Intel Pentium III 500E CPU and 550E FC-PGA Review

An anonymous reader says "This article on the Intel Pentium III "Flip Chip" 500/550E shows some great overclocking potential for this CPU. " Its a fairly technical article, but a nice one.

Comments

Interesting

[jawad]

I was reading this article last night...

Seems like the PIII-xxxE's are more in line with the Athlon performance numbers than the straight up PIII-xxx's. I have a PIII-550, and according to those charts, the PIII-550E blasts my chip out of the water.

Now it seems like it's a question between the PIII-xxxE's and the Athlons... Does anyone have information on motherboards?

Not For Me...AMD all the way.

[Peale]

Personally, dollar for dollar, I'll stick with AMD. I always vote for the underdog. First Post? 8)

I'm sorry. What I meant to say was 'please excuse me.'
what came out of my mouth was 'Move or I'll kill you!'

Re:Not For Me...AMD all the way.

[pulski]

Some people complain that the AMD CPU's performance is often lacking in some areas, but if you compare the AMD with an equally priced Intel CPU, you will find that the AMD can hold it's own in most situations.

-----

Re:Not For Me...AMD all the way.

[vanguard]

I'll stick with Intel. Their greater market share leads to greater testing (according to me). That's worth a few extra dollars to me. To be honest, if I ran into problems with my CPU I wouldn't be able to pinpoint it down to the CPU anyway. I'll just go with something that's "mostly" bug free. Also, I'm not convinced that an AMD dominated world would be better than an Intel dominated world anyway.

Re:Not For Me...AMD all the way.

[jawad]

Who's to say either has to dominate?

Re:Not For Me...AMD all the way.

[arivanov]

underdog?!?!?

Underdog, with these benchmarks?

Actually I find the article extremely funny. They explain how good and advanced the new CPU is and every time they quote a bench, errr, why is this CPU beaten up by Athlon?

In btw, I stick with AMD as well. Actually, I have done it for quite a while. At the same price always beats Intel by as much as 30% under linux or BSD for typical server duties. And it usually has better power saving mode which for a permanently working machine is a definite plus...

Re:Not For Me...AMD all the way.

[seaportcasino]

But the Athlon motherboards do tend to cost twice as much as their intel counterparts, although I'm not sure if that's true with the new intel boards. But then who'd want to pay $800 for a 128meg RDRAM module anyway. Sales of those new intel boards must be down, way down!

Re:Not For Me...AMD all the way.

[seaportcasino]

But the Athlon motherboards do tend to cost twice as much as their intel counterparts, although I'm not sure if that's true with the new intel boards. But then who'd want to pay $800 for a 128meg RDRAM module anyway. Sales of those new intel boards must be down, way down!

Re:Not For Me...AMD all the way.

[seaportcasino]

Sorry bout the double post. I thought Slashdot was hung. It is running awfully slow today.

Re:Not For Me...AMD all the way.

[Peale]

Who moderated me down as #^(&()% FLAMEBAIT? This isn't flame, moron, it's OPINION. However, by using the word 'moron' in this post, THIS is flamebait.

I'm sorry. What I meant to say was 'please excuse me.'
what came out of my mouth was 'Move or I'll kill you!'

PIII500

[nospoon]

Unfortunately I bought a couple of the older PIII 500 before the coppermine - in the Slot1 style. They both have been overclocked to 560mhz but I can't get them to post at anything higher - the next FSB setting on my particular MB is 124mhz I think. I look forward to getting one of these PIII500E's that can handle 133mhz+ FSB with a multiplier of 5 - looks like a good bargain for the money.

Celeron first onboard L2 Cache??

[heimy]

The second generation Intel Celeron design incorporated the L2 cache directly into the CPU core itself, and was the first Intel core to feature an on-die L2 cache design Someone correct me if I'm wrong, but was the Pentium Pro the first Intel core with an on-die l2 cache?

Re:Celeron first onboard L2 Cache??

[Anonymous Coward]

Someone correct me if I'm wrong
You are corrected, then.

Re:Celeron first onboard L2 Cache??

[Nick Mitchell]

The Pentium Pro had an on-chip, but separate die, L2 cache. The Ppro "chip" actually had two dies. Still ran at full speed, though. Being off-die, they had to limit the line size to 32 bytes.

In contrast, the Celeron, Alpha, R10k, and I guess these new P3e's have on-die L2 caches.

Re:Celeron first onboard L2 Cache??

[Defiler]

The PPro had the L2 cache integrated into the CPU package, but it wasn't actually etched onto the same wafer as the CPU core. All of the disadvantages, but none of the advantages.

Re:Celeron first onboard L2 Cache??

