Anandtech Looks At 'Celeron 2' 82
Oscarfish writes, "Anandtech has the scoop on the new batch of Celeron chips hitting the market. They're more or less Coppermine chips with half the L2 cache removed, so you basically have a Coppermine core with 128K L2, a 66MHz front side bus, and FC-PGA packaging. A decent choice for the "Value PC" segment, he says, but not for performance machines. "
BP6? (Score:1)
Re:Apparently not... (Score:1)
I dozed off for a moment (Score:1)
It's slower than Athlon and about the same price (Score:1)
ICQ#2584116
Flip Chip (Score:1)
"Flip chip" refers to mounting the chip that is the CPU "upside down" on the circuit board its on. This puts the metal substrate "down", against the circuit, and the silicon "up", where is it that much (1-2mm) closer to the heatsink. The idea is to make them easier to cool.
Re:Still 66MHz FSB? (Score:1)
End of marketing lesson.
Re:It's slower than Athlon and about the same pric (Score:1)
Re:*rolls eyes* (Score:1)
So you're saying that just because YOU don't new a new CPU right now, Intel should altogether stop development of new products, until YOU do need a new CPU? Okay..
Re: That's not how it's done (Score:1)
Re:Why the small cache? Because... (Score:1)
What do you mean ILLUSION? They will have different amounts of working cache... sound different to me...
Does anyone think that Intel bean-counters really care about the OPTIMALITY of the product? No, they care about it's PROFITABILITY
Its hard to make a PROFIT if you are making bad PRODUCTS.
Intel just wants to produce a variety of chips to hit a lot of different price points... what's so wrong with that?
Following the Unix nomenclature... (Score:1)
AMD wins at the low end too (Score:1)
AMD has 1GHz pricing power over Intel
AMD Spitfire beats Intel Celeron 2
AMD Sledgehammer will crush Itanium (esp. 32 bit)
AMD copper interconnect soon, Intel in a year
Be grateful for AMD's competition, or Intel would also be pushing the overpriced underperformingn Rambus down our throats, rather than market realities having to force it to gradually accept that it's a crock of shit.
Re:Didya' ever feel like you were getting ripped o (Score:1)
Re:*rolls eyes* (Score:1)
This was just to say that with the right GFX card you too can join the fun [counter-strike.net]. Your Celeron 433 needn't be half bad, so happy fragging!
Re:Apparently not... (Score:1)
Re:BP6? - nope. (Score:1)
Re:It's slower than Athlon and about the same pric (Score:1)
I haven't met an onboard sound set that I've liked yet, but hey, if you are only planning on paying $499 for a system, the integrated boards are almost acceptable... just don't try to use them for anything besides freecell (they can't handle Space Cadet Pinball 8^D)
That's not how it's done (Score:1)
You're always best off to make your cache footprint as small as possible, all else being equal. In multitasking OSes, if you have carefully used all the available cache, then one task switch kills your context anyway. OTOH, if you have two tasks with 64K working sets, and you have 128K cache, then the task switch doesn't cost much. Similarly, if you have a resident set of 32K, you can run four of these things.
So, different cache sizes don't make all that much difference for compilers or programmers. You'd use the same techniques to write code for 128K cache as for 512K cache.
Also, the 128K thing is not so much about about a 95% hit rate; it's more about the resident set. If you graph performance of a program vs. cache size, you'll an elbow in the graph where a small reduction in cache size causes a large reduction in performance. The location of this elbow in terms of cache size is the resident set of the program.
Studies show that most programs have a resident set of less than 128K, so that will do for most applications.
Actually, most programs have several elbows in the graph, and so have several resident sets. Usually most of them are less than 128K, so you get most of your performance benefit from a 128K cache size.
--
Patrick Doyle
Re:It's slower than Athlon and about the same pric (Score:1)
Intel is in the motherboard business now. The days of upgrading a CPU without changing the motherboard are behind us. If it ain't the socket/slot, it's the VRM.
