How fast is Pentium D 2.8G with super-pi?

firstep

Member
Jan 3, 2006
27
0
0
Hi, there. I'm considering to get a Pentium D 2.8G machine, or a single Pentium 640 (3.2G with 64-bit support). But I want to know how fast the Pentium D 2.8G is for one core with super-pi 1M run . Is the result similar to P4 2.8G?

If so, it seems a little bit slow. I can get 40s for Pentium 640. I think one single fast CPU is still better than dual-core. Any comments?
 

RichUK

Lifer
Feb 14, 2005
10,341
678
126
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.
 

stevty2889

Diamond Member
Dec 13, 2003
7,036
8
81
Originally posted by: firstep
Hi, there. I'm considering to get a Pentium D 2.8G machine, or a single Pentium 640 (3.2G with 64-bit support). But I want to know how fast the Pentium D 2.8G is for one core with super-pi 1M run . Is the result similar to P4 2.8G?

If so, it seems a little bit slow. I can get 40s for Pentium 640. I think one single fast CPU is still better than dual-core. Any comments?

The results for the pentium-d 2.8 will be the same as they are for a single core 2.8, super PI isn't SMP capable. And why would you base your CPU purchase primarily on calculationg digits of PI?? I don't think I know anyone that uses their computer just to do that?? Anyway, the pentium-d's are a waste of time in my opinion, when they are much slower than Athlon X2's, while running hot. Not sure about the temps on my 920, since I can't get accurate readings, but my 830 was running at around 70c on air, and 47c on water, while my overclocked overvolted X2 runs at 51c on the stock AMD heatsink..


But you also have to let us know what programs you are going to be running in order to know weather or not you would benifit from a dual core at all..
 

firstep

Member
Jan 3, 2006
27
0
0
The reason is simple: super-pi reflects the cpu's computation power that in turns impact the performance/price ratio, a very reasonable factor for purchase decision. Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

There are always heavy-loaded and light-loaded jobs. So actually a wide range of selections are OK. For example, X300SE is OK for normal use. But if there's an X800 deal, why not jump?

Originally posted by: RichUK
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.

 

firstep

Member
Jan 3, 2006
27
0
0
Did I say that number is the unique factor to affect my purchase? :p

Sometimes I need background video compression while working; sometimes I need to run CPU-intensive jobs. So both dual-core 2.8G and single core 3.2G can only partially meet my requirements. In this situation, I need to consider the trade-off on performance and price. That's why I want to know how fast a single Pentium D 2.8G is.

Originally posted by: stevty2889
Originally posted by: firstep
Hi, there. I'm considering to get a Pentium D 2.8G machine, or a single Pentium 640 (3.2G with 64-bit support). But I want to know how fast the Pentium D 2.8G is for one core with super-pi 1M run . Is the result similar to P4 2.8G?

If so, it seems a little bit slow. I can get 40s for Pentium 640. I think one single fast CPU is still better than dual-core. Any comments?

The results for the pentium-d 2.8 will be the same as they are for a single core 2.8, super PI isn't SMP capable. And why would you base your CPU purchase primarily on calculationg digits of PI?? I don't think I know anyone that uses their computer just to do that?? Anyway, the pentium-d's are a waste of time in my opinion, when they are much slower than Athlon X2's, while running hot. Not sure about the temps on my 920, since I can't get accurate readings, but my 830 was running at around 70c on air, and 47c on water, while my overclocked overvolted X2 runs at 51c on the stock AMD heatsink..


But you also have to let us know what programs you are going to be running in order to know weather or not you would benifit from a dual core at all..

 

firstep

Member
Jan 3, 2006
27
0
0
BTW, I'm not an OC fan. I prefer stability than performance.

Originally posted by: stevty2889
Originally posted by: firstep
Hi, there. I'm considering to get a Pentium D 2.8G machine, or a single Pentium 640 (3.2G with 64-bit support). But I want to know how fast the Pentium D 2.8G is for one core with super-pi 1M run . Is the result similar to P4 2.8G?

