sempron on par with fx 53??? or faster than fx 51??

i ran 3d mark 03 05 pc mark 04 and since i'm not a registerd user i cannot tell you my cpu scores for them, and can't find any other programs to bench with that are also on toms hardware (http://www23.tomshardware.com/index....l2=58&chart=14) although i have a few sandra benchmarks, but keep in mind i'm on 754 and single channel ...

setup: @ tom's hardware page


Last Update: December 13, 2004
Hardware
Processor: AMD Socket 939
Motherboard: VIA K8T800 - ASUS A8V Rev. 1.02
Memory: 2 x 512 MB - DDR400 (200 MHZ)
Corsair CMX512-3500C2 XMS3502 V1.1
Graphics Card: Gigabyte GeForce 6800 GT (GVN68T256DH)
Hard Disk: 2 x SATA Maxtor 7B250S00 (Raid 0)
250 GB / 16 MB Cache / 7200 rpm
Promise FastTrak S150 TX2plus
Network: D-Link DFE530TX
Audio: Realtek ALC850
Software
OS: Windows XP, Build 2600 SP2
DirectX: Version: 9.0c (4.09.0000.0904)
Chipset Driver: VIA Hyperion 4 in 1 V4.53
Graphics Driver Detonator 61.77
Network Driver: Windows default
Audio Driver: disabled
Other Driver: none



my setup:

socket: 754
cpu: amd sempron 1.8 ghz @ 2.4 ghz with 1.525 volts and also @ 2.2 ghz & also 2.5 ghz
motherboard: dfi ut 250gb
ram: 512 mb pc 3200 twin mos utt @ 200 mhz & 208 via 150 devider
ram channel: single
vid card: 64 bit geforce 6200
hard drive: st320424a (20 gig compaq hdd)
network:nvidia nForce networking controller
audio: nvidia nForce
vid card core clock: 300
vid card mem clock: 400

operating system: win xp pro
direct x: 9.0
raphics driver:7.6.5.0
network driver:4.4.2.0
audio driver:6.14.442.30
other driver: none




sandra - cpu dhrystone
fx 51 10106
fx 53 11025
sempron @ 2.4 11023
sempron @ 2.2 10103

sandra - cpu whetstone
fx 51 4526
fx 53 4938
sempron @ 2.4 4888
sempron @ 2.2 4480

sandra - memory float
fx 51 5871
fx 53 6000
sempron @ 2.4 3047
sempron @ 2.2 3088


sandra - memory int
fx 51 5932
fx 53 6041
sempron @ 2.4 3048
sempron @ 2.2 3087


sandra - mm float point
fx 51 22570
fx 53 24624
sempron @ 2.4 24374
sempron @ 2.2 22337

sandra - mm integer
fx 51 20982
fx 53 22890
sempron @ 2.4 22728
sempron @ 2.2 20831

all are synthetic bench try to use something like super pi 1m or better a game or 3d mark 2001

i would but i don't see how 3d mark 01 is considered relevent for a cpu, and it's not listed on the toms hardware website click the link at the top, i can't get super pi stable only at 2.2 then it runs 42 secs

3dmark2001 is like quake 3 no gpu limited so if you have a faster cpu it affects the perfomance so it is a real world bench

ok, but it's useless to do, unless i have a cpu score from fx 53 & 51 to referance it to

Hey Nemo,

LOL i hope an xip doesnt see the thread title ;) Very big differences between a sempron and an fx. Most notably in real world apps, and super pi. The cache alone is a huge difference, not to mention the lack of 64-bit technology and what not.

Though you make a valid point about the changing nature of value processors. Ultimately they'll overcome the high ends of yesteryear, but the raw power of an FX chip, even at stock, is a force to be dealt with for most of the A64s out there, a sempron wouldn't stand a chance.

Case in Point: Many find 30k a milestone worth bragging about...an XT PE coupled FX55 both at stock do more than 30k in 01 :up:

Perkam

Quote:

---

Originally Posted by perkam

sempron on par with fx 53??? or faster than fx 51??