Re:Still 66MHz FSB? (Score:1)
--
Re:Integration out the wazoo! (Score:1)
--
Hardware Central (Score:1)
Re:BP6? - almost. (Score:1)
Re:Still 66MHz FSB? (Score:1)
matt
Re:The problem with different amounts of cache... (Score:1)
the decrease of yield...
the increase of cost.
That is how is looks to me. The performance boost that comes with bigger cache only appears to effect certain server applications (DB stuff), and is next to nothing for desktop stuff, be it games, office apps, or what have you. It simply doesn't make sense to increase the cost of chips, which is the result of increasing cache due to the decreased yield.
matt
Re:Still 66MHz FSB? (Score:1)
While that was true for the very first i486sx chips, they very quickly had seperate masks for the SX chip. This mask allowed the chip to use ~1/4 less space. (less space per chip = more chips per wafer = cheaper production costs in the long run) I'm certain that that will be the same case for the celeron2 processors as well.
Here's the main reason (Score:1)
Like buying a car without the undercoating, they figure it'll turn into a rust bucket in no time.
More promotion and education is needed. Not technical jargon to confuse people (cause these people are computer illiterate), but cold hard facts that even they can plainly see.
Re:Here's the main reason (Score:1)
Now they are, thier chisp are better, faster, and have more O/C ability, and are just as reliable.
Maybe AMD sucked a$$ back then, but they kick a$$ right now.
Re:Intel has a problem... (Score:1)
BP6 Support??? (Score:1)
-motardo
It's easier to bump up this way. (Score:1)
Oh, the road map says it's time to announce the 733 MHz Celeron II. Let's up the voltage on the 600.
Oh, the road map says it's time to up the FSB on the Celeron line. Let's solder that bit back into the 600. Hey, look at all the money we saved by not designing new chips.
They can't do that with the Pentium line any more because AMD is nipping them in the speed race, so they can't leave the breathing room that they used to.
Re:BP6 (Score:1)
Re:It's slower than Athlon and about the same pric (Score:1)
Re:Didya' ever feel like you were getting ripped o (Score:1)
Two points: 1) less cache can often be easier to produce, as Intel can get better yields, which means lower cost to manufacture... so you're not getting ripped off at all. 2) Underclocking (it's a word) is actually good for the customer, because you get the same proformance at a lower cost. Look at what AMD did (does?) with their low-end Athlons -- they make so many high-speed ones (for volume production), but they can't sell them all at a high price so ship them at lower-speed -- the customer gets a better deal than before (esp if you know how to identify and properly clock it)! Win-Win.
This is the smartest thing they did since Cumine (Score:1)
Re:What about the spitfire? (Score:1)
Your argument about the FSB has its merits but you are forgeting something. Celeron's run at 66MHz anyway. That's slower than my K6-2. Now on some apps the celron will beat a K6-2 becuase it has better floating point registers, but that doesn't make up for Intels lack of effort. K6-2's aren't the greatest either, but Celeron's still suck for SMP systems. And another thing...
First off, the spitfire chipsets will have true 200MHz bus speeds. That speed cannot be atained with PC100 RAM, however becuase of the TYPE of motherboard setup they have (DEC's alpha systems [something like that]) it will be very easy for them to switch the RAM type to DDR. (Unlike Inel's RAMBUS crap)
Second, when they do switch to DDRRAM, it will be close to the same price as SDRAM, becuase DDR is an open standard. DDR is better in so many ways: it can be made in higher volumes, becuase it fabs better; because of it's open standards, it will be chaeaper. (not much more than regular SDRAM) adn best of all, it's faster. Look at the analysis you can find on tom's hardware. Lower latencys, and higher bandwidth!