If so, it seems a little bit slow. I can get 40s for Pentium 640. I think one single fast CPU is still better than dual-core. Any comments?

The results for the pentium-d 2.8 will be the same as they are for a single core 2.8, super PI isn't SMP capable. And why would you base your CPU purchase primarily on calculationg digits of PI?? I don't think I know anyone that uses their computer just to do that?? Anyway, the pentium-d's are a waste of time in my opinion, when they are much slower than Athlon X2's, while running hot. Not sure about the temps on my 920, since I can't get accurate readings, but my 830 was running at around 70c on air, and 47c on water, while my overclocked overvolted X2 runs at 51c on the stock AMD heatsink..


But you also have to let us know what programs you are going to be running in order to know weather or not you would benifit from a dual core at all..

 

RichUK

Lifer
Feb 14, 2005
10,341
678
126
Originally posted by: firstep
The reason is simple: super-pi reflects the cpu's computation power that in turns impact the performance/price ratio, a very reasonable factor for purchase decision. Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

There are always heavy-loaded and light-loaded jobs. So actually a wide range of selections are OK. For example, X300SE is OK for normal use. But if there's an X800 deal, why not jump?

Originally posted by: RichUK
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.


Ok, fine be ignorant. But please explain to me why AMD?s produce worse scores in Super PI, yet they are far superior in both single processing and symmetric multi processing?

Could it be that Intel?s reaped the benefits of the SSE3 implementation into Super PI where as AMD didn?t? hmm. Are we actually talking about a synthetic benchmark that denotes processor performance across an architectural range or all processor architectures ;) Also not taking into account that the PI code is very Intel (instruction set) friendly.

I suppose you also going to disregard the fact RAM performance (latency/bandwidth) plays a big part in the overall Super PI scores aswell :disgust:

Im trying to enlighten you here, so take note ?Super PI scores mean jack?, unless your looking to compare the same processor but at a different frequency, but that is also pointless as the result is obvious (however its good to check RAM timing performance).

Synthetic apps are not definitive, or hold any real value in the real world (unless you like to wave around your E-Peen) !!!!

EDIT: Sorry you asked me a question, so i will answer it the best i can:

Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

Answer: AMD X2 3800+ !!!!!! Now I think you need to do some research ;)
 

stevty2889

Diamond Member
Dec 13, 2003
7,036
8
81
Originally posted by: firstep
The reason is simple: super-pi reflects the cpu's computation power that in turns impact the performance/price ratio, a very reasonable factor for purchase decision. Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

There are always heavy-loaded and light-loaded jobs. So actually a wide range of selections are OK. For example, X300SE is OK for normal use. But if there's an X800 deal, why not jump?

Originally posted by: RichUK
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.

All super PI does it calculate digits of pi, it doesn't simulate real world performance. For video encoding, depending on the program you use, a dual core will be a good bit faster than a single core, but for non-smp aware aps, the dual core 2.8 is gonna be slower than the single core 3.2. Or you could spend ~$60 more, get a 3800+ X2, which will be faster than both of them.
 

firstep

Member
Jan 3, 2006
27
0
0
Interesting. As to the peformance competition, I think Intel and AMD teams are better candidates to argue which one is "superior", and some 3rd-party academia or industry instituition can be the referee. And I'm not an Intel advocate. But, I don't think you can just say AMD is far superior to Intel. I think it's safe to state that neither AMD nor Intel CPU can beat the other in every benchmark. Super-pi is just a popular benchmark. Assuming what you say is right, this benmark is more friendly to Intel. But can I say those benchmark with which AMD does better are more friendly to AMD? What's wrong with "more friendly"?

I'm not building a system, I'm not tuning up the performance. So super-pi just serves as a good enough estimation. I think for the first-order approximation, memory latency can be ignored here as long as the major factors such as DDR 400 or DDR2 533 are determined. Of course you can argue there are more parameters and brand does matter. To me, I just need a rough estimation. So I don't care.

