• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Celeron vs P4: Test Results

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
nodsetse

nodsetse

Member
​
Joined
Oct 22, 2002
Location
San Diego, CA, USA
A few weeks ago I picked up a Celeron 2.0, having heard they were good for at least 3GHz. Since I'd had a P4-1.6a for some time, I couldn't help but compare the performance. So how does a 3GHz chip with 128k of L2 do against one at just 2.4GHz with 512k of cache? Let's find out!

System Configuration
1) B0 Pentium 4-1.6a @ 2.40GHz (BIOS @ 1.65vcore, 1.59-1.64 actual)
2) C1 Celeron 2.0GHz @ 3.0GHz (BIOS @ 1.6vcore, 1.54-1.59 actual)
*Both using stock HSF
Abit BD7 (original) i845 150MHz FSB (600MHz QDR) BIOS 7_7C (MrN)
256MB Samsung (DTL) PC2700 @ 1:1 = 300MHz DDR CAS2-2-5-2-2
Gainward GF4 ti4200 64MB 3.6ns EtronTech @ core/memory (as specified)
40GB WD 7200 RPM UDMA/ATA-100
Win 98se, DX 8.1, Nvidia 40.72 Drivers

Note the completely objective comparison, in that EVERYTHING is the same, except the CPUs were changed.

Test Results

Default 3DMark 2001se ti4200 @ 300/560
P4-2.4 = 12,157 3D marks
Cel-3.0 = 10,726 3D marks

Default 3DMark 99 ti4200 @ 275/550 - VSync ON
P4-2.4 = 8,487 3D Marks, 32,570 CPU 3D Marks, 14,085 3DRasterMarks
Cel-3.0 = 8,379 3D Marks, 33,952 CPU 3D Marks, 12,525 3DRasterMarks

Final Reality, ti4200 @ 295/555
P4-2.4 = 22.33(2D), 8.04(3D), 12.08 Reality marks
Cel-3.0 = 24.30(2D), 7.55(3D), 12.32 Reality marks

Vulpine GLmark 1024x768, 32bit color, 32bit textures, ti4200 @ 295/555
P4-2.4 = 65min, 218max, 112.1avg
Cel-3.0 = 61min, 211max, 100.8avg

N-Bench Ver.1, ti4200 @ 295/555
P4-2.4 = 7847 overall
Cel-3.0 = 7216 overall

N-Bench Ver.2, ti4200 @ 295/555
P4-2.4 = 1827 overall
Cel-3.0 = 1825 overall

UT2003 Benchmark 1024x768 ti4200 @ 275/550
P4-2.4 = 157.8 flyby, 61.3 botmatch
Cel-3.0 = 138.0 flyby, 43.6 botmatch

Comanche 4 Benchmark 1024x768 ti4200 @ 275/550
P4-2.4 = 42.78 / 8,521.688
Cel-3.0 = 23.32 / 4,644.883

SPECviewPerf 6.1.2 - Average of 6 tests - ti4200 @ 295/555
P4-2.4 = 44.20167
Cel-3.0 = 45.48833

Default PCMark 2002 ti4200 @ 275/550
P4-2.4 = CPU 5805, Memory 5398, HDD 1146
Cel-3.0 = CPU 6578, Memory 4874, HDD 1096

Super PI Calculate to 256K
P4-2.4 = 18sec
Cel-3.0 = 15sec

Norton SI System Benchmark
P4-2.4 = 856.4
Cel-3.0 = 592.3

SiSoft Sandra 2002 CPU Arithmetic
P4-2.4 = ALU 4505 MIPS / FPU 2950 MFLOPS
Cel-3.0 = ALU 5410 MIPS / FPU 3730 MFLOPS

SiSoft Sandra 2002 CPU Multi-Media
P4-2.4 = Interger 9446 / Floating-Point 11549
Cel-3.0 = Interger 11927 / Floating-Point 14793

SiSoft Sandra 2002 Memory Bandwidth
P4-2.4 = Int 2279 / Float 2180
Cel-3.0 = Int 2204 / Float 2097

Well, that's it...make up your own mind on whether or not some celery might be good for you. As for me, I'm going to look at the P4-2.0a once they are available with the C1 stepping and priced under $150 USD, which could be by the end of this month! Anyway, I just thought this might be some interesting info to share, the results are presented "for what it's worth", nothing more. ;)
 
the celeron's suck because they have way to little L2 :/ sorry to say :/ but i guess thats why the OC so high
 