[barleyguy]

Well, it did have one advantage - an onboard cache controller, which allowed it to cache large amounts of RAM, about 512 Megabytes. Most of the other Intel chipsets of that era (FX, VX, TX) would only cache 64 megabytes, which left people wondering why adding more RAM slowed down their machine.

I think the biggest advantage of these new chips is the .18 process they are using, which is what leads to the greater overclocking potential. Most of the .18 aluminummines will do 800 Mhz, where the physical limitation of the .25's was about 590. Still, I'll never own one, because I'm going to have an Athlon...

Re:Celeron first onboard L2 Cache??

[billybob jr]

FWIW, the HX chipset which was very popular cached up to 512 megabytes I believe. I think the HX chipset was pretty popular. It's always frustrating when a corporation releases a supposedly more advanced solution (TX chipset) that is crippled in certain areas (64 M of cacheable memory)

Re:Celeron first onboard L2 Cache??

[Defiler]

You're right.. I'd forgotten about the TAG RAM controller.. Aah, memories! ;)

Re:Celeron first onboard L2 Cache??

[barleyguy]

You're right about the HX. It would do 512 Megabytes if you had a TAG RAM chip on your motherboard. However, it was strictly limited to 66 Mhz or less bus speed.

Re:Celeron first onboard L2 Cache??

[billybob jr]

Actually I believe many of the popular overclocking boards were based on the HX chipset. Of course you probably meant official clock speeds.

Re:Celeron first onboard L2 Cache??

[snowbird]

Yep, sure did, along with ECC support too. Funny, but Intel's own Tucson Mboard w/HX chipset only cache'ed 64MB. Too bad I didn't know enough at the time I bought mine that they were being cheap to the tune of a few bucks :(.

It'll be a long time and a very close look at the fine print before I'll buy another Intel MB.

Is overclocking really that important?

[Poe]

I have spent some time trying to think of a use for overclocking, and I can't. I seriously question the sanity of anyone who overclocks a production system. The speed of modern games seems more governed by the graphics card than the processor. (So I guess I understand overclocking your graphics card.) I can understand overclocking your chip for the sake of cool. But I can't imagine any good use for that kind of speed that couldn't be accomplished with SMP.

Re:Is overclocking really that important?

[nospoon]

1 word - Money - you can save money and get a faster system by overclocking.

Re:Is overclocking really that important?

[Mr. Flibble]

I am currently running a Celeron 300A @ 450 Mhz.
I can certanly testify that the speed increase in games (and kernel compiling!) is signifigant. I have not benchmarked it I admit, but I can certanly feel the difference when it is at 300 vs. 450.

SMP won't help with Quake 3 Arena (It will under NT, Linux SMP for Q3A is not yet avalible) Or Half-Life, or many other games.

Half-Life for example is a very processor-dependant game. Many games currently are. Although graphics card technology is growing rapidly, few current games fully exploit the cards abilites. So, in short a high clock speed is VERY important.

SMP might help if you are running BeOS, as it can force multi-threading, but the current state of gameing does not support SMP, hence there will be no bonuses associated with it.

Having said that, My next system will be a Dual Celeron 366 PPGA on a Abit BP6 motherboard. I will clock the chips to 500 (possibly 550 with water cooling, just because its geeky). I am doing this because:

1) Linux will FLY on that system.

2) Quake 3 Arena should have Linux SMP by the time I get the Motherboard.

I personally am not all that interested in overclocking my Graphics Card, I dont think that the yield in framerates is enough. The CPU is a different story IMO.

Re:Is overclocking really that important?

[Buaku]

I doubt many people do overclock production servers. I would also doubt the sanity of anyone doing that. However, why do you assume that production servers are where the overclocking is taking place?

Overclocking makes perfect sense for people who need maximum power, or need more power than they can afford. People using their computers for games are good candidates for overclocking. If their chip is old enough to have trouble with newer games, they can overclock it and delay the time when they have to buy a more powerful chip. Other people overclock because they know they can buy chips that are a 100 MHz slower than the top of the line for very cheap and overclock it faster than the top chips, saving them a lot of dough. If it burns out too soon and they have to buy another, they're no worse off than if they had bought the expensive one anyway. Other gamers may have enough cash, but they want the most powerful machine possible, so they buy the best and then overclock it. They don't care if the chip will burn out in six months since they are planning on upgrading in six months anyway.

Re:Is overclocking really that important?

[Anonymous Coward]

I can't imagine any good use for that kind of speed that couldn't be accomplished with SMP.

That's the point; not everyone wants to buy a new mobo and processor to get higher speeds. And remember that SMP is useless if your application can't utilize it. Gamers don't want SMP. Thank you for making a fool of yourself.