AMD is already making DDRRAM modules (not AMD specifically, they have companies actually doing the MFG)
People need to keep supporting AMD, I know that Intel has made donations to Linux companies, but I for one don't want to pay that much for something that doesn't even work the way it's supposed to. Competition is good, especially when the competion to the norm is better.
A 300MHZ Celeron is *plenty* (Score:1)
Seriously, I play Quake, Star Wars Racer, Need For Speed III, etc on a piddly $499 box (we did add a TNT2 card for $72). Maybe I'm missing something I'd be getting on a 'kick ass' $2-3K box, but I dunno what it would be...
Re:Didya' ever feel like you were getting ripped o (Score:1)
Doh, thats what chevy did with thier 4.3L V6, its just a 350 with the front 2 cylinders lopped off.
Re:Why the small cache? Because... (Score:1)
Re:Still 66MHz FSB? (Score:1)
Memory Bandwidth (Score:1)
Not again... (Score:1)
kwsNI
Re:Apparently not... (Score:1)
And if they didn't like them, who came up with Celery^Hon?
Overclocking tips? (Score:1)
Why I like it... (Score:1)
Re:Still 66MHz FSB? (Score:1)
Re:Here's the main reason (Score:1)
Thankyou and Goodnight!
Anandtech (Score:1)
Re:Anandtech (Score:1)
Re:Still 66MHz FSB? (Score:1)
AMD has really broken down the fat margins Intel is used to getting for high performance chips, so a move by Intel to try and regain their prior business model isn't overly surprising. Now, whether that model is valid anymore is another can of worms...
I don't have heat problems like that... (Score:2)
Maybe FreeBSD is cooking those CPUs... :-)
New XFMail home page [slappy.org]
/bin/tcsh: Try it; you'll like it.
But UltraSPARC III is coming... (Score:2)
New XFMail home page [slappy.org]
/bin/tcsh: Try it; you'll like it.
Re:But UltraSPARC III is coming... (Score:2)
Agreed. It's kinda late, but still, this is going to be one hell of a processor.
Notice how Suns keep their value? I mean, I'm trying to get an Ultra 1, and can't seem to get one for under $1000 on eBay.
What's with the pissant moderators around here?
Dunno. I see that you are at a 0, but don't see any moderations on your post. Do you have negative karma or something?
New XFMail home page [slappy.org]
/bin/tcsh: Try it; you'll like it.
Re:Still 66MHz FSB? (Score:2)
Personally, I view the new Celeron CPU's with this: yawn.
The best way to speed up computers nowadays is NOT by getting a faster CPU. More effective solutions include adding more system RAM (so your system uses the virtual memory on your hard drive a LOT less) and switching to 7200 RPM hard drive.
In fact, I've played with an Abit AB-ZM6 motherboard running a Celeron "A" 500 MHz CPU with 192 MB of PC-66 SDRAM and a 7200 RPM IBM Deskstar 10.2 GB ATA-66 hard drive and it runs Windows 2000 Professional and Microsoft Office 2000 Professional quite well with no appearent speed problems!
Re:does this chip have the P!!!-style S/N? (Score:2)
------------------
Re:This is the smartest thing they did since Cumin (Score:2)
Re:Integration out the wazoo! (Score:2)
Re:doesn't sound like a big deal (Score:2)
Re:*rolls eyes* (Score:2)
Buying a P2 over a Celeron A is a bad decision. Similarly, buying a P3 over a Celeron 2 will probably be a bad decision.
Instead of padding Intel's profits, try spending the money where it'll actually do some good.
Didya' ever feel like you were getting ripped off? (Score:2)
In the automobile industry there is certainly a place for both the 4 cylinder and the 8 cylinder and we expect too pay less for the 4 cylinder. After all you get a lot less metal in the average 4 cylinder automobile than you do in the 8, but this analogy just doesn't hold up for me when applied to silicon.
Am I being totally unreasonable here?
BTW, I'm actually hoping somebody will set me straight about this. I've recommended Celerons to a few people but always half-heartedly and for some people it does seem like a lower end machine would suffice.
