Celeron D vs Sempron Socket A's

Status
Not open for further replies.
D

dabster

In the zone
Hey,
I am on bugdet, and have to decide over Celeron D 2.4 Ghz and the AMD Sempron Socket A 2400+ ....I can't go for the Newer socket 754 for the newer AMD's As mobo's are much costlier.
Also my work includes a little of games,though mostly for designing Websites, developing new components...in .NET...
I have heard tht Amd is far better , But What about the New Celeron D' which are having FSB's 533 MHz, As compared to AMD' s 333MHz, Any computer geek out here who can tell me how this FSB actually matters during work.
Also many AMD mobo use shared(with RAM) memory of the order of 32 MB, Is it really Better than what we have in P4 (8 MB only...)
Anybody out here who has recently buyied a celeron, Do post how it fares...
 
K

Keith Sebastian

Padawan
That's a hard question to answer. Both processors are quite similar in performance AND price vs performance ratios. AMD seems to have a slight edge performance wise while Intel has a slight price advantage. Have a read through the following articles. The second article should be of more use to you as it compares full systems (with onboard graphics) as opposed to processors only.

*www.xbitlabs.com/articles/cpu/display/sempron.html

*www.pcper.com/article.php?aid=61&type=expert

And BTW, welcome to the hardware forums. :D

Cheers,
Keith
 
OP
D

dabster

In the zone
Superb links, I suppose tht will be going for the Celeron D with some newer mobo..... Since i was to decide b/w sempron 2400+ or Celeron ,I suppose tht it is right choice for me
thanks.. Keith
 
I

indro

In the zone
Celeron is never better than sempron .Sempron has the FPU performance benifit .A lot of other stuff as well . But , both the platforms are going to be extinct real soon .
 
OP
D

dabster

In the zone
indro said:
Celeron is never better than sempron .Sempron has the FPU performance benifit .A lot of other stuff as well . But , both the platforms are going to be extinct real soon .
Click to expand...

The Newer Celeron D is having 533Mhz FSB and the Cache Memory is also comparable....I tried the UT 2004 game on AMD sempron 2400+ even when the shared graphics memory is 32MB, the Graphics was not that good as compared to 8MB on pentium 4 2.4GHz, Though for Celeron I don't Know....Hopefully that's also Gonna perform Better....
 
M

Mahesh Babu

Journeyman
Yes go for Celeron D 2.8 Ghz processors over AMD sempron 2600 and the performnce will be the same in both.But there is a 1000 rupees gap between the two and the celeron runs much smoother.So celeron would be the much better choice and yes infact the higher FSB does make a lot of difference and performance will be lot smoother.
 
I

indro

In the zone
dabster said:
indro said:
Celeron is never better than sempron .Sempron has the FPU performance benifit .A lot of other stuff as well . But , both the platforms are going to be extinct real soon .
Click to expand...

The Newer Celeron D is having 533Mhz FSB and the Cache Memory is also comparable....I tried the UT 2004 game on AMD sempron 2400+ even when the shared graphics memory is 32MB, the Graphics was not that good as compared to 8MB on pentium 4 2.4GHz, Though for Celeron I don't Know....Hopefully that's also Gonna perform Better....
Click to expand...



See , 2.4 ghz Pentium 4 was not supposed to be compared with a sempron in the first place , anyways the intel chipset motherboards with the extreme graphics feature has the "Dynamic Memory Allocation" option which dynamically assigns video memory upto 64MB ,which of-course you cant see within the game .So that point is not justified.
 
kumarmohit

kumarmohit

Technomancer
Celeron D is 64 bit and since both are 64 bit the no. crunching ability advantage is gone for Sempron and Cel with its high FSB gets an edge

I wud first of all tellu what is FSB (tho I am not a geek) FSB is the speed at which ur processor interacts wih ur RAM and/or Northbridge(Northbridge in turn ctrls all other devs like ur cards) so IF u pump in a lot of RAM (min 512 MB) and keep virtual memory low U get really good performance.
since u will be goin for Intel 915 MB u get Intel extreme gfx 2 aided with the amt of VRAM(Video Ram that gets shared and 32 MB shared VRAM + the FSB env will make it an acceptable gaming rig which along with established trackrecord of Celeron is where Sempron looses out .