Re:Is overclocking really that important?

[Kvasio]

Well ... for all those folks running dnet & seti@home clients it is REALLY important. How otherwise you would get faster ? By overclocking beloved celeron 300a to 450mhz the rc5 keyrate rises from 890kkeys/s to 1260 kkeys/s. That makes the difference. And price of the celeron 450 was more than twice higher.

Re:Is overclocking really that important?

[Tackhead]

> I seriously question the sanity of anyone who overclocks a production system.

I'd echo that - but with one very important qualifier that completely inverts the intent of your original post, namely "without knowing what they're doing".

The rules for successful overclocking are actually pretty simple - know the relationship between FSB speed and the multipliers and dividers that turn that FSB speed into AGP and PCI speeds, and how that relationship varies as a function of motherboard vintage.

Would I run a C366 (5.5 x 66 MHz) at 75 MHz FSB, to get 412.5 MHz? Hell no:

• AGP: 2/3 of 75 is 50 MHz, and I'm (wastefully) underclocking the AGP bus, or if it runs at 1 * 75 is 75 MHz, I'm (far-beyond-dangerously!) overclocking my AGP bus.
• PCI: 1/2 of 75 is 37.5 MHz, and I'm overclocking my PCI bus by maybe 10%, which most hard drives can handle, but perhaps not all.

That "hell no" argument goes double if you ask me to run it at 83 MHz FSB * 5.5 for 458 MHz.

But if I crank that puppy up to 100 MHz FSB, I have - modulo heat - rock-solid stability:

• AGP: The mobo should detect 100 MHz FSB and switch to 2/3 of 100 MHz, or 66 MHz. You're running at spec.
• PCI: The mobo should detect 100 MHz FSB and switch to 1/3 of 100 MHz, or 33 MHz. You're running at spec.

That's right - as long as it can be cooled effectively, a C366 running at 550 MHz is more stable than one running at 412 or 458.

A similar set of calculations can be performed for other CPUs running at other core multiplier frequencies, and will reveal different "sweet spots" where the PCI and AGP busses run in spec. (In an ironic twist, buying a "faster" CPU of the PII and Celeron vintages actually makes things worse :)

So would I allow some random kiddie who says "0vercl0cking iz k00l 4 gamez, d00d!" crank a production system of unknown processor vintage and motherboard capability to "as fast as you think you can get it, kid"? No, that's madness. He might have a CPU that'll run fine at 550 MHz at 2.1V, but not at 2.0V. Instead of turning the voltage to 2.1V and testing heat dissipation, he leaves it at 2.0V and "steps back" to run the FSB at 83. Two weeks later, the hard drive explodes and everyone blames overclocking. It wasn't the fault of overclocking, it was the fault of one particular overclocker.

But would I, if starved for speed (and cash - in that I couldn't afford to buy the speed off the shelf), allow an overcautious doomsayer dissuade me from applying overclocking in such a way as to get speed without sacrificing stability? No again - that would be just as insane.

Like everything else in computing - blind adoption and blind rejection is madness. A clued admin will realize that there's a middle ground.

If an admin has the power to select quality equipment, and the clue to tweak said equipment without risking stability, the right answer to "Admin, we want a fast server but are on a low budget" is "yes, I can do that, but only if you let me select every component and not quibble if the motherboard I choose costs $10 more than the cheapest one on the market."

Re:Is overclocking really that important?

[Anonymous Coward]

"SMP might help if you are running BeOS, as it can force multi-threading"

What a triumph of marketing over reality! Running BeOS on a 4 CPU machine lead to little to no usage of anything but the first two CPUs, under normal usage. It took pathological cases (2 povrays, gcc and netpositive) to utilize all CPUs.

Don't believe everything you read.

Re:Is overclocking really that important?

[Anonymous Coward]

good info in your post for hobbyists, but you're quite wrong. If you're an admin with more than ~4 machines, overclocking will not save you enough money to justify the extra time you WILL spend (its just a matter of time) replacing failed components. The overclockers will argue, but no real sysadmin (that is, one with an actual job) will.

For a gamer, you'd almost be stupid NOT to overclock. For an admin, you'd CERTAINLY be stupid to overclock.

Re:Is overclocking really that important?

[WNight]

You say it WILL fail... Well, everything will fail eventually. But I've had OCed systems that have run just as long as non OCed systems.

As the poster you replied to said, increasing a bus from 66 to 100 will almost always (and it's easy to check) simple increase the PCI and AGP divisors, and you'll end up with a system where all the components except the CPU are running at rated speed.

