Slashdot Mirror
New Intel Celerons
jht writes "Intel announced new Celerons (a 366 and 400 model) this morning - still a 66 MHz bus, but now they offer a low-cost chipset (the 440ZX-66 AGPSet) that allows you to use the new Socket 370 form factor - finally, good old baby-AT size motherboards can reappear! Welcome back, I missed you.The reference board they show is nice and small. I wonder just how overclockable these might turn out to be, and how long before Intel really starts to cannibalize the Pentium II cash cow with Celerons. "
If you think Intel are going to be silly enough to
let us overclock 300A's to 464, think again..
Oh, shucks, I hoped we would get rid of those evil things. Nothing tests my wits more than those crappy Baby-AT cases with cables dangling all around inside. And they're fugly!!!
"By now we are pretty much at the end of the sweet days of overclocking. The upcoming Intel chips Katmai and the socketed Celeron will not allow any form of overclocking anymore and it would be no surprise if even the well known Pentium II and Celeron chips we know now, will soon also be equipped with Intel's new overclocking protection. AMD seems to consider the same, so that there will be hardly any new chip available that will allow us to tweak its performance by running it faster than it is supposed to."
Read his Overclocking Special
Probably not, but you can go see a real crazy person who did manage to do it....
Click:
HERE
And see how!
As much as I'd like to see the socket boards to return, we don't need YET ANOTHER propetairity (sp? blah) socket to come out.
Intel is only doing this because the Celery chips will fit just fine into a plain old Pentium II socket, and work with the standard chipset, without modifications.
(Remember the NX (I think) chipset? That was supposed to be the only one that worked with the Celery chip. I don't think anyone even made boards with that chipset on them)
Oh well.. at least in this day and age Intel doesn't have the market share it thinks it has. What is it down to, 53%? There's plenty of choice here. I just hope the other manufacturers will adopt a standard, and leave Intel out in the cold.
Consider this - the majority of people who overclock are us poor bastards who can't afford the speed / stability that a true PII-450 supplies. So what is the other possibility for people looking for high(er) performance at a reasonable price? AMD, of course. The Celeron is a double edged sword - for the low end pc market, the Celeron is a worthy competitor. For the high-end gaming market, an overclocked Celeron blows away a AMD K6-2 with 3D Now! My predicition - Intel holds on to all of the market share they can (effectively capturing two major markets; they are the only high-end option), squeezing AMD dry. Once AMD is out of the picture I can guarantee you there will be no more overclocking...
Intel 2
AMD 0
James
... and not the first time.
1. Intel once again released a proprietary chipset, incompatible with anything else. If they had released it for socket 7 (or Slot 1) they'd have much more luck selling it. The new Celerons might be cheaper but you have to buy a new board for them.
2. The new boards are AT? WTF???
3. AMD K6-3 will be a much better choice for upgrades and I think for new systems too. AMD has always had a better performance / price ratio.
4. In the event that these Celerons do catch on, they'll eat the sales of high-end PII's the way current selerons ate sales of lower-end PII's
This chipset's name is a bad omen re speed...
How many of you remember the Sinclair ZX series? My first computer was a ZX Spectrum.
A friend of mine has gone over two weeks without having to reboot Win98 and has a K6-2 300. I also have a K6-2, and I haven't had to reboot my computer in almost that long as well. I am running NT 4.0 with SP 4 and usually my rebooting consists of going to the other partition to use Linux. NT has only locked up on me twice in the span that I could get 9x to lock up quite a few times...
by the way, this is Sharks... I am on winter vacation and left my password at school.
K6-2
K6-2
K6-2, BABY!
350 MHz chips for $100
AGP motherboards w/1MB cache, AGP, 100MHz bus: $90
at the local store!
Beat that, Intel!
(okay, the floating point on the k6 still sucks but otherwise it kicks ass...)
When will the K7 go out? I want one of them to have a better box than my friends.
Hmm, not sure what you did wrong then, I have a k6-2 300 with a crappy motherboard, and I have uptimes of around a month under windows nt... And that's with heavy usage too (2d/3d graphics, distributed.net, games, etc). I'll admit it's not faster than a intel 300, but I knew that when I got it. I think it was awesome for the price.
The K6-3 may outperform the PII in Winstone and other integer or system level benchmarks, but the FP performance will be no better than the K6. Until AMD can convince more developers to use 3DNow!, the K6-3 will still be a lousy gaming CPU for all but Quake 2.
What you have just written is Intel's own proganda.
As Thomas Pabst said on his page, they could solve the problem of shady dealers re-marking their CPU's simply by adding registers to the chip that contain the true specs, which can then be read by software; and hence not needing to add overclock protection.
Intel has declared war on the poor schmuck.
Prior to that, I was running a K6-200. It ran continously for almost a year with similar reliability.
I've been pleased with the performance of both processors, having no significant slowdowns in any of the apps I use: Photoshop, MS PhotoDraw (heresy I know), WordPerfect with ViaVoice, lots of games, NetObjects Fusion.
