TwiceSliced
Gawd
- Joined
- May 1, 2002
- Messages
- 525
Got one system, two processors. Does the Celeron's higher speed compensate for its smaller cache?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
P3 500E or Cel 800?
- Thread starter TwiceSliced
- Start date
robberbaron
Supreme [H]ardness
- Joined
- Sep 6, 2004
- Messages
- 6,101
For sure. When it's 60% higher in clockspeed, I wouldn't worry about the cache.
Pkirk618
Supreme [H]ardness
- Joined
- Sep 12, 2002
- Messages
- 5,737
back in the day, I remember the celly @ higher speeds giving better results even with half its big brothers cache.
The Celeron is faster.
I had a Coppermine Celeron 1.1@1.25 that was faster than a Pentium 3 1ghz. Even my old Celeron 600@1035 came close to the 1ghz Pentium 3. The Coppermine and Tualatin Celerons were not as bad as everybody says they were. The celerons to avoid are the Covington, Willamette, Northwood.
I had a Coppermine Celeron 1.1@1.25 that was faster than a Pentium 3 1ghz. Even my old Celeron 600@1035 came close to the 1ghz Pentium 3. The Coppermine and Tualatin Celerons were not as bad as everybody says they were. The celerons to avoid are the Covington, Willamette, Northwood.
Woops, I thought the 500E was a 133FSB chip but after consulting the spec finder I see now that its only a 100FSB chip and thus retract my thoughts that it might be close. Use the Celeron as it should infact be faster, and would more than likely overclock well too!
S
SirKenin
Guest
Nothing wrong with the Northwood Celerons. I have one.acascianelli said:
lithium726
[H]F Junkie
- Joined
- Jun 16, 2004
- Messages
- 8,195
compared to other celerons, the northwoods sucked. even the tualatin celerons had more cache than them, and they didnt need it near as badly.SirKenin said:
S
SirKenin
Guest
They perform just fine and overclocked as well as any of them. I have one running in a server and it hasn't choked once. What data are you using, or are you talking out your ass?lithium726 said:
http://www.sharkeyextreme.com/hardware/cpu/article.php/1497501
Incidentally, it is the 2.0 Celeron that I have.
Naw, NW Celerons aren't that great, except maybe if you get a good overclock out of one. The 400MHz FSB and 128KB L2 are a double whammy for bad performance. Overclocking helps the first out.SirKenin said:
edit: before you flame me, yes I have owned several NW core Celerons, desktop and mobile versions. If you have a server that's not under a huge load of course it will work. I can make a 400MHz G3 Mac work as a decent web server (and I do). And I would rather have a NW Celeron 2.6GHz than a Tualatin Celeron 1A GHz @ 1.33GHz. But for relative GHz performance (i.e. vs NW P4), the NW Celerons were pretty poor.
S
SirKenin
Guest
It's been proven that cache isn't the be all end all. Just look at the lack luster increase in performance between the 512k and 1MB cache. Personally I wish the Celeron had more cache, but hai, they still work just fine and are a good CPU for the money.pxc said:
Yes, there are some types of applications that aren't affected by the amount of cache. Most applications and games are not like encoding apps.SirKenin said:
The difference between 128KB and 512KB and 512KB and 1MB is a lot more than you imagine. You should read a review or two on Celeron scaling. The 128KB L2 cache and 400MHz FSB do severely affect NW Celeron performance.
start here and read the next several pages of benchmarks: http://www.tomshardware.com/cpu/20041221/cpu_charts-13.html#opengl
Except for encoding, where the NW Celeron performs half-way decently, the performance of the NW Celeron is horrible compared to NW P4s and comparable AMD CPUs.
S
SirKenin
Guest
If you ever want to me to talk to you like you are human, you will never give me another link from that jerk at THG.
With that said, I'm familiar with about every processor from both camps (with the exception of a couple of the AMD chips). I studied in depth how they all worked and what made them tick. I don't need some loser at THG who changes his story as often as he changes his underwear to attempt to tell me about CPUs.
I realize that they take a performance hit over the Northwood P4, however we are not comparing P4s and Celerons. Compared to every other Celeron, there is nothing wrong with the Northwood Celeron. I use one. It works fine. There is nothing wrong with it. The guy before me used it to play games. At the end of the day benchmarks don't matter. It's whether or not the CPU does what you want it to do. If it does that without overheating or blowing up, fine.
With that said, I'm familiar with about every processor from both camps (with the exception of a couple of the AMD chips). I studied in depth how they all worked and what made them tick. I don't need some loser at THG who changes his story as often as he changes his underwear to attempt to tell me about CPUs.