Hey jd, I was hoping you'd see this thread! ;)
Of course I agree completely with what you said about the Celeron. As the results indicate, sometimes they're beat by a CPU running 600MHz slower, but in other apps, 3GHz is 3GHz. Given the price for the performance, I've found this Celeron to be a great chip for everyday use, with all the applications I'm running. :)
 
nodsetse said:
Hey jd, I was hoping you'd see this thread! ;)
Of course I agree completely with what you said about the Celeron. As the results indicate, sometimes they're beat by a CPU running 600MHz slower, but in other apps, 3GHz is 3GHz. Given the price for the performance, I've found this Celeron to be a great chip for everyday use, with all the applications I'm running. :)
Click to expand...
i like the chip too, but needed faster seti times, than the celly could deliver. nice chip though.
 
jdmcnudgent said:
i like the chip too, but needed faster seti times, than the celly could deliver. nice chip though.
Click to expand...
How does your Cele time compare with the rig in your sig, for an average WU? I haven't tested SETI with the Cele at 3GHz yet, but my P4 setup completes a unit in 3hrs 45min.
 
nodsetse said:
A few weeks ago I picked up a Celeron 2.0, having heard they were good for at least 3GHz. Since I'd had a P4-1.6a for some time, I couldn't help but compare the performance. So how does a 3GHz chip with 128k of L2 do against one at just 2.4GHz with 512k of cache? Let's find out!

System Configuration
1) B0 Pentium 4-1.6a @ 2.40GHz (BIOS @ 1.65vcore, 1.59-1.64 actual)
2) C1 Celeron 2.0GHz @ 3.0GHz (BIOS @ 1.6vcore, 1.54-1.59 actual)
*Both using stock HSF
Abit BD7 (original) i845 150MHz FSB (600MHz QDR) BIOS 7_7C (MrN)
256MB Samsung (DTL) PC2700 @ 1:1 = 300MHz DDR CAS2-2-5-2-2
Gainward GF4 ti4200 64MB 3.6ns EtronTech @ core/memory (as specified)
40GB WD 7200 RPM UDMA/ATA-100
Win 98se, DX 8.1, Nvidia 40.72 Drivers

Note the completely objective comparison, in that EVERYTHING is the same, except the CPUs were changed.

Test Results

Default 3DMark 2001se ti4200 @ 300/560
P4-2.4 = 12,157 3D marks
Cel-3.0 = 10,726 3D marks

Default 3DMark 99 ti4200 @ 275/550 - VSync ON
P4-2.4 = 8,487 3D Marks, 32,570 CPU 3D Marks, 14,085 3DRasterMarks
Cel-3.0 = 8,379 3D Marks, 33,952 CPU 3D Marks, 12,525 3DRasterMarks

Final Reality, ti4200 @ 295/555
P4-2.4 = 22.33(2D), 8.04(3D), 12.08 Reality marks
Cel-3.0 = 24.30(2D), 7.55(3D), 12.32 Reality marks

Vulpine GLmark 1024x768, 32bit color, 32bit textures, ti4200 @ 295/555
P4-2.4 = 65min, 218max, 112.1avg
Cel-3.0 = 61min, 211max, 100.8avg

N-Bench Ver.1, ti4200 @ 295/555
P4-2.4 = 7847 overall
Cel-3.0 = 7216 overall

N-Bench Ver.2, ti4200 @ 295/555
P4-2.4 = 1827 overall
Cel-3.0 = 1825 overall

UT2003 Benchmark 1024x768 ti4200 @ 275/550
P4-2.4 = 157.8 flyby, 61.3 botmatch
Cel-3.0 = 138.0 flyby, 43.6 botmatch

Comanche 4 Benchmark 1024x768 ti4200 @ 275/550
P4-2.4 = 42.78 / 8,521.688
Cel-3.0 = 23.32 / 4,644.883

SPECviewPerf 6.1.2 - Average of 6 tests - ti4200 @ 295/555
P4-2.4 = 44.20167
Cel-3.0 = 45.48833

Default PCMark 2002 ti4200 @ 275/550
P4-2.4 = CPU 5805, Memory 5398, HDD 1146
Cel-3.0 = CPU 6578, Memory 4874, HDD 1096

Super PI Calculate to 256K
P4-2.4 = 18sec
Cel-3.0 = 15sec

Norton SI System Benchmark
P4-2.4 = 856.4
Cel-3.0 = 592.3

SiSoft Sandra 2002 CPU Arithmetic
P4-2.4 = ALU 4505 MIPS / FPU 2950 MFLOPS
Cel-3.0 = ALU 5410 MIPS / FPU 3730 MFLOPS

