- Joined
- Aug 31, 2009
.............
-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
why does super pi, finish calculations so quick?
- Thread starter umarrehman187
- Start date
- Joined
- Mar 1, 2002
- Location
- Texas
Well. Maybe your computer is just that awesome.
Does it barf out with an error or what?
Does it barf out with an error or what?
- Joined
- Mar 8, 2008
- Location
- Middle of Nowhere, Wisconsin
Yeah its a benchmark. You gauge performance based on the time it took to calculate. The accepted standard is the 1M test.
Intel and AMD results aren't really comparable though because they seem to calculate differently. It's a good way to gauge gains from OCing though.
Intel and AMD results aren't really comparable though because they seem to calculate differently. It's a good way to gauge gains from OCing though.
- Joined
- Mar 8, 2008
- Location
- Middle of Nowhere, Wisconsin
Yeah I know they are but that's not the point. Intel chips MUCH slower than my X3 get better scores in Pi. It's a matter of how they calculate it.
Mind you I've been building for several years. This is my 4th full PC build(not counting dozens upon dozens of customer builds). I know full well which chips were faster
Mind you I've been building for several years. This is my 4th full PC build(not counting dozens upon dozens of customer builds). I know full well which chips were faster
That is the point. BECAUSE they are slower clock for clock in this particular calculation (ALU?) it takes a noteably faster clockspeed for AMD to match Intel. Its calculated the same across the chips. 1M Pi digits to calculate are 1M Pi digits to calculate bro. If you have any LINKS that state it calculates different that I can read, I would be more than happy to stand corrected. But AFAIK, its the same. AMD chips are just slower in that CPU function.
- Joined
- Nov 25, 2008
Ding.
Just gonna pop in here and say, don't get in a which CPU is better argument. Right now the intel chips are faster, just like how the AMD chips were faster in 2004 (Good ol A64 and A64x2s). It changes every few years, get used to it. No use to sit crying over how your favorite CPU company doesn't make the fastest chips, either buy the faster ones or don't complain. AMD chips are cheap as... well chips right now, even though they aren't faster than intel chips they make up for it with price.
God I hate brand loyalty in things like this.
Oh, and intel has always had a thing for superPI. No use getting worked up about that unless you do a lot of benchmarking or are trying to find some obnoxiously large pi number.
QuietIce
Disabled
- Joined
- May 7, 2006
- Location
- Anywhere but there
SuperPi has always been optimized for Intel. Just before the C2D came out Intel could still beat AMD clock for clock in SuperPi.
But that doesn't mean the Intel chip was faster "clock for clock" - it only meant Intel was faster at running SuperPi - period, End of Line
Geez, five years later and still having to explain that ...
But that doesn't mean the Intel chip was faster "clock for clock" - it only meant Intel was faster at running SuperPi - period, End of Line
Geez, five years later and still having to explain that ...
See...now THAT I must admit I didnt know. For that generation, you are correct in regards to performance...SuperPi has always been optimized for Intel. Just before the C2D came out Intel could still beat AMD clock for clock in SuperPi.
But that doesn't mean the Intel chip was faster "clock for clock" - it only meant Intel was faster at running SuperPi - period, End of Line
Geez, five years later and still having to explain that ...
How can that be optimized though? Im not a programmer so I dont get it? Can you explain?
QuietIce
Disabled
- Joined
- May 7, 2006
- Location
- Anywhere but there
I'm not a programmer either (at least, not at that level) but I've seen the results of optimization in SETI and know what an optimized program can do. Optimized rigs perform anywhere from 30-70% better (depending on the CPU model) than a non-optimized twin - and the optimization programs themselves can be different for AMD and Intel even when using the same language set (i.e., SSE3) ...
- Joined
- Mar 8, 2008
- Location
- Middle of Nowhere, Wisconsin
1.475 was a 1.25VID chip. As far as I know up to 1.5 is fine as long as your temps are good.
Don't worry about it. I totally understand your logic, but this is a case of it being optimized for Intel even though it normally shouldn't matter. It's dirty tactics, and it's been working against AMD/ATi for years.
- Joined
- Dec 28, 2009
- Location
- under a heatsink
Is there any chances that Intel had it's hand in making SuperPi optimized for its chips? Or maybe its just that the easiest route to calculating Pi happens to be a calculation path that works very well on Intel chips vs. AMD. Then again... the SuperPi developers could just be Intel fan boys lol
And... regarding AMD vs. Intel (I know its been said a million times before, but I thought I'd bring everyone back down to Earth) Intel is a HUGE company compared to AMD and has a lot more money and time to develop its chips. AMD specifically targets low-mid end consumer PCs with their "e" edition low power chips and also targets anyone building a budget pc. Intel targets pretty much everyone, and has a line of chips for every type of PC user/consumer. Many people argue that AMD has better price to performance compared to Intel, and I agree. However, Intel can name any price they want and because people buy similar chips from Intel for 40% more then AMD they don't have a reason to have comparable pricing. Therefore, Intel and AMD are very different companies as are their chips. So AMD vs. Intel isn't just PIIx4 vs. i7.
And... regarding AMD vs. Intel (I know its been said a million times before, but I thought I'd bring everyone back down to Earth) Intel is a HUGE company compared to AMD and has a lot more money and time to develop its chips. AMD specifically targets low-mid end consumer PCs with their "e" edition low power chips and also targets anyone building a budget pc. Intel targets pretty much everyone, and has a line of chips for every type of PC user/consumer. Many people argue that AMD has better price to performance compared to Intel, and I agree. However, Intel can name any price they want and because people buy similar chips from Intel for 40% more then AMD they don't have a reason to have comparable pricing. Therefore, Intel and AMD are very different companies as are their chips. So AMD vs. Intel isn't just PIIx4 vs. i7.
Last edited:
- Joined
- Jan 7, 2005
SuperPi has always been optimized for Intel. Just before the C2D came out Intel could still beat AMD clock for clock in SuperPi.
But that doesn't mean the Intel chip was faster "clock for clock" - it only meant Intel was faster at running SuperPi - period, End of Line
Geez, five years later and still having to explain that ...
QI,
I have always found SuperPi (32 mb) to be a good memory test used in between Memtest86 V2.xx and Prime 95 / Orthos too.
QuietIce
Disabled
- Joined
- May 7, 2006
- Location
- Anywhere but there
It can be though I'm not sure it applies as much as it used to with 6 & 8 Mb of L3 cache on many CPUs today. If nothing else it's still good for the non-L3 CPUs ...
- Joined
- Dec 28, 2009
- Location
- under a heatsink
Quick semi-off topic question, but how do higher levels of CPU cache effect performance. I am assuming that L3/L2 cache can be used to store temporary mid-step data between cpu-level instructions?
QuietIce
Disabled
- Joined
- May 7, 2006
- Location
- Anywhere but there
Exactly ...
Similar threads