All of this with the distributed.net Bovine client running (continously under both OSes), and lots of other background tasks (a LapLink remote control host always running; Netscape Messenger always running [though it seems to crash after about a week on Win98]) and lots of moving in and out of lots of apps.
So there :-)
I have the 300A running at 472mhz on the asus p2b board. Enabling turbo in the bios increases the 103mhz bus to 105mhz. It will still show up as 463mhz but if you run benchmarks you can clearly see that it IS faster with turbo on. It may be less than a 2% increase in speed but every little bit counts.
-A.P.
--
"One World, One Web, One Program" - Microsoft Promotional Ad
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
Check out Anandtech's Celery Report, a report on the overclocking of Celerons. A cheap $CND 129 Celeron 300A can quite easily be brought up to 450, with performance at or above that of a Pentium II for many tasks. I've seriously been considering getting one...
I'm not sure how accurate this assertion may be but... I heard it from a tech at one of the big Retail chains. To wit: The Celeron is going to cease to be available in it's current configuration and will go to Socket7. Oh, and they are seeking to get rid of overclocking too.
:)
Mind you, this is just a rumor but it is from a fairly reliable source.
If this is true then it means that Intel is trying to retain Slot1 space for the PII only and capture some of AMD's market in the socket7 space by moving the Celeron there. Such and action would kill two birds with one stone... or is it kill one bird with two stones... Oh well, you get my meaning.
I guess I'll be sticking with AMD until Intel changes thier tune...
Codifex Maximus ~ In search of... a shorter sig.
According to german c't there are certain hopes that the new Celerons can't be faked and repackaged. That would also be the end of overclocking, though.
Unlike what one might believe by reading jht's comment (he didn't say anything explicitly about this, though), it seems Intel won't go back to AT boards. They leave that to other manufacturers, as they have done until now. The reference board depicted on the marketing/promotional document is actually a rather nifty MicroATX design, with the misterious Socket 370 instead of the damned Slot 1.
:o).
I won't be buying one of these, though; I'll intead wait for the K6-3
In Soviet Russia, Jesus asks: "What Would You Do?"
Not exactly old news.
Anyway, I doubt the new parts would overclock
much better than the 300A's.
I don't think people understand why Intel is so against Overclocking.
It isn't because some poor schmuck can't afford a 400 Mhz chip, so he buys a 350 Mhz chip.
It's because some poor schmuck who can afford to buy a 400 Mhz chip walks into CULater computers plunks down the cash to buy what he thinks is a machine with a 400 Mhz chip in it, only to actually get a 300 Mhz chip overclocked.
Yeah, sure it might work. But it's dishonest. The customer essentially gets screwed, paying more for less. Intel certainly cares because they had a customer willing to pay for the more expensive chip, but instead the money went to line the pocket of the dishonest dealer.
I had a company pull that on me once. It was quite a few years ago, but a machine came overclocked and we had tons of trouble with it. That was when I discovered the machine was mislabeled and boy did I raise holy hell with the vendor.
I'm confused. Is Socket 370 the same as Socket 7?
(How many programmers does it take to change a lightbulb? None, it's a hardware issue, call technical support)
Andrew Oakley - www.aoakley.com
>The K6-3 has ONBOARD L2 cache. It's going to be slightly faster. Note: I said SLIGHTLY. The >K6-3 reviews I've seen have shown they're faster than equivallent P2's by only a VERY slight >margin. A couple points one way or the other does not a spanking make. Definitely not what I'd >call "considerable".
No, the "considerable" is the difference in price. In terms of Price/Performance AMD always beats Intel. But in terms of pure performance, AMD still lags behind Intel's high end chips. But, they have never been closer then they are right now to overtaking Intel's lead.
-Paul
"I'm nobody suspicious... That makes me sound even more suspicious, doesn't it?" - Spike (Cowboy Bebop)
If socket 8 is electronically compatible, why not just make a socket converter so those of us with PPro systems can go somewhere other than the expensive PII overdrives. Maybe dual celerons would work on a dual socket 8 board...
The new packaging is going to be nearly identical to the old PPro socket. But it won't be pin-compatible (you cannot plug a PPro into a Socket 370, and you cannot plug a new S370 Celeron into an old PPro board).
What's this going to do to Celery Sandwiches though?
Chas - The one, the only.
THANK GOD!!!
Chas - The one, the only.
THANK GOD!!!
Bah!
Bah!
Oh, and before I forget....
Bah!
The K6-3 has ONBOARD L2 cache. It's going to be slightly faster. Note: I said SLIGHTLY. The K6-3 reviews I've seen have shown they're faster than equivallent P2's by only a VERY slight margin. A couple points one way or the other does not a spanking make. Definitely not what I'd call "considerable".
Wait till there's solid benchmarking on OFF THE SHELF procs for both the K6-3 and the new Celeron-A's. Don't start blathering on about vapourwares until then.
Anyhow, I stole the following quote from off the site.
