Intel "Northwood" vs. Athlon XP 2000+

Augustus writes: "LinuxHardware.org has just published their results in the Pentium 4 verses Athlon XP war. In this review, the new Pentium 4 'Northwood' 2.2GHz is pitted against the Athlon XP 2000+. To level the playing field, both platforms use DDR memory which make for some interesting results."

Comments

I have one, and..

[Evanrude]

I am running Linux on Dual Athlon XPs. Does that make it Linux XP?

Re:I have one, and..

[jd]

I take it you're running the XP's on an MP chipset. You do know that you're being very naughty, by not buying AMD's hyper-expensive, ultra-slow MP processors, and having a decent computer to use instead?

(I asked AMD if such a configuration was possible. I got a LONG e-mail back, which said that it wasn't recommended, documented, approved, etc'ed, to run a mere (read: cheaper) desktop processor with a multi-processor chipset.)

Re:I have one, and..

[Toraz Chryx]

uh, the AlthonMP is _one_ speed grade (67Mhz) behind the top of the line XP

that's hardly 'ultra slow'

Re:I have one, and..

[Evanrude]

Actually, you are right. However, Tyan now manufactures a motherboard that is "approved" (not necessariyl by AMD) to work with Athlon XP Processors (not just XPs masquerading as MPs).
The board is the Tyan Tiger MPX (S2466)

Re:I have one, and..

[ryusen]

this is the same thing like intel not wanting to certify that even their celerons can go smp on those abit boards?

Re:I have one, and..

[Emil Brink]

Really? It doesn't say so on the page you linked to, not that I could see anyway. It did list the XP as being supported for the board, but also said "Single only!" right next to it. Too bad, I almost got my hopes up there. ;^)

Re:I have one, and..

[Alan Partridge]

overclockers, don't forget that the Athlon MP is "unlocked"...

Need better testing

[Triode]

We need better testing. You can not (really)
expect the same test to give you a good judgement, and then on the other hand, having different test for each CPU would obviously not give you a good judgement... Hmm, perhaps some speed testing on regular apps. I know they do that on quake and the like, but then that was written for a specific architecture also. Just a thought.

Re:Need better testing

[Triode]

Even then, I think it is more of a "what you are after" situation. For instance, the chipset itself can dramatically change the performance. As well as ram, etc, etc... so the idea is are you benching the CPUs against each other or the CPU/chipset/RAM system against each other? Even then the question is "how do I verify that the results are not skewed based on the testing methodoligy?"

Re:Need better testing

[Chris Burke]

The XP is the same mask as the MP. The only difference is the AMD MP Seal of Approval(tm).

Just pointing that out.

boom!

[feed_me_cereal]

and 5 seconds later, linuxhardware.com gets slashdotted. Does anyone know of another place I can find a simmilar conparison? I'm interested to know how it plays out.

/.'ed already

[ZaMoose]

Perhaps they should have used the test machines to host their website...

Re:/.'ed already

[Tower]

The last /.ing of LinuxHardware was due to provider bandwidth rather than server constraint. I'm guessing this is the same issue.

Re:/.'ed already

[marcop]

Well, maybe they ditched all the load balancing and extra hardware and placed only these machines as the webserver. It's a new benchmark: how long does a system last a nice /.'ing?

Re:/.'ed already

[harakh]

Yeah well not everyone has fat pipes to the Internet I guess - its pretty expensive...

Why would it be tough to be bilingual when you get hit in the head btw? :)

Re:/.'ed already

[Progoth]

I could be mixed up, but isn't linuxhardware run by a georgia tech student with a cable modem in his apartment? and doesn't it get trashed every time it's linked on ./?

Why DDR on P4?

[kawaichan]

P4 was designed with Rambus ram in mind.

They should really use Intel's i850 motherboard to pit against the Athlon.

The p4 platform is simply not designed for DDR in mind, adding DDR in the i84x boards are afterthoughts and IMO I would much rather use Intel boards with Intel processors.

Athlon is doing quite well right now, seems like there might be a delay for the .13 Athlons, hopefully it will work out ok.

Re:Why DDR on P4?

[ergo98]

For these comparisons to really be valid they should base them on price : i.e. "We have $2000 to spend on each platform", and if the Athlon gets to use an 8MB cache HD because of the money saved on the RAM, well then so be it. Most people do vary their options based upon the price, so it does seem to be the most pertinent factor.

Re:Why DDR on P4?

[aminorex]

Because nobody with an ethical bone in their body
would buy RAMBUS.

