as you can see from my avatar, i am a [email protected] freak. i know little about SETI, but i am wondering how amd fares versus intel in SETI crunching speed. if you are not aware, amd cpu's are faster than intel cpu's at Folding. but, i think i have heard the edge goes to intel in SETI. am i wrong? thanks for the insight.
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
amd vs. intel seti performance
- Thread starter KFB
- Start date
![[P3D]AndrewM](/forums/data/avatars/m/15/15359.jpg?1626886267){kind=avatar}
- Joined
- Mar 22, 2002
KFB said:
Yes , you're wrong. AMD rulez here too!
As a simple example you might see what will happen with the Intel Corporation's SETI Team in a couple of months ...
- Joined
- Oct 5, 2001
- Location
- Ottawa, Canada
So far the Northwood are crunching faster than the XP's now. TC and his 2.2 scored 2.6 hours stock speed, while overclocked XP's do it around 3hours. Although, there Thoroughbred comes out soon, and we'll see how it competes also.
Edit:
The P4 Williamette can crunch as fast as the Thunderbird which crunches in about the late 3hr range. A Duron Morgan DDR will crunch faster than Tualatins (P3, Celeron), because of the DDR advantage.
Yodums
Edit:
The P4 Williamette can crunch as fast as the Thunderbird which crunches in about the late 3hr range. A Duron Morgan DDR will crunch faster than Tualatins (P3, Celeron), because of the DDR advantage.
Yodums
- Joined
- Mar 22, 2002
2,6 hours looks pretty quick for a 2,2 Northwood cpu… What does the workunit’s angle range look like? If it’s the standart 0,417 – then I truly don’t believe it!
According to the “performance calculator” program , the 2,2 Northwood is as fast as a 1,9Ghz Palomino , but that is rather caused by the higher bus speed of the P4.
And finally , how much does it cost to somebody possessing P4 Northwood. I could buy dual Palomino sys for the same money and have two WUs crunched in 3 hours. So AMD rulez , now and ever!
According to the “performance calculator” program , the 2,2 Northwood is as fast as a 1,9Ghz Palomino , but that is rather caused by the higher bus speed of the P4.
And finally , how much does it cost to somebody possessing P4 Northwood. I could buy dual Palomino sys for the same money and have two WUs crunched in 3 hours. So AMD rulez , now and ever!
- Joined
- Dec 18, 2001
- Location
- Prattsville, NY
A white box A7M266 costs $59. w/ free shipping at newwegg.
A 128mb stick of Crucial PC2100 costs $49. also with free shipping or you can shop eBay and pick up a stick of Crucial PC1600 for about $30. that will O/C to 155-166? MHz fsb.
With a 1GHz Morgan Duron for around $45. you should get 5.5-6 Seti wu/day. With an XP 1600+ for around $95. you can expect 7.5-8 wu/day.
I've been buying 300W ATX PS's for $19.95 (250's work just as well) and ThermalTake Volcano 7cu+ for $8.95 at theoverclockerzstore.com but they've disappeared off the map lately.
If you figure $ per wu, AMD DDR is tough to beat.
Harvey
A 128mb stick of Crucial PC2100 costs $49. also with free shipping or you can shop eBay and pick up a stick of Crucial PC1600 for about $30. that will O/C to 155-166? MHz fsb.
With a 1GHz Morgan Duron for around $45. you should get 5.5-6 Seti wu/day. With an XP 1600+ for around $95. you can expect 7.5-8 wu/day.
I've been buying 300W ATX PS's for $19.95 (250's work just as well) and ThermalTake Volcano 7cu+ for $8.95 at theoverclockerzstore.com but they've disappeared off the map lately.
If you figure $ per wu, AMD DDR is tough to beat.
Harvey
- Joined
- Dec 17, 2000
- Location
- San Diego, CA
I dunno about some of those numbers... my P3 with a stock HSF runs WUs in the 3-4 hour range... mostly mid 3s...
with better cooling I could get another 200-300MHz too....
with better cooling I could get another 200-300MHz too....
- Joined
- Mar 22, 2002
3,5h with a P3 ?! Come on – it’s just not possible! Or if it really happens on your sys you should make sure that your cpu is OK. I mean its fpu is working normally.
I think I read somewhere in this site some tips for checking the cpu integrity. Check what the result looks like when you’ve completed the workunit and then run the same wu on another system . If there’re some differences between both results then you’ll know…
If your P3 cpu is OK but somehow is working at 1,6Ghz then I'll be agree. Otherwise you must be cheating....( I hope you're not).
I think I read somewhere in this site some tips for checking the cpu integrity. Check what the result looks like when you’ve completed the workunit and then run the same wu on another system . If there’re some differences between both results then you’ll know…
If your P3 cpu is OK but somehow is working at 1,6Ghz then I'll be agree. Otherwise you must be cheating....( I hope you're not).
Last edited:
- Joined
- Dec 17, 2000
- Location
- San Diego, CA
this CPU will run Prime95 torture test with no errors... 100% stable or there would be no reason to run SETI for me...
WU times:
best time to 100% = 3.17 hours
previous 5 WUs =
3.821
3.442
3.458
3.583
3.822
So, not only possible, but happens on a regular basis... and, as I said, I am not pushing it, yet...
WU times:
best time to 100% = 3.17 hours
previous 5 WUs =
3.821
3.442
3.458
3.583
3.822
So, not only possible, but happens on a regular basis... and, as I said, I am not pushing it, yet...
- Joined
- Jul 17, 2001
- Location
- perth, Australia
probably the extra cache on the Tulatin giving it a good boost
and Kris,look out ,your on my hit list to pass on the stats
and Kris,look out ,your on my hit list to pass on the stats
My [email protected] has a fast time of 2.624hrs but on average it is probably over 3hrs.