So, you can go from 366 -> 550 in your Celeron, and the only part that is running above spec is the CPU. And CPUs all came out of the same batch anyways, meaning you've probably got a CPU from the same set as others that ended up marked at 450s and 500s, so it's probably barely above spec.

So, if anything fails, it'll be the CPU, because it's the only thing that'll be running fast, and CPUs are some of the most robust parts. With a proper fan you'll get five or six years, easily, of 24/7 usage from a good CPU. If OCing it drop that a year, who cares? It'll still be many times longer than that system will end up in production usage.

Cons: Perhaps shorter CPU life. (unproven, with good cooling)

Pros: Cheaper, faster.

With the savings from buying a 366 instead of a 550, you could probably buy 256MB of ram, definately 128... That would boost the performance even more.

In a cost/benefit scenario. Your wages - at least $150 a day... The CPU price difference between what you buy, and what it would cost to buy what you OC it to - probably $100 - $200... So you're spending at most 30 minutes to overclock, toss a decent fan on, and test it. The math works out strongly in favour of doing the overclocking.

can we say.... athlon?

[frknfrk]

Is it just me, or did the Athlon outperform everything on every single test. In most cases it was around 10 percent better or so. While the K6 was certainly not 'quite there', the Athlon is really pushing Intel to develop cooler technology, which in turn pushes AMD, etc, etc.

Which, of course, leads to better, cheaper technology for all of us.

Gotta love competition. If only a certain OS had competition all through the 80s like Intel is getting from AMD now.

We might have had free UNIX quite a few years earlier...

Re:can we say.... athlon?

[Anonymous Coward]

Did you notice how the benchmarks were performed?

Please note that while the i810E is not Intel's most robust chipset, its support for 66/100/133 bus speeds makes it very easy to compare the various Intel processors under a consistent test bed.

and

Athlon benchmarks have been included for reference only, since the Gigabyte GA-71X motherboard utilizes a higher performance chipset...

While Athalon is a good processor, it is not that much better. The chipsets can easily account for such a difference.

Re:can we say.... athlon?

[frknfrk]

I never read the fine print, my mistake.

FSB Overclocking

[Defiler]

Ever since the 486/66, I've been a straight Intel user.. The K6 was completely disappointing, as was the K6-2. However, I recently dropped my P3 and switched over to Athlon. Wow.. It's great to be able to overclock again without altering the PCI/FSB clock speeds.. I've got my Athlon 500 running at 800MHz, and it smokes. Some of these new Coppermines show good overclocking promise, but your shiny new AGP card is unlikely to work when you have your FSB up there at 150MHz.. Most of these Coppermine reviews are using unlocked engineering samples.. Nice benchmarks, but in the real world, you'll be setting your sound card on fire if you try them.

Re:FSB Overclocking

[J4]

I've got my Athlon 500 running at 800MHz, and it smokes.

I'm sure it does

Re:FSB Overclocking

[Xamot]

Correct me if I'm wrong, but does the locking on the Intel chips have anything to do with the speeds of your PCI or AGP buses? They run off speeds based on the FSB. The multipliers for them are located on the chipset(BX/i820/etc) not the CPU. And these multipliers are not locked. Check out the new VIA Apollo Pro Plus 133 it allows you to up your FSB to 133 and multiply down your AGP bus to within spec (something not possible on the BX chipset).

--

Yes it is.

[pulski]

Here's a thought. SMP costs money. Overclocking does not. It's as simple as that. If you want more speed and you don't feel like paying for 50 extra MHz, overclocking is the way to go. Not all of us have a lot of money to spend on our computers.

-----

Motherboards...

[Fiore2]

The Athlon is a GREAT chip.
The problem is the fact that the motherboard manufacturers are scared to manufacture the Athlon boards. They are scared at what Intel will do to them.
Currently there are 5 motherboards:
Asus, Biostar, FIC, Gigabyte, and MSI
Some of these companies won't even mention these boards exist!
You can find all the info at www.tomshardware.com

Re:Motherboards...

[barleyguy]

If you go to www.slota.com/motherboards/ they list 45 different slota motherboard models. The 5 you mention are just the mainstream OEM boards, there are a lot more.

However, I've seen 4 out of the 5 you mention (Asus, FIC, Gigabyte, and MSI), and my favorite is the Gigabyte. It has onboard temperature and voltage sensors, is incredibly stable, and doesn't seem to need a 300 watt power supply. It also hasn't had the compatibility problems the FIC and Asus have. (The FIC had a voltage regulation bug that caused random reboots on their early boards, and the Asus has had problems with Ultra class video cards. There are other issues as well...)