Re:*rolls eyes* (Score:2)
It's not about convincing people who have something that they're happy with that they should upgrade. It's about having a product which is half-way competitive with AMD for those people who are looking for a new computer.
There are still a lot of folks looking for a first (cheap) computer, and an amazing number of antique systems out there. My company just upgraded my software development system from the Pentium-166 that I'd been using. We have tons of 486 systems still in use, including a bunch of SX-25's. We will be buying new computers to replace those over the next few years, and we certainly will be looking for cheap ones.
Integration out the wazoo! (Score:2)
Between Intel & AMD we have:
Socket7
Super7 (100Mhz)
Socket 370
Socket ??? (100 & 133Mhz Coppermine)
Slot 1
Slot 2 (SECC2)
Slot A
With all these different "standards" for putting the processor onto the motherboard, how long will it be before the processors are just stamped onto the motherboard?
We're at the point now, where any processor upgrade requires a new motherboard (either for FSB speed, or connector type, or chassis type). If you buy early and expect to upgrade, Intel changes the rules, if you buy later, your upgrade choices are less than 100Mhz different because the clock multiplyer won't go that high. And of course, everything else under the sun has been integrated onto some motherboards (Video, Sound, Modem, Ethernet, Serial, Parallel, USB, IDE, FDC)
This industry is rapidly moving just like the Automotive industry did about 10 years ago. The trend moved from a car that could be 'tinkered' with in the driveway to eke out more horsepower or gas mileage, to cars that should never be touched by a hobbyist without expensive diagnostic equipment.
The days of tinkering are numbered, and I for one am disappointed.
Am I Missing Something Here...? (Score:2)
Intel releases a chip for less than $200 that easily overclocks to 83/750 MHz with simple air cooling. And that's before anyone's gotten fancy with it...or even considering that the cheaper 566 MHz chip has the option of going up to 100/850 MHz (and remember this is the same core stepping as the new CuMine 850s and 866s).
And people are bitching about this???
Talk about history repeating itself...I remember when the 300A came out, everyone whined about its low bus speed and small cache. Anyone else see the irony?
Re:Integration out the wazoo! (Score:2)
> one am disappointed.
Well, get over it.
Back in the good old days (before ZIF (zero insertion force) sockets) motherboards have ALWAYS been tied to the motherboard they were running on. (8086, 286, 386, early 486). There was little overclocking too, since they used quartz timing chips that were difficult to remove or replace. You basicly used what you had.
Even during the years of the 486, pentium and pentium II, cpus were still more or less tied to the motherboard. There was simply no point in only replacing the CPU. The main reasons for this are:
1) Better resale value: most people looking for older hardware need the motherboard that came with the CPU and pay much better.
2) Newer motherboards have features and performace better than the older motherboards, Eg higher FSB's, better RAM support (although RDRAM is a bit dodgy). What is the point of sticking a 1Ghz cpu into a motherboard that can only do 66mhz, or cannot supply enough voltage. So the most logical thing to do is to remove the old motherboard and sell it with the CPU and get a decent motherboard.
It may be that intel are pushing through extra physical changes to their boards to boost chipset sales, but even if they didn't, the technology changes quickly enough that old motherboards are just useless with older CPU's (the only exception that I will mention is the BX board, which seems to scream no matter what pentium-II/III class processor you use in it)
The days of tinkering may be numbered, but don't imply that they existed of more than a few lucky years.
Re:*rolls eyes* (Score:2)
I don't play those kind of games (I don't like 1st person shoot em up games), and my Celeron 433 is just fine for the kinds of games I play on it. Other people with previous generation Celerons are most likely the same kind of user that I am.
Intel SHOULD develop new products... they SHOULD work on getting the high end processors even better. Not cripple their high end processors and try to get Joe user who uses office apps to "upgrade" just because an upgrade is available.