- Joined
- Jun 20, 2001
- Location
- Vancouver, WA
- Joined
- Jan 15, 2001
- Location
- Denver, CO
This is starting to look like one of those troll posts in the Intel cpu section. AMD rulez, oh it's not the fastest anymore, okay well it's cheaper then, AMD rulez it rulez 
- Joined
- Mar 22, 2002
JigPu said:
Yes , exactly what I meant! I chose Tualatin (512k) 1266 and what do I see? 4,2 hours! That's right - now that's trustworthy!
- Joined
- Aug 13, 2001
- Location
- Deep Blue Sea (Maryland)
just check out our team's benchmark page. it use the same 0.417AR WU for all the benchmark. so, it's comparable.
crunch on........
crunch on........
- Joined
- Jan 15, 2001
- Location
- Denver, CO
Hmm, what have we here.... a P3S at 1.49GHz crunching a .417AR unit in just a tad over 3 hours. Often finishes slightly higher units well under 3 hours:
- Joined
- Dec 17, 2000
- Location
- San Diego, CA
[P3D]AndrewM said:
I suppose the suggestion above is that I am less than trustworthy...
be that as it may, the above number does verify my claims with this chip (as can TC since I bought it from him)... I am running roughly a 200 MHz OC, so 1/6 more speed than in Roeoff's calculation... so... 6/4.2 = 0.7 and 4.2-0.7 = 3.5 hours...
I have seen 1725 out of this chip, too... just need a better cooler
- Joined
- Mar 22, 2002
TC said:This is starting to look like one of those troll posts in the Intel cpu section. AMD rulez, oh it's not the fastest anymore, okay well it's cheaper then, AMD rulez it rulez
Well ,I suppose you’re a kind of disappointed since you’re the man with the Northwood ( I think your name was mentioned…) Don’t worry – it’s normal !I was also disappointed in the past when I use to spent a lot of money for the Intel’s garbage and everyone was laughing at me…But now I’m not tired to repeat it : AMD rulez ! Indeed , it rulez!
- Joined
- Oct 25, 2001
TC said:This is starting to look like one of those troll posts in the Intel cpu section. AMD rulez, oh it's not the fastest anymore, okay well it's cheaper then, AMD rulez it rulez
I think KFB is seriously wondering...I know I have. If this thread deteriorates into another battle in the unending war between the CPU giants then it'll be because of somebody else.
I have run SETI before and now I Fold because my spread is mainly AMD. I have observed this:
SETI relies heavily on memory bandwidth and even more importantly a large cache. [email protected] will store as much of the cruncher and WU in the L2 cache as it can, so this tends to give many Intel systems an edge with SETI.
Folding relies almost exclusively on FPU performance. AMD processors have a much higher FPU performance level than Intel CPUs so they really pull ahead when it comes to [email protected]
My P3 Xeons will get really good SETI performance due to their massive 2mbs of full speed L2 cache allowing almost the entire WU and client to be kept right there in the processor cartridge. They grind to a near halt with Folding because of low FPU performance though. This is a good example of how important memory times and cache sizes are for [email protected] but how FPU performance determines Folding speed.
For SETI, the cache is the key so Intel is often better. For Folding...as [P3D]AndrewM put it in terms often seen elsew[H]ere, "AMD Rulez".
And please guys, keep this decent....way too many of us are tired of seeing the bickering about who is better, aren't we?
edit- Oops...just realized that I forgot to WELCOMEAndrew to the oc.com forums!
- Joined
- Feb 16, 2002
- Location
- Highlands Ranch, CO
According to the download page at [email protected], the clients have been optimized for Intel chipsets.
http://setiathome.ssl.berkeley.edu/unix.html just above the table for Operating System Compatibility:
The binaries have been compiled with a blended optimization to give good performance on all versions of the Intel chipsets.
My new Dell laptop running a 1 Ghz PIII-M, runs WU's about 20 minutes faster than my 1 Ghz TB running at 1.103. That's with the PIII running on a standard 133Mhz bus and the TB running at 147Mhz.
Being an AMD fan, I can whine about an unfair advantage, but I'd rather stick with the AMD's for the extra performance they bring me for everything else I do with my PC's. I make decisions based on my wants and needs, and others do what is best for them. I don't judge a person by the CPU they use and don't rely on my processor choice to make statements about what kind of person I am.
Can't we all just get along?
http://setiathome.ssl.berkeley.edu/unix.html just above the table for Operating System Compatibility:
The binaries have been compiled with a blended optimization to give good performance on all versions of the Intel chipsets.
My new Dell laptop running a 1 Ghz PIII-M, runs WU's about 20 minutes faster than my 1 Ghz TB running at 1.103. That's with the PIII running on a standard 133Mhz bus and the TB running at 147Mhz.
Being an AMD fan, I can whine about an unfair advantage, but I'd rather stick with the AMD's for the extra performance they bring me for everything else I do with my PC's. I make decisions based on my wants and needs, and others do what is best for them. I don't judge a person by the CPU they use and don't rely on my processor choice to make statements about what kind of person I am.
Can't we all just get along?
Last edited:
- Joined
- Mar 22, 2002
Morpheus said:
I suppose the suggestion above is that I am less than trustworthy...
be that as it may, the above number does verify my claims with this chip (as can TC since I bought it from him)... I am running roughly a 200 MHz OC, so 1/6 more speed than in Roeoff's calculation... so... 6/4.2 = 0.7 and 4.2-0.7 = 3.5 hours...
I have seen 1725 out of this chip, too... just need a better cooler
I see you’re good in math. I hope your cpu is at least as good...
Sorry , but you didn’t mention about any overclock so far! Please , be exact when you express yourself!
Similar threads