I see you have some AMD system that you are proud of. However, does it really meet your real requirements, or it just get some high scores in 3DMark and you feel satisfied? To measure whether a system really meets personal use, probably you need some log software to record your personalized actions. Then you analyze it and get the conclusion whether a prospective system meets what you need. But I don't think PC users have intention to do that. :)

Originally posted by: RichUK
Originally posted by: firstep
The reason is simple: super-pi reflects the cpu's computation power that in turns impact the performance/price ratio, a very reasonable factor for purchase decision. Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

There are always heavy-loaded and light-loaded jobs. So actually a wide range of selections are OK. For example, X300SE is OK for normal use. But if there's an X800 deal, why not jump?

Originally posted by: RichUK
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.


Ok, fine be ignorant. But please explain to me why AMD?s produce worse scores in Super PI, yet they are far superior in both single processing and Symmetric multi processing?

Could it be that Intel?s reaped the benefits of the SSE3 implementation into Super PI where as AMD didn?t? hmm. Are we actually talking about a synthetic benchmark that denotes processor performance across an architectural range or all processor architectures ;) Also not taking into account that the PI code is very Intel (instruction set) friendly.

I suppose you also going to disregard the fact RAM performance (latency/bandwidth) plays a big part in the overall Super PI scores aswell :disgust:

Im trying to enlighten you here, so take note ?Super PI scores mean jack?, unless your looking to compare the same processor but at a different frequency, but that is also pointless as the result is obvious (however its good to check RAM timing performance).

Synthetic apps are not definitive, or hold any real value in the real world (unless you like to wave around your E-Peen) !!!!

 

firstep

Member
Jan 3, 2006
27
0
0
The thing is Dell doesn't sell AMD. :) I actually built a 3800X2 recently but felt tired about DIY. Too much hassel. So I prefer grabbing a Dell deal.

I agree super-pi is not everything. But I just want it as an indicator.

Originally posted by: stevty2889
Originally posted by: firstep
The reason is simple: super-pi reflects the cpu's computation power that in turns impact the performance/price ratio, a very reasonable factor for purchase decision. Let me know if you have any "smart" influential factor, if you consider performance/price ratio is silly. LOL.

There are always heavy-loaded and light-loaded jobs. So actually a wide range of selections are OK. For example, X300SE is OK for normal use. But if there's an X800 deal, why not jump?

Originally posted by: RichUK
You shouldn't let PI times influence you processor purchase. Lol that?s just silly ;)

What are you actually going to be using your machine for? This question should influence your processor purchase.

All super PI does it calculate digits of pi, it doesn't simulate real world performance. For video encoding, depending on the program you use, a dual core will be a good bit faster than a single core, but for non-smp aware aps, the dual core 2.8 is gonna be slower than the single core 3.2. Or you could spend ~$60 more, get a 3800+ X2, which will be faster than both of them.

 

RichUK

Lifer
Feb 14, 2005
10,341
678
126
Originally posted by: firstep
Interesting. As to the peformance competition, I think Intel and AMD teams are better candidates to argue which one is "superior", and some 3rd-party academia or industry instituition can be the referee. And I'm not an Intel advocate. But, I don't think you can just say AMD is far superior to Intel. I think it's safe to state that neither AMD nor Intel CPU can beat the other in every benchmark. Super-pi is just a popular benchmark. Assuming what you say is right, this benmark is more friendly to Intel. But can I say those benchmark with which AMD does better are more friendly to AMD? What's wrong with "more friendly"?

I'm not building a system, I'm not tuning up the performance. So super-pi just serves as a good enough estimation. I think for the first-order approximation, memory latency can be ignored here as long as the major factors such as DDR 400 or DDR2 533 are determined. Of course you can argue there are more parameters and brand does matter. To me, I just need a rough estimation. So I don't care.

I see you have some AMD system that you are proud of. However, does it really meet your real requirements, or it just get some high scores in 3DMark and you feel satisfied? To measure whether a system really meets personal use, probably you need some log software to record your personalized actions. Then you analyze it and get the conclusion whether a prospective system meets what you need. But I don't think PC users have intention to do that.