Re:Puh-leeeease! (Score:2)
The problem with the processor wars is that they just focus on a number (now many MHz they can get up to) when there are lots of other factors in computer performance. People see the 800 mhz and think it must be 2x as fast as their 400 mhz chip. If some of the bottlenecks in the hardware would be resolved, there'd be a lot better performance everywhere.
*rolls eyes* (Score:2)
When will the insanity end??
Re:What about the spitfire? (Score:2)
Yea using PC100 ram... (Non-DDR) see the probelm here?
Of course, the CPU bus does not always talk to the RAM. It also acts as a conduit for information to the PCI/AGP Bus as well, which means that while the CPU waits for a data from the decoupled RAM connection, it could also read/write to the peripheral busses. No, there isn't a problem here.
Even if the PC100 was the only game in town, there are many ways to speed it up, like having multiple banks of RAM. Instead of having one bank for the memory, you have two. Now, in optimal cases, the cpu can read/write two words at once, effectively doubling the memory bandwith.
Re:Why the small cache? Because... (Score:2)
Have you looked at any benchmarks for the high speed chips (where the athlons are running at 1/3 speed cache)? What will you see? The coppermines soundly thrashing (well, by a noticeable, but not very large, margin)
Oh wait, I forgot, AMD is the underdog. Therefore, since a while back their chips were faster (the lower speeds with the 1/2 speed cache) Mhz by Mhz, we're still going to say they're faster even though it's no longer true.
Really though, read up on the performance of the Athlons with the 1/3 speed cache vs. the performance of the Coppermines... Yup, the Coppermines have edged by the Athlons.
Re:BP6? - almost. (Score:2)
Re:BP6? (Score:2)
Interjecting a bit of math
x times the log of y is equal to log y ^ x
so 2 times a celeron 2, would be a celeron 4, because it's celeron 2 ^ 2.
Of course, this is entirely stupid, but what can you expect when you have math on the brain
----
Don't underestimate the power of peanut brittle
As a tried and true overclocking weenie... (Score:2)
Compare this to overclocking a normal P3 - you have to start pushing your FSB OVER 100 - suddely your PC100 ram isn't good enough anymore... your PCI cards start to act fruity - when you up your FSB, after all, you're overclocking EVERYTHING... Forget about even TRYING to overclock your Cu that is designed to run on a 133 Mhz FSB... your PC133 ram will bite the dust around 140-145.
So, once you got your celeron running at 600, 100mhz FSB, you just get your Abit BP6 board, run two of them, and whoosh. Nice power. 'Cause remember, the other thing that made the celeron so great was that the cache RAN AT THE SAME SPEED AS THE CHIP. As opposed to half. So just because there was less cache didn't matter so much 'cause it was a HECK of a lot faster.
Then there's the celeron two - same slow cache as the P3, just less of it. And the multiplier's going to be pretty damn high to make a 600+ Mhz chip out of a 66Mhz FSB... so if you up it all the way to 100, just sit back and watch the fireworks... the era of the Celeron being the perfect chip (value-wise, and even power) looks like it's coming to a halt.
God, I ramble like Katz...
Re:It's slower than Athlon and about the same pric (Score:2)
My understanding on the cache issue and clock (Score:3)
Performance-wise, the average Joe who buys a Celery-based system isn't going to overclock, and the overclockers are a small, but devoted group that will find a way to crank the processors up anyway. The multiplier locks on processors don't really stop overclockers, but they do help stop people from remarking the chips since there's no easy way to tamper.
On the overclocking front, interestingly enough, Athlons are actually pretty overclock-friendly, and the Golden Fingers cards that simplify the process are pretty inexpensive. I run my K7-600 at 750 and it was quick and easy.
Basically, the new Celerys are a pre-emptive strike against the AMD Spitfire CPU's, which replace the K6-2 processor family but use an Athlon core and bus. I think AMD will continue making inroads into the Celery target market, though - the box companies have been burned too many times by Intel of late to put all their eggs in one basket. Dell is the last Intel-only holdout and we'll see how long that lasts.