SiSoft Sandra 2002 CPU Multi-Media
P4-2.4 = Interger 9446 / Floating-Point 11549
Cel-3.0 = Interger 11927 / Floating-Point 14793

SiSoft Sandra 2002 Memory Bandwidth
P4-2.4 = Int 2279 / Float 2180
Cel-3.0 = Int 2204 / Float 2097

Well, that's it...make up your own mind on whether or not some celery might be good for you. As for me, I'm going to look at the P4-2.0a once they are available with the C1 stepping and priced under $150 USD, which could be by the end of this month! Anyway, I just thought this might be some interesting info to share, the results are presented "for what it's worth", nothing more. ;)
Click to expand...


Thanks for the benchies! I think that in order to be more competitive you need more memory bandwidth on the celly rig. It wont be a freak out difference, but it will mean something since these chips love bandwidth. Just a suggestion though.
 
Re: Re: Celeron vs P4: Test Results

Extol said:
Thanks for the benchies! I think that in order to be more competitive you need more memory bandwidth on the celly rig. It wont be a freak out difference, but it will mean something since these chips love bandwidth. Just a suggestion though.
Click to expand...
You're quite welcome, and I agree with your opinion. It's somewhat ironic, but at the time I got the Cele my Samsung PC2700 died on me. It was stable for several months at 200MHz/DDR400 with just 2.7v, then suddenly it's performance started degrading, now to the extent that it only does 150MHz with slow timings and provides just 1700MB/sec bandwidth. Point is at the moment I have no way to find out how much the Cele performance might improve using faster memory. Oh well, I'll see how it does once the GB chipset mobos are available, as maybe dual-channel DDR will help! :p
 
DDR-PIII said:
the celeron's suck because they have way to little L2 :/ sorry to say :/ but i guess thats why the OC so high
Click to expand...
Well I wouldn't say they "suck", because that's simply not the case when it comes down to daily use of the processor...it's actually got enough power to do whatever I care to, with very reasonable performance. Sure in many/most applications it falls behind the P4, but that in no way implies the performance "sucks".

Now when it comes to OCing, having much less L2 cache MAY enable it to OC higher, but we really don't know if such is the case with the Cele 2.0 yet. Since this is the first C1 stepping Northwood Celeron, I think we'll have to see what the C1 stepping P4-2.0a can do before drawing any conclusions on how 128k vs 512k of L2 might affect OCs. ;)
 
I'm not all that impressed with that CPU. I have been checking the MadOnion ORB and reading reviews around the web and they all show that stock @ 2Ghz, my Duron can beat that Celeron and when OC'ed, an Athlon XP1800+ can still beat it. I was thinking of making a move to Intel, but after comparing prices and performance, I think I will stay AMD.
 
quegyboe said:
I'm not all that impressed with that CPU. I have been checking the MadOnion ORB and reading reviews around the web and they all show that stock @ 2Ghz, my Duron can beat that Celeron and when OC'ed, an Athlon XP1800+ can still beat it. I was thinking of making a move to Intel, but after comparing prices and performance, I think I will stay AMD.
Click to expand...
Though I didn't do a comparison with an AMD system, I've also got a 1600+ at 157fsb for 1649MHz, so you can bet I know how they compare...that rig and the celeron at 3GHz+ are actually fairly close. Of course we all know that 3DM isn't the only benchmark, and there's more to PCs than just benching. If I needed to buy an entire system and had very little cash, I'd have to agree with your synopsis. However in this case, I already had a complete Intel rig sitting around without a CPU, so I thought a little 1000+MHz OC up to 3GHz might be fun! ;)
 
quegyboe said:
I'm not all that impressed with that CPU. I have been checking the MadOnion ORB and reading reviews around the web and they all show that stock @ 2Ghz, my Duron can beat that Celeron and when OC'ed, an Athlon XP1800+ can still beat it. I was thinking of making a move to Intel, but after comparing prices and performance, I think I will stay AMD.
Click to expand...

I was also going to make the change... still thinking 'bout it, AMD do run HOT
 
Want a real price/performance? hehe

Get a P4 2.4/533 and a good copper cooler.
I did 800fsb ( 3.6GHz) on a Gigabyte sq800.
It's less than 130$ now and worths the buck.
Besides, you wipe the floor with any AMD using that 3.6 800fsb P4.

Cheers
 
You must log in or register to reply here.

Similar threads

Maximus Nickus
Tweaks-The Definitive Guide
4 5 6 7 8
Replies
149
Views
61K
I.M.O.G.
I.M.O.G.
Back