. Another measure you can use is simply your average CPU score as listed by the Agent. Again, if you have a very busy machine, it may vary quite a bit, but normally is within a few points every time. e.g my AMD64 3000+ normally scores between 197 and 200, while my 1.6GHz Pentium M has varied between 133 and 155. Now that's a much bigger gap between the systems than by using the points mechanism. In this case, the Pentium M CPU score is 22.5% less than the AMD64 but at 20% slower clock speed. So does this mean the AMD64 is crunching data faster per MHz than the Intel? Funny thing about statistics is you can make them say whatever you want 
Rick Alther
----------------------------------------Former World Community Grid Developer [Edit 3 times, last edit by Alther at Jan 24, 2005 11:41:46 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I know exactly what you mean. Still I'm looking for a way to get an apples to apples comparison, may be it is just a phenomenon of my observation time that I'm seeing.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Funny thing about statistics is you can make them say whatever you want George Canning once said "I can prove anything by statistics - except the truth.". (But there are billions of funny quotes about statistics. And all of them make me laugh  )By the way: i've been studying statistics  Andrea |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Funny thing about statistics is you can make them say whatever you want George Canning once said "I can prove anything by statistics - except the truth.". (But there are billions of funny quotes about statistics. And all of them make me laugh )By the way: i've been studying statistics Andrea Hello Andrea; A saying I've heard that I like is "Liers figure and figures lie" I've been guilty of both.  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi Hairy4Paws
----------------------------------------Simply put, With this Device, AMD blows the doors of Intel. After 100 or so work units the Long protein claim is debunked. (Sizes Average out) Honestly put, the points relate more to your electric bill than contributions. ( "Not to our machines Speed or the amount of actual work it does.") Look at anyones Point Graphs on the member pa...ual amount) of work done. At the Grid.org we mostly laughed about it. Getting from the beginning to the end of the Batch, in the real world, is all that counts. The points are just silly motivational tools like our Teams! OK Bring it on LOL Robert [Edit 1 times, last edit by Former Member at Jan 24, 2005 2:34:39 PM] |
|||
|
||||
|
RT
Master Cruncher USA - Texas - DFW Joined: Dec 22, 2004 Post Count: 2636 Status: Offline Project Badges: 
|
I put a bunch of reports together to rate the various processers with the intent of answering questions like yours. Please see HERE
---------------------------------------- |
||
|
|
|