Depending on pricing, these new Celerys could be a pretty good bargain for the "enthusiast". I may check out the possibility of slapping one in to replace the PII-350 I have in my old Mandrake box at home.
- -Josh Turiel
... cache size isn't everything ! (Score:3)
The problem is a direct mapped cache has only one place for each RAM location (repeating every cachelength). Say you have a program that accesses a memory location, then accesses one [cachelength] away. The first will be flushed from cache to make room for the second. And over again = thrashing. You won't generally be able to predict physical addresses because paging remaps everything.
Two way set associative caches are a big win. Every memory cell can go into _two_ cache locations. That way, an LRU algorithm is used to decide who replaces what, and recently used data is much less likely to be flushed.
I'm very pleased with Intel's P6 four way set associative caches. When comparing cachesize, I multiply size by associativity. So a 128 KB 4way cache is as good as a 512 KB direct mapped.
This is definietly an oversimplification. There are a few problems/OSes that can keep their entire working set in L2 cache. Then large direct mapped is a win. But they are rather rare compared to problems that exceed L2 and need to coprocess 2 or more data elements.
Still 66MHz FSB? (Score:3)
--
Apparently not... (Score:3)
I'm still a little surprised they didn't go with Pentium Value for the Celeron. I mean, they paid the hundred bucks for the trademark, and they'll be damned if they can't squeeze every last cent out of it.
Re:BP6? - almost. (Score:3)
Why the small cache? Because... (Score:4)
The 'new' Celeron, with another cache level is being introduced to add confusion to the market. It will be marginally cheaper than the larger cache version, so cost of production is not the issue.
In fact, this Celeron will probably start life as the larger-cached version, and have half it's cache disabled (intentionally destroyed) to offer the customer an ILLUSION of having more choices.
Intel has done this before:
386SX was a 16 bit 386DX. Valid difference.
486SX was a 486DX, with a non-functioning floating point unit. The DX chips that burned out their FPU during testing were relabeled as SX and sold at a discount. This way, Intel sold that which it would have otherwise thrown away.
How is this different? This is intentional. Intel knows that they can't beat AMD on performance, so they will flood the market with variations of the same chip. Intel is betting on the fact that Joe Q. Average will see how many 'different' processors Intel produces, and conclude that Intel is a better investment.
The performance penalty that Joe Q. Average will suffer by running a chip with a cache that is half the size for which most code is optimized is not really an issue for Intel.
Does anyone think that Intel bean-counters really care about the OPTIMALITY of the product? No, they care about it's PROFITABILITY. The stock-holders are the company's first priority.
As for the technical details.. CAVEAT EMPTOR, as always.
More on the Celeron 2 (Score:5)
Here's a HardwareCentral review [hardwarecentral.com] of the Celeron 2 (today must be the day NDAs were lifted; look for other butt-kicking sites like Thresh's [firingsquad.com] and Sharky's [sharkyextreme.com] to maybe have something on it later today.
Here's BP6.com [bp6.com], an excellent reference for those of you with that funky Abit board. Check out the video preview of the Powerleap FC-PGA adapters - basically they plug into Coppermines and allow two of them (new stepping ONLY) to run in SMP mode. Of course, your BP6 would be running at 100MHz FSB by default - and overclocking well past 100MHz (which is what is required to unlock the true potential of Coppermines) is flaky on any BX board.
Coppermines seem, for me, an excellent buy. I have a 500E running at 733Mhz (147MHz FSB) on an MSI MS-6309 [firingsquad.com] Apollo Pro 133A board. Excellent performance, and super stable.