Nor, for that matter would anyone who cared about
getting a fair value for their dollar.

Re:Why DDR on P4?

[liquidsin]

But the test isn't about biggest bang for your buck...AMD wins that test hands down every time. The test is about comparing performance. If we had to do it by dollars, we'd be running the Athlon 2000 XP against a P4 1.9Ghz (according to pricewatch).

Re:Why DDR on P4?

[/ASCII]

Sounds overly complicated (Read: stupid). First of all price on identical products vary so it would be a bit arbitrary, and besides that, it would open up to the question of which piece of hardware is the culprit.

By using nearly identical hardware, one can be sure of the performance impact of that piece of hardware. Afterwards, you can compare prices, and try figureing out which processor fits your wallet best.

Re:Why DDR on P4?

[kenneth_martens]

For these comparisons to really be valid they should base them on price : i.e. "We have $2000 to spend on each platform"

That's a good idea if you are comparing the options for the general user looking to buy a computer. In this case, they are reviewing the top-of-the-line processors (which people on a budget won't be buying anyway.) Only those with a large budget will be buying a machine with one of these processors anyway, so price shouldn't be an issue.

Also, their decision to use DDR RAM for both platforms is misguided. Those who pay the big bucks for the fastest processors won't skimp on the rest of the PC, so the comparison really should be done using the RD RAM the Pentium 4 is designed for. Instead of trying to artificially force the platforms to be as similar as possible, they should compare real-world hardware configurations.

Re:Why DDR on P4?

[bbqBrain]

That would turn out really ugly for the Intel products involved. The amazing thing is that one can build a damn fine, complete AthlonXP 2000+ system (sans monitor/keybd/mouse) for about $1000. With Intel, that same $1000 gets you a 2.2GHz P4, nice mobo, 512 MB RDRAM, and case/PS. Save those pennies, and you can eventually get your video card, sound card, CD-RW, and HDD. And you'll still only see performance on par with your friend's XP2000+ system.

Intel certainly seems capable of producing some really nice products, and they aren't even pushing the 0.13u chips yet, from what I've read. You'd think that if they wanted to show AMD up, they'd just clock them at > 2.5GHz now and keep going. I'm puzzled.

Re:Why DDR on P4?

[GigsVT]

Art of war.

When you are strong, appear weak, when you are weak, appear strong.

All warfare is based on deception. [1:18]
Hence, when able to attack, we must seem unable; when using our forces, we must seem inactive; when we are near, we must make the enemy believe we are far away; when far away, we must make him believe we are near. [1:19]

Re:Why DDR on P4?

[Master Bait]

That's a good idea if you are comparing the options for the general user looking to buy a computer. In this case, they are reviewing the top-of-the-line processors (which people on a budget won't be buying anyway.) Only those with a large budget will be buying a machine with one of these processors anyway, so price shouldn't be an issue.

Well then if price is no consideration, then why don't they just benchmark the P4 against IBM's Power4?

Re:Why DDR on P4?

[ryusen]

i've got an idea.. why not run the test on BOTH rdram and ddr? that way we can actually see what the difference in the two chipsets are? and then add in a little thing about th eprice differences so we cna judge that for ourselves too

Re:Why DDR on P4?

[ergo98]

Well my contention is that the processor alone does not a PC make. Perhaps the AMD processor requires a more expensive motherboard and more esoteric cooling system, meaning that they'll have to sacrifice elsewhere. Perhaps RDRAM exacts a cost penalty that hurts the Intel elsewhere.

Money is never "no object", and just about anyone who states that quickly changes their tune when the $s add up. Why don't the big comparison tests include 15,000RPM hard drives and U160 SCSI? What about some of the high performance server backbones? They don't because those features exact some hefty costs, and when Joe Average who thinks he's going to max out his system sees what that costs with his GeForce3 Ti500 64MB video card, super 16 channel 24-bit soundcard, etc, something always gives. I've done that classic spreadsheet game a million times where I settle on a lesser harddrive but up the RAM, etc.

In the end the $ is always the deciding factor. Just because a Nissan Altima and a Ferrari Testarossa have 4 wheels doesn't mean that they're directly comparable.

Re:Why DDR on P4?

[connorbd]

Actually, I believe the Northwood P4 w/DDR is the overall leader just at the moment. And I believe Tom's Hardware just managed to get it cranked up to 3.1GHz...