---

That made for a great laugh!! Thanks. :clap:

i know that 3d mark 01 is odd cuz i used to get 8.3 @ stock everything and now i get 7.3k with cpu @ 2400 i dunno though perk i need a program that can bench real world....links plz and they must be from that website (toms hardware) so i can compare to the fx 53 & 51

Just for comparison ;) This is the lowest cpu speed i can go, its still 11mhz more....
http://img176.echo.cx/img176/6016/screenshot0454ru.jpg

damn that is still pretty close except for the mem

Nemo try this...http://www.softpedia.com/get/System/.../CPUMark.shtml..This is my svcore with the same config...http://img68.echo.cx/img68/2933/screenshot0465zy.th.jpg

ok results are in

Thats a really nice score.. :toast:

thank you, infact the first score i beat you, i really din't expect that ...so any more links i wann c if i can hang with the big boys well last years big boys anyways is ur htt 200.9 cuz that times 12 which i think ur multi is..that would explian 2411

K, this another one and just try the rendering....http://www.cinebench.com/

okk buddy i really hope i'm at par with you, you got a fx 53 right???




results are in

I run fx55 underclocked to 2411Mhz and this mine...
http://img108.echo.cx/img108/623/scr...t0484qv.th.jpg

omg ur only 1.7 seconds faster with a lot more chache lol if you want my speed go 267*9

Your sempron definately outstanding...I still have my XP2800 t'bird B that oc like crazy, next time we will comparing notes again.. :toast:
Now, I'm back to my normal oc...
http://img165.echo.cx/img165/4585/sc...t0497ud.th.jpg

no more tests ???? lol

Try run Aquamark3 with your highest oc and see your cpu score...

umm ok will do @ 2.4 as i have been

interesting but day to today apps a sempron cant compare to fx.

run cpu mark 99 @ 2.4 ghz here's my score

kamaleon i wish i knew for sure instead of guessing, tha's what i'm trying to see here so far we're pretty close neck and neck imo. other than the cache difference they seem to be doin' the same work clock per clock remember sempron is basically a single channel 32 bit only winchester or clawhammer i 4-get which one

Quote:

---

Originally Posted by i found nemo

kamaleon i wish i knew for sure instead of guessing, tha's what i'm trying to see here so far we're pretty close neck and neck imo. other than the cache difference they seem to be doin' the same work clock per clock remember sempron is basically a single channel 32 bit only winchester or clawhammer i 4-get which one

---

Its winchester 9.0nm..

you mean 90nm lol anyways run those tests see where we end up and remember 2.4ghz ;) thanx for doin' this for me

Quote:

---

Originally Posted by i found nemo

run cpu mark 99 @ 2.4 ghz here's my score

---

For some reason, I cant run AM3 with cpu underclock.....Heres cpumark...

woooo :banana::banana::banana::banana: you really got it there must be the l2 cache taking effect

any other bench marks we can think up???

Run 3DMARK05 cpu test and compare to other cpus on the orb.

run an unreal engine game with everything in low like ut2003 or ut with everything ultra low and low res and you will see the difference

Quote:

---

Originally Posted by leviathan18

3dmark2001 is like quake 3 no gpu limited so if you have a faster cpu it affects the perfomance so it is a real world bench

---

I saw some crap like this over at overclockers forums the other day, started by the one and only idiot OC550.

Sure, what you say is true, to a certain extent, it's very dependent on HTT/FSB timmings, etc, but at the end of the day, it's still a Video card bench.

A Sempron is not in the same league as a FX. There's a reason why one costs 80quid and the other costs over 500. :)

i need links can't do nothin' without them i already posted this, i don't care if it's not in the same leage i want to see where it ends up when the dust settles quit telling me stuff i know and assume apperently ppl don't read to well cuz i've already posted this once

belive me quake 3 unreal tournamet are not longer gpu limited are cpu limited you can try this put a 6800gt with a athlon64 3000 and then run the same bench i.e quake 3 with a 3500 you will see a big difference in FPS.

3dmark2001 is not longer gpu limited now is cpu limited and that is why many ppl here uses it like a cpu bench more than a video bench

wow that toms cpu guide is really helpful.

nicke i can't don't have access to it don't have full version i already said this, any body with a unreal tounament download link i'll do the test if dumo will

screenies @ 2502 mhz memory @ 208 via 150 devider

Here are some close comparisons.
http://www.xbitlabs.com/articles/cpu...ron-3100e.html
http://www.xbitlabs.com/articles/cpu...n-3100-oc.html
http://www.xbitlabs.com/articles/cpu...pron-2600.html

Quote:

---

Originally Posted by Major_A

Here are some close comparisons.
http://www.xbitlabs.com/articles/cpu...ron-3100e.html
http://www.xbitlabs.com/articles/cpu...n-3100-oc.html
http://www.xbitlabs.com/articles/cpu...pron-2600.html

---

Nemo, you can beat those sempron rev.E :toast:

Nemo:

Great way to test this Sempron. I think your little thread basing it on facts and not "oh man.... any FX blows this away" is just the thing that ppl need to see. Makes the OC'er in us stand up and take notice. If I were to build a new system and price was of utmost importance, Sempron is it!