I agree that the Athlon is a great chip. But most of the motherboards (with the exception of the Gigabyte, which I think is awesome) suck. I hope this gets worked out, so there can be some fair competition for a change.

Re:Motherboards...

[bdowne01]

This is all too true. Try to find information from Asus on the K7M. Good luck.

Re:Motherboards...

[Anonymous Coward]

Hmmm Explain why AMD is actively SCREWING ABIT? Is it because Intel if forcing AMD to screw one company that WANTS to build a MB? AMD is in Microsoft pocket. We all know it. The screwing of Abit by AMD is directly related to the dual Celeron board being aimed at Linux users. Screw AMD and Microsoft both.

Just another way to solve the same problem

[handorf]

Personally, I make use of both. With the wondrous Abit BP6 motherboard, I've got SMP Celeron 366s OCed to 458 (I'm Chicken). Since I was able to buy the cheaper CPUs I could justify getting an SMP board, which does wonders for my performance under Linux.

Also, as another poster pointed out, many games don't support SMP, so you're stuck with core speed as your way to go faster.

It's certianly a risk, but to each their own, eh? I could probably push my system to 550, if I wanted to, but I'm happy with my price/perfomance curve at the moment.

Good thing.

[Emphyrio]

It's a good thing the pentium III finally came out in fpga form factor - slot1 boards are simply too expensive, and not handy enough.

It smokes?

[Anonymous Coward]

I've got my Athlon 500 running at 800MHz, and it smokes.

Uh, if it's smoking, maybe you're overclocking it a bit too much? ;-)

RDRAM prices!?!

[seaportcasino]

What is intel thinking trying to sell us RAM at these prices? I could buy a whole eMachine, including monitor and everything for the price of one 128meg stick of RDRAM. Is anybody actually buying these systems and paying these prices? I imagine some must, probably the big corporations that believe nobody ever lost their job buying the latest intel. Does anybody know when the prices for this ram might be going down? Or even better if intel has any plans to support the relatively cheap new DDR ram? I am really dismayed that intel thinks they can gouge us this way. We need to stand up and say "I'm mad as hell and I'm not going to take it any more!" and then go buy a nice new Athlon.

Re:RDRAM prices!?!

[eomir]

Of course RDRAM is going to be expensive. It is pretty new on the market, and they've got to pay off their research and development costs. After that RDRAM prices will drop. Also RDRAM is a lot faster than SDRAM so you are gonna pay more for the performance increase. That is just the way it goes.

Re:RDRAM prices!?!

[Anonymous Coward]

-=[Slightly off topic with RDRam]=- Actually, RDRam just has a higher bandwidth than does SDRam. The latency is quite a bit more on the RDRam...It can chug alot of information, just slower gettin it to ya.

Gettin' too hot

[Skinka]

Remember the days when CPUs didn't have to be cooled at all? No noisy fans, not even heatsinks, just the bare naked chip. Who wants to bet that the Itanium will be flip chip with a built-in dual layer peltier sandwich?

Overclocking=rip-off?

[jued0001]

As much as I understand wanting to push a processor to it's limits, let's think about this from Intel's point of view. They *KNOW* someone's going to overclock it. With this new XXXE chipset, just like the first Celerons, they knew they could give these chips to the public at whatever speed they desired. If these were coming out in 6 months, do you think these same chips would be running at 500 or 550? No way, Intel would be forced to sell them at near the peak of the their performance-from what I've seen, almost 800mghtz. If you're going to buy one of these, wait for a month or two, the price will be significantly lower and you'll still be able to overclock the piss out of it. If you want something to overclock now, go to Pricewatch and buy a 300-400 Mghtz Celeron for well under $100. I have yet to find a game that won't run just fine on my 400 Mghtz non-overclocked Celeron system.

________________________

Mello like the Yello, but without the fizz.

What the heck?

[H3lldr0p]

What does Intel think they're doing here?
A release of yet another chip based upon their out-dated and out-preformed PPro core? Who are they kidding?
It's not as if the new design actually came close to touching what AMD has in the Athlon, nor does it prove that this revamp is actually worth that much to an everyday user. Okay, some of what they did to the chip to speed things up are cool, but IMHO, these are ideas that should have been floating around for some time now. The pushing of the overclocking is cool, but again, this proves little to nothing when comparing it to anything outside of Intel chips.

Re:What the heck?

[Anonymous Coward]

Uh actually... it comes *very* close to beating the K7. (when SSE is used it beats it)

Re:What the heck?

[H3lldr0p]

This is just adding to my comment above.