I realize that they take a performance hit over the Northwood P4, however we are not comparing P4s and Celerons. Compared to every other Celeron, there is nothing wrong with the Northwood Celeron. I use one. It works fine. There is nothing wrong with it. The guy before me used it to play games. At the end of the day benchmarks don't matter. It's whether or not the CPU does what you want it to do. If it does that without overheating or blowing up, fine.
In your clouded rage, maybe you couldn't notice that Tom didn't write it. Regardless, it's an accurate review. Type Celeron 128kb review into google and you'll find dozens of reviews with the same results.SirKenin said:
I see no point continuing this because you ignore facts. I don't have time to waste with unreasonable people. Sorry.
lithium726
[H]F Junkie
- Joined
- Jun 16, 2004
- Messages
- 8,195
i was speaking in relative terms. take a 1ghz p3 and run it along side a 1ghz celeron. the difference is not as big as a 2ghz p4 to a 2ghz celeron.SirKenin said:
look at the page you linked - the PC mark 2002 CPU and memory test looks decent. then all the rest are abysmal. (all the rest are gaming graphics, though) the 2.0 NW (and the 3.0) is getting its ass kicked by a 1.8A. thats pretty sad.
note the only thing it looked halfway decent in were the synthetic benchmarks.
He's taking that it works in place of performance.
I can't think of any CPUs that don't work well except the old overheating 5v Cyrix processors, some messed up K6 and K7 processors (3dnow!/MMX validation bug), the Pentium I 60MHz-100MHz FDIV bug and the overclocked Coppermine PIII 1.1GHz. This excludes buggy clones (cough cyrix cough amd cough) from the pre-Penium I era.
Of course all processors work. But a NW Celeron has severe performance problems compared to P4s and AthlonXP/Durons.
I can't think of any CPUs that don't work well except the old overheating 5v Cyrix processors, some messed up K6 and K7 processors (3dnow!/MMX validation bug), the Pentium I 60MHz-100MHz FDIV bug and the overclocked Coppermine PIII 1.1GHz. This excludes buggy clones (cough cyrix cough amd cough) from the pre-Penium I era.
Of course all processors work. But a NW Celeron has severe performance problems compared to P4s and AthlonXP/Durons.
I agree. 128KB difference in cache is nothing compared to the 300MHz clock speed difference when both have 100MHz FSBs.pstang said:
It would be a harder choice between the PIII 500MHz (100MHz FSB) and Celeron 733MHz (66MHz FSB). The Celeron 766MHz (33MHz faster) isn't much faster than the PIII 500MHz, and even lags in a couple of gaming benchmarks: http://www.anandtech.com/showdoc.aspx?i=1368&p=6
S
SirKenin
Guest
Nope. I got it on a trade, and a very VERY good trade at that.pstang said:
S
SirKenin
Guest
You are ignoring the facts because you're the type that thinks that bigger numbers = bigger e-penis.pxc said:
I go by what works and I stand by the fact that it works very well, benchmarks or no. I don't need that crap website to tell me whether or not the Celeron works, thank you.
S
SirKenin
Guest
True. So to contribute I would go with the Celeron as opposed to the 500.pstang said:I may be out of line, but
SirKenin, your northwood celeron is awsome, but you have not even contributed to the thread topic, all you are doing is thread crapping
I've seen several benchmarks of a Tualatin Celeron 1A GHz outperforming a NW Celeron 2.4GHz. I linked it above.acascianelli said:
The NW Celeron is really pathetic in some things.The Celeron D was a very good thing to replace the NW Celeron. I upgraded my nephew's system a couple of months ago from a Duron 1.5GHz @ 1.8GHz to a Celeron D 2.4GHz @ 3.3GHz. The Celeron D @ 3.3GHz performs between a P4 2.66GHz and P4 2.8GHz/533MHz FSB, and is noticably faster than the overclocked Duron.
pxc said:I've seen several benchmarks of a Tualatin Celeron 1A GHz outperforming a NW Celeron 2.4GHz. I linked it above.
The Celeron D was a very good thing to replace the NW Celeron. I upgraded my nephew's system a couple of months ago from a Duron 1.5GHz @ 1.8GHz to a Celeron D 2.4GHz @ 3.3GHz. The Celeron D @ 3.3GHz performs between a P4 2.66GHz and P4 2.8GHz/533MHz FSB, and is noticably faster than the overclocked Duron.
I had my Celeron 1.3@1.6 in an old motherboard, and it could match my brothers 2.8C in some synthetic CPU benchmarks.