The 66MHz FSB for these new Celerons is a double-edged sword. It's good that the 66MHz+ gap is open, which is really what made the original Celerons such good overclockers; but besides the performance hit (naturally), the lower FSB means a higher multiplier. The internal multiplier (locked by Intel) for the 600MHZ Celeron 2 is 9.0x. That's ass-high, people. I don't think many motherboards currently support that. At the very least I think a BIOS upgrade is in order, unless you're absoluely sure the board can handle that high a multiplier - but getting back to the performance hit, not only is your memory, etc. running at only 66MHz, but with the high multiplier your chip is running 9 times faster than your system. That's a low of waiting on its part.
My advice? Get a 500E or 550E (both can be had for around $200, if you know where to look) and overclock them beyond insanity. 150MHz FSB is not out of the question for these chips, especially the ones with the new core stepping. I'll be going for a 600E (FC-PGA) as soon as school [ssu.edu] lets out for me for the Summer.
For a truly bent journalistic look at the Coppermines, check out this piece [overclockers.com] I wrote for the fantastic Overclockers.com [overclockers.com] over Winter Break.
The problem with different amounts of cache... (Score:5)
When you are optimizing memory locations in your code (the OS or the applications), you optimize for vitual memory page sizes for memory usage, and for cache sizes to make sure that your application gets the most use of the cache as it can.
In a world where we get precompiled binaries (ideally you could teach the compiler to optimize for the different levels and compile it for different amounts, additionally programmers could put flags in their code for the different optimizations), how to we optimize our applications?
If the Celeron2 becomes really popular, then we optimize for 128K and 256K machines or 512K machines get little to know benefit from their larger caches. If we optimize on 256K or 512K, our smaller caches may get unacceptable cache miss rates and suffer poor performance.
The idea behind this is that 128K is aleady ~95% cache hit rate, and doubling the cache only gives small improvements. However, when Pentium Classic machines used to ship with 512K on the good boards and K6s and K6-2s were shipping on mobo with 1MB cache, why are we now switching down the amount for higher speed? I understand why it is faster (90% of L1 no slow down, 95% of L2 1 missed cycle; faster than 90% of L1 no slow down, 98% or 99% L2 of 3 or 4 missed cycles and memory access is 6-8 missed cycles)? Are our programs out there optimized for the older "larger" caches or the newer, "smaller" caches. It seems like we are going to see less and less advantage to the expensive systems with lots of cache when we are optimizing for the workstations with much less cache.
Alex
Intel has a problem... (Score:5)
From where I'm sitting, Intel currently has an interesting problem. Their flagship processor line, the Pentium III Coppermine, tops out at 1GHz . But there aren't that many around, and the current crop of PIIICu's seem to top out at 800MHz. Overclockers can squeeze this up a bit, but it seems that the PIIICu's in the shop are close to their limits anyway. Celeron's have always proved to be seriously overclockable - most 500MHz will go to 600MHz plus, which leads me to believe that Intel is not being anything like as strict in the CPU speeds it's binning its processors into.
In other words, the Celeron processors are not being tested hard as they come off the production line as there is an adequate supply of the speeds that Intel wishes to see. Why doesn't Intel want it's Celerons to be sold as close to the limiting speeds as possible? In my opinion, it's because they don't want to eclipse their PIIICu flagship chips with Celeron chips running at the same clockspeed as this would dilute their market with their own product.
So what you cry? None of this would matter so much if AMD wasn't quietly pulling ahead in the high-speed chip fabrication stakes. With the Athlon coming off the production lines at clock speeds 850, 900, 950 and 1GHz, there is considerable pressure on the PIIICu's to remain visible in the marketplace, since they are the direct competitor to the Athlon. The Celeron 2 looks to be throttled by it's bus speed (66MHz?) in comparison to the Athlon at 200MHz, and won't compete in heavy memory fetching tasks, such as games, art programs, complex DTP etc. Until Intel can successfully ramp up the speeds on it's headline brand, the increasing speeds of the Celerons and Celeron 2's present an interesting dilemma.
Cheers,
Toby Haynes