So all in all go for Celeron preferably and since both games and gfx demand RAM so try to to go for atleast 1GB of RAM (preferably DDR2)
 
infra_red_dude

infra_red_dude

Wire muncher!
i think the thread starter wanted a comparison between a celeron d (wid 64bit instructions) and a "socket A" sempron. afaik, no socket A sempron has 64bit extensions! correct me if i am wrong.

there can be no comparisons b/w a 32bit proc and a 64bit one! a socket A sempron can be compared wid a normal celeron. and sempron really wins the hands down.

but wid the cache improvements,fsb, 64bit thingy etc the celeron d edges out the sempron 32bit in most general purpose areas.

also hafing a higher fsb over a higher clocked proc is advantageous as the higher fsb reflects the overall system performance as most of the components are benifitted from it. however in case of a lower fsb but higher clocked proc, its only tht proc thats provides the edge. so the overall effect may not be the same.
 
darklord

darklord

Cyborg Agent
krishnathelord said:
CELERON CAN ANY DAY BEAT SEPRON

GO for celeron
Click to expand...

Make it clear that its CELERON D you are talking about cos the older Celrons would be owned by Semprons. :wink:

Celeron D is 64 bit and since both are 64 bit the no. crunching ability advantage is gone for Sempron and Cel with its high FSB gets an edge
Click to expand...

Celeron D is NOT 64 Bit :shock:
Nor are Socket A semprons.

since u will be goin for Intel 915 MB u get Intel extreme gfx 2 aided with the amt of VRAM(Video Ram that gets shared and 32 MB shared VRAM + the FSB env will make it an acceptable gaming rig which along with established trackrecord of Celeron is where Sempron looses out .
Click to expand...
How exactly do you suggest to make a Socket 478 CPU work on a Socket T board ? :roll:

32 MB shared VRAM + the FSB env will make it an acceptable gaming rig---------> What exactly does this mean ? cos it literally makes no sense. :roll:

So all in all go for Celeron preferably and since both games and gfx demand RAM so try to to go for atleast 1GB of RAM (preferably DDR2)
Click to expand...
Wow DDR2 for Celeron D....pretty ambitious
AFAIK,there are no Socket T Celeron in Indian market yet. :wink:

@ Thread starter,
Think about Socket 754 based Semprons too,they thrash out even the mighty P4s...lol
Also they are just round the corner in India.Whats more, if you are lucky, you can even get the 64 Bit one.

Also Socket 754 boards are pretty cheap.
So 754 based Sempron will -
1] Run extremely Cool cos of 90nm fab tech.
2] Literallly Humiliate a P4 [2.4-3.0Ghz] leave aside Celeron D
3] Extremely Cheap.Price should be equal or less than Celeron 2.4D :wink:
4] They are wicked Overclockers....
 
infra_red_dude

infra_red_dude

Wire muncher!
lets us try to understand wot this 64-bit is all about. consider the example below:

suppose u have a reservoir wid a hole of diameter 1 meter and a pipe of the same diameter is connected to it, the other end of which is connected to a tank wid a 1m diameter hole in it. the task to be accomplished is to transfer a specific volume of water from the reservoir to the tank.

now consider another reservoir wid a 2m diameter hole connecting a tank wid a 2m dia hole wid a suitable pipe.

consider a third situation where the hole in the reservoir is 1m in dia but the hole in the tank is 2m.

from the three scenarios you can see that the transfer of specific volume of water in the first case takes the longest time and a bit less time in the third case. but it takes the shortest time in the second case.

the analogous terms in the computer world would be :
reservoir - software
pipe - bus
tank - processor registers

so wot basically is 64bit processing? procs contain registers which are analogous to memory cells. if some data has to be processed it should be brought to a register for processing. if the data is a 128 bit and the processor is 32 bit then the data is broken into 4 parts. the cpu is capable of executing an instruction (like adding) on the data that is equivalent to the register word size (32bit). lets say that two 128bit words are to be added. so the 1 first part of the first data (first 32bits) is brought to the cpu register (called general purpose register) and the first part of the second data (first 32 bits) is brought to another register (usually called the accumalator) and added. lets say it takes 1 second to do it. the processed first part is saved at a location and when all the parts are added they are joined together to obtain the 128bit (or more) result. it takes 1x4=4 secs to do it.