Over at ABCNews (beware, its an article from CNet news, and will spawn some java your way) they're talking about Intel some more and their current "accelerated" roadmap for 2000. The highlights being that there will be a 1 GHz by Q4. Does this make no sense whatsoever to anybody other than me? AMD has a chip out that is a generation ahead of the PIII, and they're preparing to simply up the speed-rating? And not soon, but several months from now!? Admitily, the article does mention that Intel wants to "diversify into other markets". But it sounds like Intel wants to loose now.

Re:What the heck?

[Anonymous Coward]

Since this is thier LOW end chip. AKA Celeron's big brother I'm not sure what your point is. Notice they tested the chip with some crappy MB and the AMD chip still only beat the low end chip by around 10-15%. You think with a quality MB it might be closer? How about the fact it will cost 1/2 the price of the Athlon? I'm not letting AMD gouge me 50% for MAYBE a 5% speed increase.

But it does cost money.

[Poe]

You are going to pay for your overclocking in cooling equipment or reduced processor life. A hotter processor simply doesn't last as long.

Re:But it does cost money.

[Troed]

A Celeron 300A at 450MHz lasts just as long as a Celeron 450A would (so, let's take the 466 as an example instead since there's no 450)

You relly need to learn a thing or few about CPU manufactoring before trying to lecture the ones that do.

Read the article. The die in a 550E is the same as in a faster model. The reason they're stamped with a lower MHz rating is pure marketing.

Re:But it does cost money.

[pulski]

You can get away with overclocking many processors without any additional cooling equipment. Also, if you are the kind of person that is going to be overclocking your system, you are probably also going to be the one to replace your system after a relatively short period of time. If you are just going to overclock a little to save some money, you aren't going to have your CPU die on you before you're done with it.

-----

Re:But it does cost money.

[Anonymous Coward]

I guess you could call it "pure marketing" since Quality Assurance (at least when it's paraded all over the place) is marketing.

What chip vendors do is lot samples of the chips produced in each production run. Some runs, the yield isn't as good as other runs, so the chips don't run reliably at the higher clock speed.

The really bad chips, of course, get thrown in a dumpster, out of which the scrap dealer pulls them. From the scrap dealer they go to a remarking facility. They end up for sale at Ma and Pa computer stores. This has been documented more than once.

Re:But it does cost money.

[Anonymous Coward]

So you spend $20 for a heatsink/fan combo instead of $10. The CPU *might* die sooner, but if you're an overclocker you're after speed RIGHT NOW, by the time that CPU dies you're on to the next CPU and you're overclocking it too.

Overclocking is great for the home user, just don't bet your business or your job on it. 1 more GPF per week means squat in a game machine, but your business could go under if it is crashing all of the time

Are the dual processor boards still scheduled?

[Greyfox]

I'm going to be keen on picking up a dual processor Athlon motherboard and another Athlon 600 processor in the next few months. Are we still looking at 1Q or 2Q 2000 for the dual processor Athlon boards?

Re:Are the dual processor boards still scheduled?

[barleyguy]

According to JC's PC News and Links. (www.jc-news.com/pc), there should be dual Athlon boards out in Q1, probably January of February. Tyan is supposed to be the first to manufacture one.

Re:Are the dual processor boards still scheduled?

[Ozric]

This is what I have been waiting for. I always
buy dual only now days. I have a dual PPRO
180 oc'ed to 200 and I now have a duel 300A oc'ed to 465@112mhz FSB. SMP is for me, and Intel has had a lock on me because of it.

It costs less. Do the math.

[Mr. Flibble]

Current cost of a Celeron 366 (Overclockable to 500, often to 550):

$37 US

Current cost of a P-III 500:

$227 US

Lets do the math:

227/37 = 6.1

That means that I can buy six celeron 366's for the price of one P-III 500. So, if I blow one processor a year, it will take me 6 years (at the current cost) to work up to the current cost of a P-III 500.

Despite the nice current speed of a 500, I doubt I will be using a 500 six years from now.

This is why my Celeron 300A is clocked to 450, if the processor blows (and if you see the cooling I have, I doubt it will blow) I don't care I buy another chip.

This coupled with the Celeron's core-speed cache, the more you clock it the better it performs. My celery performs faster than a Pentium-2 450, and about 5% slower than a P-III 500 because of its cache speed (benchmarks like photoshop excluded).

So, given all this, which would you choose?

I chose price and performance, I have no use for just price.

Re:It costs less. Do the math.

[Anonymous Coward]

I guess it doesn't matter to you what happens to the rest of your system when the processor goes borderline. If it corrupts your filesystem, or is marginal for a week and corrupts all binaries you build during that time, or sends out bad packets that take out the rest of the network, you can always plug in one of the other five chips you budgeted.

Right?