Ok first off the fact of the matter is, that in reality right now AMD offers a much better desktop solution and that is a given.

Better bang for buck or price performance ratio.
Produces less heat under load and idle operation
Doesn?t throttle
Literally holds the performance crown
HT well that is now moot, since both AMD and Intel are in the SMP era

?Just rolled off the top of my head?

And so on, I am not bias before you try and make this distinction. I run an AMD setup like most on this forum, not because I am a fanboy loyalist but because for my limited home use/needs, it suits my profile. Not to mention Intel doesn?t bring anything to the table with its current netburst processors which shouldn't interest anybody (except if your only upgrading the processor on your current setup). FYI I have just bought my second DELL laptop, which is a Yonah Core Duo laptop, and previous to that was a Dothan variant, and I am also awaiting on the new Conroe when it is released later this year, as a possible upgrade solution. I make educated decisions on what I purchase, so to make sure it fore fills my needs. My sig computer is just purely my enthusiast computer So don?t try and brand me. I use a range of computers that satisfy my every need (work and home related).

Also my reference to more friendly, wasn?t that the Super PI coders purpose built this to be bias for Intel, but infact it just ended up this way, as the program exploited an Intel processors more so than an AMD. And my main focus was that this has no relevance to real world performance, which is what each end user seeks right? For example if you analysed data from a range of processors tested by Super PI, could you then conclude a definitive winner from you results? I should hope not, and yes the argument is true for AMD also on a vice versa basis. But in reality AMD is in fact better, even in your scenario. Please do some research

To measure whether a system really meets personal use, probably you need some log software to record your personalized actions. Then you analyze it and get the conclusion whether a prospective system meets what you need. But I don't think PC users have intention to do that.

^^ Not really, for a single user, if you find out what their specific needs are (keeping in mind a PC is just a tool for a job), you can then ascertain there requirements quite quickly and work off a cost basis, and tailor a system so forth. Of course this is different in a corporate situation, in my job I have had the opportunity to work within an integration team, whom are responsible for designing complete system network infrastructures, and designing/working off custom ?life cycles? for given projects/rollouts (which have been designed around a business need), this can be tremendous work, but this is on the flip side from your situation which is quite simple.

Please be aware I am not stating AMD is superior as a company (which is evident if you read what I have previously said), in fact my opinion was aimed at the desktop market, in which AMD currently dominates. You must have been living under a rock if you haven?t realised this over the last year or so! Desktops aside, I would personally say that Intel have a firm grasp on the laptop (Mobile market), and as a result I have two laptops (well soon to be two) which has Intel hardware incorporated (processor, chipset, wireless), and at a good price.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
27,236
16,106
136
I have both, but I don;t run either at stock. The X2 3800 @ stock on every benchmark I have seen beats the 820D, virtually every bench of the 830D and many of the 840D. You want a space heater ? get a Dell PD. If you want good price/performance, and not DIY ? HP and many other places make them.

Edit: Just for kicks, I ran it on my slowest X2 @2370 (below the speed of a 4800+ and it was 1 minute 33 for the 2M test. Then I ran it on my 820D@3420, far above the 840D in speed, and it lost big time: 1 minute 50 sec for the same test.

As we are all saying, the X2 creams the 820D. And specifically in this test !
 

TSS

Senior member
Nov 14, 2005
227
0
0
http://www23.tomshardware.com/cpu.html?modelx=33&model1=237&model2=203&chart=63

you wanted to do video encoding right? there ya have it...

Originally posted by: firstep
The thing is Dell doesn't sell AMD. :) I actually built a 3800X2 recently but felt tired about DIY. Too much hassel. So I prefer grabbing a Dell deal.

I agree super-pi is not everything. But I just want it as an indicator.

oh, so your just lazy. coulda just sed so that you wanted to get a dell instead of making up some elaborate BS.