The K7 is AMD's Xeon
And that's about DEAD ON. The price point that AMD is looking for with the K7, while not quite on the level of the bleeding-edge Intel chips, is aimed at a CONSIDERABLY higher strata than the current K6-2 and 3 chips are. Besides which, you're going to have to buy, in ADDITION TO THE PROCESSOR:
Do the math. Intel's STILL going to come off cheaper.
In addition, you're probably going to see a LOT of complaints (similar to the launching of the LX chipset) and compatibility issues with the K7 arise early on.
The only people who are probably going to wind up buy the K7 are those poor sods who must ABSOLOUTELY have THE fastest setup of the moment.
I have better things to spend my cash on.
Chas - The one, the only.
THANK GOD!!!
Chas - The one, the only.
THANK GOD!!!
Slot A IS going to be used for K7/Alpha.
This is socket 370. It's another format brought out by Intel.
Chas - The one, the only.
THANK GOD!!!
Chas - The one, the only.
THANK GOD!!!
The folks at FiringSquad (http://www.firingsquad.com/hardware/celeron366) managed to overclock their pre-release Celeron 366 to 550 mhz. The chip they tested was an engineering sample so it didn't have any overclocking-protection circuitry.
According to the reviewers, even at 550 MHz, the chip never exceeded 53 degrees C.
LOL! a lot!
I see even classic Slashdot is now pretty much unusable on dial up anymore.
Nope ... another proprietry one.
:))
...
Pretty soon we'll have 5 to pick from
Slot 1 - Dechuttes, Katmai (spelling error detected
Slot 2 - Xeon (can't remember if this is Slot 2 or S.E.C.)
Slot A - K7
Socket 7 (and Super 7) - K6-3, Pentium and the rest
Socket 370 - Celeron
Joy Joy Joy
And the word is defintely no overclocking on the new Celeron's too.
Well, I've had a K6-2 for about three months now (333 overclocked to 350), and it only crashes in one game (Jane's WWII Fighters), and runs continuously otherwise. I run Win98, occasionally Linux, with a variety of applications, including Bryce 3D, photo-editing packages, games etc. Prior to the K6-2, I had a K6-200 overclocked to 225, and that never caused problems either.
Graeme
NO.
The K6-3 has onCHIP L2 cache, turning your motherboard's cache into L3 -- this will give it very good performace, especially when multi-tasking several different apps. And even if the K6-3 is within spitting distance of the PII in performance benchmarks, the introductory price will be around $100 less than the comperable Intel chip (unless Intel drops their prices again, which wouldn't make me cry at all:)
As far as the K7 goes, bashing vaporware is about as productive as praising vaporware. I don't think anyone has ever doubted that it is targeted at the high-end server/workstation market. And as far as buying new supporting hardware goes, Intel mastered that trick when they pushed ATX and DIMMS on us all. I know this much: I will be looking at buying a new web server in about 10 months, and I will definetly be looking at a K7 with 8 MB of cache!
Same thing; AMD liscensed the technology from DEC/Compaq when Intel wouldn't let them use Slot 1. Guess AMD just didn't want to use the "Alpha" name in their product.
Intel isn't trying to reserve the Slot X market for PII's, it is just that that dinky little card that the celeron is currently built on raises the price, both for Intel to manufacture and for the big OEMs to integrate (it's harder to build a robot to plug the card in, make it firm, etc. than to just plug a PGA chip in to a socket). The socket makes it easeier for OEMs to make computers with 1 giant board that has all the components integrated on it (ick) -- somethinng that is very popular with P5-class machines.
BTW, the LX and BX chipsets were always designed to run Celerons as well as PIIs. The EX chipset (the one the poster was trying to think of) was designed to be a very low-cost, Celeron-only platform (it leaves out SMP, half the DIMM sockets, and some of the PCI and ISA slots), but I think Intel was the only company to make a motherboard with it (a crappy little thing, but you could buy them in bulk for about $40 each).
IMO the prevention of overclocking is so that
Intel can maximise its revenue by selling chips
of different clock speeds. I'd like to see Intel
come clean and instead of fixing the clock speed
use a 'We've tested it at xxx MHz and it works
ok. Overclock at your own risk' approach.
My 233 k6 OC'd to 250 has been up for 43 days! no problems, linux with loads up to 4.6! Of course win95 won't even boot while its OC'd, don't know why, don't expect i ever will...
$ uptime
1:20pm up 43 days, 14:17, 2 users, load average: 1.00, 1.00, 1.02
James Flom
OK,
For now at least, it has been said on anandtech and (IIRC) bxboards,
that these things are overclockable, for now at least.
83 might be the highest I'd try the bus at without some wicked cooling.
What I consider the most promising piece of tech associated with this chip,
are the slot 1 adapter cards.
I'd really like to see if anyone can alter them to get them to run dual on a BX board.
doing it on the riser board should have the advantage of not having to
purposely damage the CPU itself to do the hack.
I mean, heck, if they cut out the OC-ability of them,
but we can still get them to run dual...
well I could certainly have fun with a dual 400.
My K6-300 (not the K6-2 you gripe about, alas) has been up for 20 days, since it came back up from a power failure that outlasted the UPS. Matter of fact, the previous reboot was due to another power failure...