Re:It costs less. Do the math.

[Mr. Flibble]

When I was pushing the current chip I have way too far (just to see what would happen) I was getting build errors when trying to recompile my kernel.

I run tests to check that the system IS stable, I would never suggest that an enterprise level server use overclocked chips. The cost vs. performance is just not worth it. What many people fail to realize is that many of the chips released are EXACTLY the same as other speed chips. They are only marked "down one bin" or "up one bin". So there is some margin for error. I have had a 166 that was running at 200 for quite some time (just gave it to a friend0 and I never ran into ANY hardware or software problems on the system.

At my current FSB speed my Voodoo 1 (pity me) flakes out sometimes. But I am running a gaming machine, I do no critical builds on it.

Thats the point I guess, I have not had any problems as a result of overclocking - but my framerates in Quake have helped me frag a little more! Over the years that I have been overclocking systems I keep haveing people telling me that I am destroying my system etc. etc. etc. If you don't think that the risk is worth it, then by all means don't do it.

Maybe It's me...

[KaosDG]

But I can't really see why we squabble over SMP > OC'ing, or OC'ing > SMP...

In my work, OverClocking can't be done... I run production servers and i can't take the risks of having a board crap out on me because i needed an extra 50mhz... Even on workstations, where I would see a benefit in OC (3ds max and After Effects) i have gone SMP Xeon, because:
1) I didn't have to pay for it (HUGE plus)
2) It's more reliable (i've tried the OC with a test workstation and Adobe AE has GPF'd more than I can tolerate on a windows platform)

But at home, I'd love that Refigerated 1Ghz Athlon for UT/Q3a... hell, even solitaire.
My home pc isn't that important to me at this point.

It all depends on your needs...

Most people knew that Intel Released lower clock chips, it's nothing new. I remember being able to overclock my 386. OOoh... 40Mhz.
I remember reading that it was because they tried to make the chips run as fast as possible, but if it had too many problems at that speed, it would be lowered and tested again... (or something like that)
So apparently they would work OC'd, but with more "faults."

Quality Assurance I suppose?

Oh yeah, we have 2 of the Xeon monsters, and 8 SGi O2's... which in this case the O2's are actually cheaper than a Single processor Xeon and run better... Plus they look really cool.

*Blech* biege PeeCees suk.

Re:Maybe It's me...

[Anonymous Coward]

Sure overclocking is not for everyone but to buy a $300 P3-500e and run it at 700mhz or so is a pretty good bargin.

Re:Maybe It's me...

[Anonymous Coward]

The thing is eventually you get to the point that all the chips you make are good enough to run at the highest speed. Now what do you do for lower speed chips? You could setup a special line to produce crappy chips or you could take the good chips and label them slower.

OC your desktop - but NOT a production machine

[in8]

No one in their right mind should ever overclock(OC) a production machine. It's not worth the added risk and uncertainty.

However, personnal workstations/PC's - well that's a different story. It's always nice to be able to improve your price/performance - even if you're not playing the latest and greatest games. Looks like you can get a 33% increase OC'ing the PIII 500E - not a bad deal if you got PC133 SDRAM. (ps - RAM quality is a big factor in successful OC'ing.)

Re: OS competition.

[Anonymous Coward]

Hey, nerdboy. There _was_, contrary to your twisted view of reality, plenty of competition to Microsoft in the 80's. They just all sucked at the consumer level, and people _chose_ Microsoft. Similarly, there is unarguably competition right now to Microsoft. It's a freaking tautology.

Why is this a problem?

[adamsc]

I bought an FIC SD-11. It's fast, had every feature I wanted, stable & they had tons in stock at a good price. The other boards seem to comparable.

My question is just why this is a problem. You've got a selection of good boards at good prices, all of which are in stock. Even if there are only a few companies making them, why would this prevent you from buying an Athlon? Isn't this really a good thing in that it gives all of the manufacturers better volume than the more contested Intel market, thereby increasing the benefit for standing up to Intel?

Re:Why is this a problem?

[Anonymous Coward]

Perhaps if there were better mobo manufacturer support and demand the chipsets would have been available to fully exploit the Athlon at 133 and AGP 4x.

Rambus Performance - on the next In Search Of!

[adamsc]

Also RDRAM is a lot faster than SDRAM so you are gonna pay more for the performance increase.

For values of "a lot faster" approaching 1-2%, yes. None of the benchmarks I've seen for RDRAM are that impressive, certainly not "pay 5 times as much" impressive...

Next page

[Anonymous Coward]

I just hate "Next page" buttons.