now consider a 64bit processor. the 128bit is broken into 2 parts. so the effective adding time should be 1x2=2secs (Which is not true we'll see it later why)

thus a 64bit cpu helps in reducing the time factor.

now lets see why it is not exactly half the time taken to process in a 64bit cpu. first thing the adding time may not be 1secs. it may take 1.5secs to add a word. so effectively it takes 3secs instead of 4 secs that a 32bit takes. secondly the software maybe a 32bit version. so the software can supply wid only 32bit words. and two 32bits are joined to make a 64bit word in the cpu register. all these and many more factors make the 64bit cpu not exactly doubly fast.

the two examples given above were very crude examples. but i've tried to make it as simple as possible for understanding puposes. hope it helps :D
 
K

Keith Sebastian

Padawan
infra_red_dude said:
lets us try to understand wot this 64-bit is all about. consider the example below:

suppose u have a reservoir wid a hole of diameter 1 meter and a pipe of the same diameter is connected to it, the other end of which is connected to a tank wid a 1m diameter hole in it. the task to be accomplished is to transfer a specific volume of water from the reservoir to the tank.

now consider another reservoir wid a 2m diameter hole connecting a tank wid a 2m dia hole wid a suitable pipe.

consider a third situation where the hole in the reservoir is 1m in dia but the hole in the tank is 2m.

from the three scenarios you can see that the transfer of specific volume of water in the first case takes the longest time and a bit less time in the third case. but it takes the shortest time in the second case.

the analogous terms in the computer world would be :
reservoir - software
pipe - bus
tank - processor registers

so wot basically is 64bit processing? procs contain registers which are analogous to memory cells. if some data has to be processed it should be brought to a register for processing. if the data is a 128 bit and the processor is 32 bit then the data is broken into 4 parts. the cpu is capable of executing an instruction (like adding) on the data that is equivalent to the register word size (32bit). lets say that two 128bit words are to be added. so the 1 first part of the first data (first 32bits) is brought to the cpu register (called general purpose register) and the first part of the second data (first 32 bits) is brought to another register (usually called the accumalator) and added. lets say it takes 1 second to do it. the processed first part is saved at a location and when all the parts are added they are joined together to obtain the 128bit (or more) result. it takes 1x4=4 secs to do it.

now consider a 64bit processor. the 128bit is broken into 2 parts. so the effective adding time should be 1x2=2secs (Which is not true we'll see it later why)

thus a 64bit cpu helps in reducing the time factor.

now lets see why it is not exactly half the time taken to process in a 64bit cpu. first thing the adding time may not be 1secs. it may take 1.5secs to add a word. so effectively it takes 3secs instead of 4 secs that a 32bit takes. secondly the software maybe a 32bit version. so the software can supply wid only 32bit words. and two 32bits are joined to make a 64bit word in the cpu register. all these and many more factors make the 64bit cpu not exactly doubly fast.

the two examples given above were very crude examples. but i've tried to make it as simple as possible for understanding puposes. hope it helps :D
Click to expand...

WTF?? You get first prize for totally irrelevant reply of the month. Also known as "Foot in the mouth award"

-Keith
 
infra_red_dude

infra_red_dude

Wire muncher!
no palz just trying to throw some light on the 64-bit hype, thats it. well if you feel its soooooo irrelevant then plz delete my post.

lotsa ppl were trying to compare a 32bit proc wid a 64bit one. so i thot that some info on 64bit thingy would clarify most of their doubts.

mods: plz delete my post if its totally off topic.
 
I

indro

In the zone
Relax Dude ,

Nothing Like that , its a forum which is public , everyone has the right to speak .Myself did find it to be interesting , involved a lot of Physics etc .I apologise if you felt bad :)
 
Status
Not open for further replies.
Top Bottom