Thanks!
Magik

You happy we happy.:) But Toms is the LAST place I'd be looking for AMD benchmarks. He should change his domain to intelzone.com;)

Uh oh... The FSB 267 is the only thing that sempron has on a FX. Seems everyone missed that? Get Hardcoreclocker here and have him run your exact multi and FSB and you will see the diffrence is very big.

i think is funny sempron fx killer ok if you overclock one but lets overclock a little the fx to see maybe just get a higher htt and a lower multi in the fx to see if the sempron keeps winning.... i dont think so....

hey hold on you all make it seem like i'm saying my cpu is better than fx, point out where i said that but it would be nice to see how they hold up with the same multi's and htt :) don't get me wrong and thanx dumo for the help. where my cpu is @ 2.5ghz the fsb is 278 with memory @ 208 via 150 devider if any one feels the need to compare @ exact fsb and multi plz do thank you very much for your support - thanx majik it's about time someone ecouraged me :)


edit- also i'd love to see fx overclocked versus mine but i can't get past 2610 @ 1.8 volts so it's not happening at least not for FAIR comparison and don't tell me my htt is unfair, cuz i lack dual channel :) lol j/p

edit- also note that originally this was comparing to fx 51 &53 and you throw a fx 55 that has higher work per clock cycle than my sempy could ever IMAGINE (@ stock anyways i think) to have and i still hold the candle if you ask me note once agian i'm short 896kb of l2 cache i'm sorry some of you are offended and want "another test with same settings" i din't think it'd come to that but agian it would be appreciated if someone did do that

edit- just pm'd opb hopefully he will help us decide if my sempy deserves it's credibility :)

edit- just pm'd hard core clocker :)

i don't care which fx pl refer to my "experement" @ http://www.xtremesystems.org/forums/...ad.php?t=61161
plz keep cpu @ same fsb/htt and multi for "precise" results plz :)

Quote:

---

Originally Posted by Grov

I saw some crap like this over at overclockers forums the other day, started by the one and only idiot OC550.

Sure, what you say is true, to a certain extent, it's very dependent on HTT/FSB timmings, etc, but at the end of the day, it's still a Video card bench.

---

Not really. With the same CPU, a 9800 Pro will probably get a better score than a standard 6800 in it, despite obviously being slower. It's mostly CPU dependent now.

mostly but not 100%

Lol, really nice.

But i dont understand that a underclocked fx-55 @ 2.4GHz get's 1:16 @ rendering.
When it's at stock (2.6) it does 52 seconds, that quite a big difference for 200Mhz more...

i have a FX-51 if that helps? :stick:

but its not in my pc right now.

oh hell yes that will thank you so much for your contribution to my experement :)

maybe it's a very "mhz" sensitive program???

this is an old screenie 2610mhz @ ~1.8 volts so can someone show me what an o/c'd fx 51 fx 53 and stock fx 55 do on sandra @ 2.6 on these bench marks?

http://www.xtremesystems.org/forums/...chmentid=27996

but its not in my pc right now. its in its retail box.

it's okay you do know how to put the cpu in the socket right??? c'mon now don't tease me with the fact that you have but won't/can't help...plllllz take it outta box and run it 4 meeeee?

well i cant, 1st of all i dont have registerd ram. (sold it)

2nd of all, if i did have ecc ram, i would have to build a whole new pc again.

you wanna buy it from me?

i'm skt 754 and no, cuz i don't have that kinda money laying around..got any old sandra benches or any thing saved to contribute sir?

Why dont u just use the countless reviews of the FX51/53/55 and instead of asking ppl to do tests at different fsb, just put your sempy at that speed. For one thing, the FX-53 runs at 2.4ghz stock...so just find an fx-53 run, clock your sempy to 2400 and fire away :)

Most with FX-55s here ARENT running it on air ;) and those with fx 53/51s just dont put it in their sigs or have moved on from them...so unless you're willing to go back in time when the AMD forums were literally flooded with FX-53s on the orb and everywhere, I'd suggest you use FX runs on air in reviews (anandtech, etc) and on xs to compare with your sempy.