BANG! cheap horsepower ...just like chevy350

[Cmdr. John Koenig]

Let's see a Celly366 goes for 40 bucks (I'll take two)+ BP6...140 bucks...hmm! = easy 550x2 with oem fans

W2K/SMP takes advantage of both

not to mention...It's just seat of your pants fun! Like hotroddin' use to be

enjoy life, eat out more often

Re: OS competition.

[Cmdr. John Koenig]

Remember TRS-dos? (trashdos), L-dos, Newdos-80, AppleDos, CPM, PC-dos(IBM says...woops!), Atari dos...oh yah unix and aix?

Long live scarfman!

Fanless Freak!

[Cmdr. John Koenig]

Yah, but those diskdrives sure were much louder back then, whinin' and a whistlin'...And before Hard drives in PCs, those floppies were constantly going...cartridges by the pile.

Go 550MHz dude

[Cmdr. John Koenig]

I was worried A BIT myself about 55O before going for it. I've read a lot about fans and cooling etc. Frightening me etc.
After overclocking 2.5 months with cheap AOC fans, I'm not at all frightened anymore. I've read somewhere that the sink size itself is more important than the fan. AOC is pretty big. A little Radio shack heatsink goo. Had to bump core voltage to 2.10 to get Q2 and MP2 Playback to run. Other than that I'm doing cartwheels.

My buddy was impressed enough to have me build him an identical machine. I got two more 366's from a completely different source (access micro) and they overclocked as my first set, with no surprizes or locks.

BeOS has some nice SMP/multithread benchmarks and demos to burn both chips in...I don't do command line..gave it up for Lent back in '91.

as the little voice said to charlie m. "do it"

Re:Go 550MHz dude

[handorf]

Kewl. I just may try it after I pick up some more beefy heatsinks (got the standard Intels at the moment).

On another topic, do you have the BP6? If so, do you know if it will handle the coppermine chips discussed in this article? I meant to check the voltage range last night, but I fell asleep.

Re:Go 550MHz dude

[Cmdr. John Koenig]

Indeed,

I have the BP6.

point your browser at:
http://www.bp6.com/

in answer to your question; NO abit says it can't do the FC boogy-woogy. Bummer!

at first folks thought maybe a bios upgrade and an adapter board, but abit has responded with a "NoGO...sorry!" I'm sure they'll have a kickass board by the time I'll want to upgrade (7-8mos)

go to bp6.com for more info.

On another note, I just remembered I had to switch the AGPCLK/CPUCLK ratio to 2/3 (ie. 66Mhz) for my ATI AIW 128 32mb to work properly. May not be critical to other new boards. MPEG2 playback and 3D games would lock the system up otherwise. You know...I might just try 1:1 again, I can't remember if I had bumped the cpu(s) core voltage to 2.10 before or after trying 1:1. I mean logically, it should run fine at 1:1 (100MHz) since PII and PIII mobos agp are natively running @100.

On still another note, abit has a last minute flash bios update out, again available at bp6.com; says it's supposed to fix something y2k related...who knows? bp6.com has links to easytofollow instructions on flashing; basically make a bootable disk with upgrade, boot from disk, run a .bat file, reset bios and enjoy.

Re:Go 550MHz dude

[handorf]

3117! Thanks for the link. I'll try and flash the BIOS tonight, but unless the power goes, the system will be up anyway, so I'm not too concerned.

Bummer about the FCs! Oh well. Hopefully they'll aim another board at the same cheap SMP market segment. Unless they foobar BADLY, they've got a customer for life here!

Smoking not good

[Anonymous Coward]

Last time I checked smoking was a bad thing for processors. It usually means that you did something wrong. You may want to look into that.

AMD gouge of 50% $ increase? NO!

[hodags]

Actually the current street value of an Athlon 550 is only about $250USD where as the PIII 550E is marked at $350USD. That when looking at comparitable clock speed pricing. But also note, the comparison on sharkey was to an Athlon 600, and even this case Athlon 600's are running at about $375USD street price. So, you're only paying ABOUT $5USD more for that 10-15% speed increase... as well as getting a better chip and (as stated earlier by others) "overclockability" to upwards of 800 with an Athlon 600... sounds rather worth it to me ;^)

Re: OS competition.

[frknfrk]

I said that I wished MS had competition in the 80s similar to the competition Intel has from AMD now. the competition MS faced in the 80s was nothing like the current Intel vs. AMD wars. You said yourself that all of the MS 'competition' 'sucked at the consumer level'. Well, I was saying that AMD does NOT suck at the consumer level, and if OS/2 had not sucked at the consumer level, perhaps windows and OS/2 would both be around, and both be much better. Buf, of course, who knows. at least I don't go around calling people nerdboy :)