Ouch. I never bought higher Tualatin Celerons than the 1A GHz. I got overclocks to 1.33-1.45GHz on those easily (no voltage bump) and the CPUs were pretty fast.acascianelli said:
I meant above that I measured real world performance on the Celeron D @ 3.3GHz between a P4 2.66 and 2.8GHz (non-HT version). Synthetic tests showed the Celeron D @ 3.3GHz at about the same speed as a P4 3.2GHz, but that's not true.
lithium726
[H]F Junkie
- Joined
- Jun 16, 2004
- Messages
- 8,195
yeah, i know sandra shows my dual 1.75ghz pIII's topping a 570 in some things, but i would question the validity of that (note: this is only is multithreaded things, and HT was onpxc said:I meant above that I measured real world performance on the Celeron D @ 3.3GHz between a P4 2.66 and 2.8GHz (non-HT version). Synthetic tests showed the Celeron D @ 3.3GHz at about the same speed as a P4 3.2GHz, but that's not true.
)
S
SirKenin
Guest
Well you could always be like any NORMAL human being and just use them for what they are instead of caring about the size of the e-penis. At least the Celeron didn't melt down like the Axp did. It is a reliable workhorse in all it's flavors. Not the fastest chip around, but it works great for the money.Josh_B said:
SirKenin said:
I'm not worried about the size of anything. Sounds like you're the one worried about size...
Both Intel and AMD have produced CPUs that ran slightly warmer than most people liked. Very few "normal" reviews complain about the blast furnace created by the Prescotts, but whenever someone mentions an Athlon XP, it's a big deal. (With proper cooling, either CPU can be tamed.)
If you're going to complain, then at least try to be objective.
That's because 84W LGA775 Prescotts up to 3.4GHz have been available for a long time (vs "cool" 82W NW 3.2GHz CPUs) and even 89W S478 E0 Prescotts up to 3.4GHz (or 89W NW P4 3.4GHz) aren't exactly a "blast furnace." It's hard to complain about the past when a better product is in hand, unless you want to cling on to a belief.Josh_B said:
There are older 103W Prescotts (1 sSpec in LGA775 @ 3.4GHz, 3 sSpecs in S478 @ 3.2GHz), but those are pretty old steppings. The 3.6GHz and 3.8GHz Prescotts are their own sad story @ 115W, so complain away about those.
The 2.8GHz-3.4GHz versions sell in much higher volumes and also run much cooler (84W in LGA775, 89W in S478).I detailed the power consumption of NW vs Prescott-1MB here.
pxc said:That's because 84W LGA775 Prescotts up to 3.4GHz have been available for a long time (vs "cool" 82W NW 3.2GHz CPUs) and even 89W S478 E0 Prescotts up to 3.4GHz (or 89W NW P4 3.4GHz) aren't exactly a "blast furnace." It's hard to complain about the past when a better product is in hand, unless you want to cling on to a belief.
There are older 103W Prescotts (1 sSpec in LGA775 @ 3.4GHz, 3 sSpecs in S478 @ 3.2GHz), but those are pretty old steppings. The 3.6GHz and 3.8GHz Prescotts are their own sad story @ 115W, so complain away about those.
I detailed the power consumption of NW vs Prescott-1MB here.
I am not arguing that the newer steppings are much better in regard to power mgmt., but rather that there seems to be a bias in some reviews when it comes to different vendors' products.
Regardless, I still maintain that Celerons are relatively underpowered when compared to various other CPUs available at a similar price point. Even going with a used P4 might be better IMO.
In any case, I am sorry to take the thread off topic.
It went off topic several posts ago, with the few posts on topic recommending the Celeron 800 for the orignial question.Josh_B said:
I would agree if E0 was "new," but it's been out for almost a year (in OEM systems, 9 months in mail order/retail). And Prescotts have only been out for 20 months.
With the Celeron 800... was that not the time at which Intel finally allowed the Celeron to run at 100MHz FSB?
If it was, then the Celeron is most likely the better choice. The PIII 500E runs at the same bus speed, but would face a 300MHz clock speed defecit. A PIII 500E may very well be at least marginally slower than the Celeron.
If it was, then the Celeron is most likely the better choice. The PIII 500E runs at the same bus speed, but would face a 300MHz clock speed defecit. A PIII 500E may very well be at least marginally slower than the Celeron.
Josh_B said:Ack!
Newer doesn't have to mean "released yesterday", you know.
Now we're just bickering. I could argue that the vast majority of Prescott 1MB chips ever produced are E0 stepping and making statements about hot Prescotts up to 3.4GHz vs NW P4s is not accurate since they are very close or equal in power consumption. But this horse has already been beaten to death. C0/D0 = bad, E0 = good.
S
SirKenin
Guest
Pentium Pro pwnz j00.