..Yes, 2....and counting ;) .

Perkam

i dunno you must have this all wrong perk first off the first one was comparing the two then controversy over my htt being higher than 200 was the reason for my cpu's success and this is not 4 threads at all indeed but only 2

Quote:

---

Originally Posted by i found nemo

i dunno you must have this all wrong perk first off the first one was comparing the two then controversy over my htt being higher than 200 was the reason for my cpu's success and this is not 4 threads at all indeed but only 2

---

My point isnt the no. of threads, its that couldn't this have been continued on in the previous thread ???

Anyway, you can ask Dumo to help you again or any other good willing member...but it really is a waste of time considering you're not even on the same platform. There are too many factors not accounted for: benefits from dual channel, benefits from the cache, benefits from bank interleaving, benefits from tighter timings, etc.

IMO, a more worthy comparison would've been to get someone with a Turion and then compare clock for clock, or even a 3700+ Claw, considering you're aiming quite high.

Perkam

it don't matter perk i wanna do comparisons of cpu not memory so dual channel won't be benificiary, chache yes the fx will stomp me but w/e apperently i already eat up the new sempron rev e's and if i go for a turion barly any one owns one so results will take for ever, even if the other person has one does not mean they are willing and i think but am not sure don't think hcc or opb will be helping with results and dumo just quit posting results i thought he sold i to charlie i'm not sure

Calm down, guys. And one thread on the same topic will suffice. New thread has been merged with the original thread. ;)

noticed that :) any opinoins on this conrad?

My opinions on this topic are as follows:

- FX = Flagship AMD Processor, catering to the high-end enthusiast and PC-user
- Sempron = Low-end AMD Processor, catering to the enthusiast and PC-user who craves maximum performance on a budget

The two processors can not be compared. The obvious is the fact that you are dealing with two different sockets. Other differences include memory controllers (single channel vs. dual channel), instruction sets (MMX, 3DNow, SSE, SSE2, SSE3, etc), OS compatibility (32-Bit vs. 64-Bit), and so on.

If you are able to obtain XYZ clocks on a Sempron, which happens to be the exact frequency someone else's FX is running at, you will naturally notice null differences in certain applications. However, the moment you start comparing the two processors on an application that makes use of the extra memory bandwidth, or instruction set, or OS-specific architecture, the differences between the two should be quite apparent.

Sempron vs. FX will always be an apples vs. oranges comparison. ;)

ask to some one with a 3800 or 3500 i think they are more common or somebody with a 3700 sandy

but my bet are on athlon in real world bench if both cpu have the same htt

Quote:

---

Originally Posted by perkam

or even a 3700+ Claw, considering you're aiming quite high.

---

will a 3700+ newark work?


Nemo: what's your cpu score in 3dmark05?
this is at 240x10 with my ballistix at 2.5-2-2-10, 240mhz

Quote:

---

Originally Posted by Zebo

You happy we happy.:) But Toms is the LAST place I'd be looking for AMD benchmarks. He should change his domain to intelzone.com;)

---

LOL! Do you go to extremetech.com ? It should be ExtremeIntelfanboys.com!

Tom's Hardware quide is actually quite reasonable, whilst Anandtech does seem to have an AMD Bias (or is it that they are objective hence apparent AMD Bias is actually superiority of AMD product?)

ozzi c'mon now tha's a vid card bench not to mention i've got a 64 bit pos 6200

Quote:

---

Originally Posted by i found nemo

ozzi c'mon now tha's a vid card bench not to mention i've got a 64 bit pos 6200

---

Look closely at his SS, he's asking to compare CPU Marks ;)

Perkam

wish i had more cache show some of the other benchmarks that were posted up here


"Sempron vs. FX will always be an apples vs. oranges comparison. " <---that means "leave me outta this" lmao

try running the bench twice. i always always always get higher scores the second time through. i should also note that that is from the rig in my sig, 3200+ clawhammer

edit: though, it looks like 3dmark loves cache.

uuuuh huh ok let me know when you get the rest of the results in

ok...quick bump

c'mon man lol hurry up ozzmark get ur cpu in here

semprons are deff alot of bang for the buck. IMHO I would say in 3dmark01 tho and alot of games a fx has 300MHz over you, so what you do at 2500 an fx can do at 2200.