As for using Rambus and the i850 -- nobody really wants it. Notice: as long as the P4 was tied to Rambus, it was a flop. These days a P4 is not particularly expensive (largely because practically every P4 consumer system on the market is SDRAM-based).

/Brian

Re:Why DDR on P4?

[ergo98]

As I mentioned in another thread: Why not benchmark with a RAID 5 array of 7 15,000 RPM U160 SCSI hard drives?

It always comes down to $, and I really think for these comparisons to have any relevance they should be completely based upon $, rather than just blanket comparing systems built around certain clock speeds.

Re:Why DDR on P4?

[liquidsin]

I agree with you one hundred percent. I'm currently trying to build myself a new system and constantly weighing options. However, from what I can tell from the story on slashdot (linuxhardware.com is *still* inaccessible) the testers were simply comparing the processors. You can always add more bells and whistles, but plugging high-end scsi drives into both systems *shouldn't* produce a larger performance gap on the test systems. The point I was originally making is that this test isn't to see if you get more performance out of AMD or Intel based on the same budget - it's to see which company's top of the line processor performs better on relatively similar hardware. We could put an 80Gb, 15,000 rpm scsi harddrive on both. Hell, we could add a dvd burner and a twelve foot flat screen monitor, but it won't change the system performance of one system relative to the other (or at least it shouldn't).

Re:Why DDR on P4?

[Hamshrew]

The XP series have thermal diodes, so that comment isn't valid in this case. Besides, out of all the computers I have worked with(over 1000) I can think of only one that ran without a heatsink... and the guy forgot to put it on, it didn't fall off.

And before you make comments on what an idiot he was... you try building 200 systems in a row with no mistakes :)

Re:Why DDR on P4?

[liquidsin]

er, I guess that's linuxhardware.ORG, not .COM

I know, I know...READ THE ARTICLE, and don't reply to my own posts, or something...

Re:Why DDR on P4?

[scd]

P4 was not particularly designed for RDRAM. It was designed to utilize large amounts of memory bandwidth while not caring as much for large-ish latencies. RDRAM most certainly has these characteristics, of course.

Plus, given the price disparities between DDR and RDRAM, comparing these systems shouldn't be too out of hand.

Re:Why DDR on P4?

[Chromium_One]

The XP series have thermal diodes, so that comment isn't valid in this case.

I've heard a few nasty rumors about Athlons and the thermal diodes... namely that the diodes aren't used as a marketing tactic partly because the CPU likes to blow in order to protect the diode. I'd like a pointer to more info before my next upgrade.

Re:Why DDR on P4?

[RedWizzard]

Why didn't they test price/performance? A hundred other sites have already done raw performace comparisons, and they don't mean anything to me because I don't have an unlimited budget. I already know that different processors have different performance so what use is the information they are providing? I don't follow the benchmark scene closely but it seems to me that no-one ever compares price/performace, yet that is by far the most important measure for 99% of computer buyers.

Re:Why DDR on P4?

[BigBir3d]

Those thermal diodes on the AMD XP series are very problematic. It turns out, the diodes were designed for a change of temperature of only 1 degree per second! This is so slow as to be useless. Think of the rate of change of temperature when a fan fails or a heatsink & fan fall off/partially disconnect. You get the result that Tom's Hardware did here.

Re:Why DDR on P4?

[ergo98]

but it won't change the system performance of one system relative to the other

But my point is that if this is truly the "Best of the best", then both systems should include U160 15,000RPM hard drives (because it makes a BIG difference in things like kernel compile times), but they don't because most "normal" folk don't buy such a hard drive. Why? Cost. Well if cost is a factor then let's completely factor it in.

Re:Why DDR on P4?

[Anonymous Coward]

The Athlon XP's will not fry themselves if you pull off the heatsink.... Well, kinda. Current Motherboards do not measure the thermister, or at least not well enough. So, when an abrupt change occurs in temp the processor can fry. However, AMD has since come up with plans to add a logic circuit to the mainboard which will cut off power to the processor and board if temperature goes too high. Check the actual story at Tom's Hardware.

Re:Why DDR on P4?

[mgblst]

Because nobody with an ethical bone in their body
would buy RAMBUS.

Or go see a movie.

Or buy a new cd.

becareful, you might be labelled a commie...

Re:Why DDR on P4?

[Mr. Piccolo]

Because a RAID 0 array would be faster!

Re:Why DDR on P4?

[Alan Partridge]

err... maybe because there are ZERO socket A RDRAM motherboards?

Re:Why DDR on P4?

[ryusen]

maybe i wasn't clear, but my post was in a reply to a comment asking why they are testing the p4s with ddr only and not rdram, so i was suggesting that they test rdram and ddr on the p4 solution, not the athlon

Re:Why DDR on P4?

[WNight]

The chance of a heatsink falling off completely is pretty small. Only once have I dropped a computer hard enough to possibly cause this (and when I did I opened it up and inspected it, reseated cards, etc).

Worse is the possibility of turning it on without the heatsink properly seated. But ideally since most boards/BIOSes start them seriously underclocked, this would be detected before they kick the speed up during the POST.

Re:Why DDR on P4?

[qurob]

Just take the # from a benchmark, and divide it by the price. Then you've got the value you're looking for.

Re:Why DDR on P4?

[VAXman]

If the Power4 ran an instruction set which more than five people cared about, they would.

The target market clearly is PC enthusiasts, who, along with 95% of the world, only care about X86.

Re:Why DDR on P4?

[VAXman]

You'd think that if they wanted to show AMD up, they'd just clock them at > 2.5GHz now and keep going. I'm puzzled.

You're confusing sports with business. Intel's job is to make money, not to show up its competition. Don't assume that Intel's profits would be boosted by having the fastest part around. Indeed, in 2000, Intel made $10 billion in profit which is more than any other tech company has made -- before or since -- in a single year, but for at least half of the year, their fastest part was appreciably slower than AMD's (AMD, meanwhile, who is highly interested in showmanship, is losing money every quarter).

Re:Why DDR on P4?

[RedWizzard]

But they don't provide the price. I could go to pricewatch or whatever and find out what the price is but that's a pain in the ass. If you write an article comparing two systems you should include the price, without that it's useless information.

Re:Why DDR on P4?

[BigBir3d]

What you say is all true, but you know what you are doing. People that go into Gateway and say, "I want the fastest thing you sell..." probably don't. With a P4, if something goes weirdly wrong, the chances of melting it are much much less than with a Athlon.

Re:Why DDR on P4?

[Stone Rhino]

it was 3.01, WITH water cooling and a handpicked processor. and it still wasn't fully stable.

Re:Why DDR on P4?

[zeno_2]

I don't know myself, but this is my guess..

The pentium 4 isn't optimized for either ddr or rambus memory. The motherboard chipset that intel made for their processor uses rambus. My next guess is because rambus wow'd them and they signed an exclusive contract that pretty much made it so you could only use rambus and thats it. Seeing this, intel is going to market the pentium 4 and rambus as a good pair, saying that they work better together, etc.

Like I said, I don't know for a fact if its true, but it seems likely..

Re:Why DDR on P4?

[Znork]

Not quite right. The diodes _on_ the XP work just fine, there just arent many (any?) motherboards that support them.

The thermal diodes _outside_ the CPU, on the motherboards, are the ones that are slow. It's a case of the motherboard vendors not updating design to use the new thermal sensor in the XP.

Should you find a motherboard that does support the XP thermal diode it wont burn.

Re:Why DDR on P4?

[13Echo]

Typically, AMD based mobos are cheaper, and a $10 Cooler Master heatsink/fan is perfectly suitable for cooling. You need no more, unless you intend on overclocking the chip. Aside from that, every platform, regardless of CPU can at least benefit from an extra 12V fan to blow hot air out, or cool air in.

Besides... I have noticed that the Athlon XP 1600+ runs 5-10 C degrees cooler than the 1400 T-Birds. They are the same clock speed, but have a totally different core. It also depends on your mobo and BIOS among other things, but typically, the new Athlon XP line stays nice and cool.

Really though. Even if Athlons were more expensive to cool, do you think that it would cost an extra $200-$300? It's not likely.

Re:Why DDR on P4?

[Scooter]

Well yes - and that was the point he was making really. If you've got your $2000 to spend, why would you buy the inferior performing product?

I think this goes for just about any product you care to mention which involves easily measurable specifications - the budget is king.

Strangely, when trying to sell a server solution to the pointy haired ones, I may as well be saying "The gobi desert went out for a hamburger" - they just look at you like you just announced you'd be nailing banana skins to the office walls for the next 2 days. "Look at the report - it's free, goes faster and stays up longer." "yeah but why bananas?" is all you'll get back... :)

I imagine I would get the same repsonse if I said - "look - we don't have to buy Pentium based servers - check out the Athlons, we buy lots of cheaper servers rather than one monolithic 8 way pentium - which gives vastly dimnishing returns on the number of CPUs" PHBs: "What? power our new e-commerce project with hotdogs? What do you mean?!?" ...