Quote:

---

Originally Posted by i found nemo

wish i had more cache show some of the other benchmarks that were posted up here


"Sempron vs. FX will always be an apples vs. oranges comparison. " <---that means "leave me outta this" lmao

---

lol same here.

so you gunna get us those benchies on ur winnie and ur spankin' new newark??? :drool:

oh c'mon ozzi

sorry! i've been busy!! :( :p:

i'll try to get them up soon :toast:

i think i'll do 2.4ghz, 2.6ghz and 2.8ghz.. if benches run at 3ghz, i'll do that too

2.4ghz :toast:

aquamark is probably pretty gpu dependant from the looks of things.. and it's the first time i've ever tun that bench, so i have no idea how good that score is.. time to do some research and then do 2.6ghz :p:

2.6ghz, ram was dropped to 2.5-2-2-10, as opposed to 2-2-2 in the 2.4ghz run

2.8ghz, same settings as 2.6, but a higher htt speed. probably more than is needed for comparison purposes with your sempron, but i think you might just be able to hit it if you want (very cold air should get you there) :D
i guess if dumo wants to go through the 4 benches with his fx-55, we can compare the true meaning of dual channel :toast:


and i just realized. i forgot to run 3dmark. :slapass:



edit: uploaded the 2.6ghz benches by mistake :eek:
edit2: not cold enough to do 3ghz with 1.65v.. maybe later tonight ;)

on ur 2.4ghz, i stomped you on arithmatic and cinebench @ 2.4 i was only .1 second off....but the rest was kinda up there....surpized it was even that close, and in aquamark i had 9725ish and you had 10283 so i don't think mine is too bad, plus urs has sse3 and a lot larger cache, good job i wish i could get mine upto 2.6 or higher well any more reviews or what??? lol

here's more, just so ppl don't gotta brwoze around too much btw you forgot to do cpumark 2 which is posted on here earlier

56k warning lol

http://www.xtremesystems.org/forums/...chmentid=29618
http://www.xtremesystems.org/forums/...chmentid=29619
http://www.xtremesystems.org/forums/...chmentid=29626
http://www.xtremesystems.org/forums/...chmentid=29630
http://www.xtremesystems.org/forums/...chmentid=29666
http://www.xtremesystems.org/forums/...chmentid=29665
http://www.xtremesystems.org/forums/...chmentid=29667
[img]http://www.xtremesystems.org/forums/attachment.php?attachmentid=29673
[/img]
http://www.xtremesystems.org/forums/...chmentid=29674
http://www.xtremesystems.org/forums/...chmentid=29715
http://www.xtremesystems.org/forums/...chmentid=29721
http://www.xtremesystems.org/forums/...chmentid=30002

damn i just noticed ur cpu really takes advantage @ 2.6 my cpu only got 11861mips where as you got 12+

fx55 doesnt have sse3 unless is the new revision 90nm from sandy core

and why you dont run 3dmark01 is the only 3dmark not gpu dependant anymore (unless you are using a very old gfx)

look @ my aqua mark score, my gpu is my bottleneck

i don't think sse3 makes any difference for any of the benches we're running.
and sandra is extremely clock speed dependant :stick:

i have a bunch of stuff to do today, so no more benches from me for a while. sorry :(

yea..tha's true...i dunno ppl must not like the title of my thread, they feel threatened lol

hahaha, probably.. people trying to avoid the fact that a cheaper cpu is almost always a better deal ;)

or else they're too lazy/busy to do benches :slapass: :D

i have a 3500@ 2.2 if you want to run bench with me but you have to adjust to my settings to have a "equal" machine

well push 'er up to 2.4 btw broke my own personal record of 2610 2672 :) @ 20 degrees celcius i put my computer over the airconditioning vent :D in the pic it's 25 but cools down real quick other temps are in the teens

edit looking at my and yours arithmatic score, mine wins, but when it comes to graphix urs wins....wtf lol

ok hcc should be helping me within a few hours with his awesome fx 55, thanx hcc ;)

okay he told me he was....presuming this a no-show

Quote:

---

Originally Posted by i found nemo

well push 'er up to 2.4 btw broke my own personal record of 2610 2672 :) @ 20 degrees celcius i put my computer over the airconditioning vent :D in the pic it's 25 but cools down real quick other temps are in the teens

edit looking at my and yours arithmatic score, mine wins, but when it comes to graphix urs wins....wtf lol

---

he has more cache... that is probably why

prolly