Re:Why DDR on P4?

[liquidsin]

Bah, we could do this for days on end. But while "normal" folk may never buy a 15,000 rpm scsi drive, the linuxhardware.org folk are hardly normal. That said, they *could* choose to build the total high end system, but I seriously doubt that it would effect the outcome of the testing. If both systems use the same high speed scsi hard drive, the systems will greatly improve, but both should show the same improvements, and thus there's no difference. But hey, if you're buying, I'll take 15,000 rpm 100 Gb drive any day ;).

p.s. - I would love to see a comparison done with a TCO sort of thing. Sure, the AMD is less than half the price of the Intel, but is the install and config longer? Is it *really* more likely to die? Put together the best P4 and Athlon XP systems you can for, say, $2,000, including build time (set a rate for billing techs) and see who comes out on top. THEN you can do the benchmarks.

Re:Why DDR on P4?

[liquidsin]

I normally ignore ACs, but hey...I'm here. Whether you go by pricewatch (as I chose to do) or CompUSA or the scary man wearing 40 lbs of gold chains selling OEM shit out of the back of a van in the alley behind the barber shop, you'll come up the same. For the same dollar ammount, you'll get a much faster Athlon. The Athlon XP 2000 is actually *cheaper* than a P4 1.9 Ghz, and all benchmarks I've seen show that the Athlon can outperform a 1.9 Ghz Intel. From what I've seen, the 2.2 Ghz P4 is at least twice the cost of the Athlon 2000. So yeah, I don't care about cost, since I'm not in the market for either, I just want to see whose top of the line processor performs better. If the Athlon wins, great, since it's way cheaper. If the Intel wins, does it win by enough to make it worth double the dollars? THAT'S what *I* care about.

Re:Why DDR on P4?

[13Echo]

Even with T-Birds, they will most likely just lock up the machine before any real damage can be done. It doesn't take much time to fry a CPU, but the thing is normally going to lock up and puke long before it dies.

I've tested this with a Duron, and while they don't get as hot as the fastest Athlons, they still get pretty darn hot.

Re:Why DDR on P4?

[connorbd]

You didn't really read the piece, did you? They tried it with two processors, one an off-the-shelf Northwood (which did require water cooling) and one a hand-picked chip supplied by an unspecified mobo vendor.

/Brian

Re:Why DDR on P4?

[frozenray]

The chance of a heatsink falling off completely is pretty small.

Depends on how you use your PC, IMO. It's not a problem with my 1.4 TBird at home, because I assembled the system at home and do not plan to move it around.

If you want to ship the box, or even take it to a LAN party in your car over bumpy roads, you better think about removing the heatsink - especially if you have one as heavy as this beast or one with a high center of gravity like this one. Unfortunately, removing the heat sink without damaging the processor die or popping off a few SMD resistors with a badly designed clip isn't exactly trivial in many cases, not to mention that there's also the issue of correct handling of thermal paste/tape involved.

In my PC, there are only two flimsy plastic noses on the CPU socket securing the heat sink clip, and the sink is quite heavy. I shudder at the thought of the havoc the heat sink could wreak if it should come loose during transport. And should I inadvertently turn on the system without heat sink in place, the CPU would be toast before I even had time to notice something is wrong (the BIOS doesn't start at a 'safe speed' for all I know).

AMD has to do three things in my opinion:

1. Provide an internal (fast) thermal diode similar to the one used in the P4, throttling the CPU down to a safe speed if temperatures get too high.

2. Provide a heat-spreader which protects the CPU die. My local PC parts dealer has half a shoebox full of AMD CPUs with broken dies. It's not really a problem for him, because it's not the dealer who pays for a new CPU if the customer screws up heatsink assembly, but it doesn't generate happy AMD customers.

3. Redesign and standardize the heat-sink/motherboard attachment and provide mandatory design guides for mainboard manufacturers. There are too many clip designs around, some sinks cannot be mounted at all in some motherboards, etc.

Raymond

Re:Why DDR on P4?

[WNight]

Two issues:

1) Was Tom's video a fair representation.

IMHO, No. Unrealistic circumstances = Fear Mongering.

2) Does AMD need to do something.

IMHO, Yes.

I agree completely with #1 and #2. Especially #2.

#3 however, has to do with an industry standard. AMD can't really change it much.

They could however say that all fans must screw into the board (like the P4 fans) or use all the plastic tabs, if the company wants to say "AMD Approved". (Or, for alternate methods, satisfy AMD engineers that the mounting method is sufficient.)

--

I really don't think there's much chance of a heatsink falling off, even the big copper ones, as long as you're careful with how you move the computer.

Re:Why DDR on P4?

[Stone Rhino]

I DID read the piece and that was what was implied to me. I was not focusing on that part of the piece when I read it.

Re:Why DDR on P4?

[frozenray]

1) Was Tom's video a fair representation.

IMHO, No. Unrealistic circumstances = Fear Mongering

It sure drew a lot of attention, though (and I bet AMD was not happy about this kind of publicity). If anyone from AMD listens: "cracked die" and "fire hazard" are two words we geeks don't want to hear when it comes to our processors.

#3 however, has to do with an industry standard. AMD can't really change it much.

I just wish they tried a little harder. Some clip design I've seen suck toxic waste, some boards have capacitors that come too close to the CPU etc. Unfortunately, Intel is in the better position when it comes to tell motherboard and heatsink designers what to do.

Raymond

Level playing field?

[Anonymous Coward]

Wouldn't it be more of a level playing field if both processors were allowed to use their optimum RAM types? Sure, keep the rest of the system honest, but don't handicap one processor by forcing it to use a RAM type it wasn't initially designed for.

Re:Level playing field?

[L-Wave]

Maybe both machines should use EDO RAM? =)

Re:Level playing field?

[Master Bait]

I'm not sure that Rambus is all that optimal. It has horrible latencies compared to DDR. The DDR333 RAM in those SiS645 motherboards definitely keep up with the intel850-based rambus motherboards in reviews I've seen.

So if you look at the different benchmarks (if linuxhardware.org ever recovers) you'll quickly see that Rambus is only better for streaming large amounts of data, such as encoding video. And we all spend a lot of time encoding video, right?

For general purpose desktop computing or serving, the lower latencies of DDR, and in particular DDR333, are all over that Rambus crap.

Re:Level playing field?

[mr qix]

i think if a chip manufacturer is going to not only make it possible to use multiple types of RAM, but also endorse that (i.e. Intel's switched stance on non-RDRAM mobos) it's only fair to use whatever type of RAM you want to test the processor.

Re:Level playing field?

[Soko]

As someone else said, the P4 is memory bandwidth hungry, but it wasn't really designed with the requirement for RDRAM in mind. It should be able to use any memory access method as well as or better than the Athlon. However, due to it's 20 stage instruction pipeline, it takes more of a performance hit when a program branch is mis-predicted. This is magnified more when you use slower memory - likely the cause of the Athlon being able to keep up with lower MHz.

Soko

Re:Level playing field?

[Chris Burke]

Well, to a degree. To a first order, the Athlon keeping up is just because it is more of a "brainiac" (high IPC) than the P4 "speed-demon" (high clk freq.), though the P4 also doesn't resemble any of the machines normally called speed-demons. Memory access is certainly a factor in this, but the fact that the XP "keeps up" is due to its higher IPC (which is more complicated than just branch mispredictions, though that certainly is a big factor for the P4).

The P4 wins in streaming benchmarks because it can use the higher bandwidth RDRAM (and by use I mean not only capatability but having the FSB bandwidth to not throttle the memory). There is no reason to think that it should be able to use any memory access method "as well as or better" than an Athlon, because no matter what it will be limited by its inherently lower IPC.

Re:Level playing field?

[kanayo]

Wouldn't it be more of a level playing field if both processors were allowed to use their optimum RAM types? Sure, keep the rest of the system honest, but don't handicap one processor by forcing it to use a RAM type it wasn't initially designed for.

If the comparison were between machines, or between chipsets, perhaps it would be a good idea to bring optimized configurations of each. The comparison however, is between CPUs, in which case adding another variable such as memory type will only serve to make the results less credible.

(Besides, they were both designed to be able to use DDR memory.)

Re:Level playing field?

[TheAwfulTruth]

How can you even do that considering the chipset on the MB is also different between the CPUs? You are automaticaly testing the "system" not the "CPU" no matter what memory you used becuase the MB/Chipset is different. :(

Re:Level playing field?

[DemiKnute]

Hey hey...here's how it goes. (I don't know if anyone has used this analogy before, but it's new to me). Transport yourself back to the 1920's or so.

The P4 is like a race horse. Really honkin' fast, but it needs to know where it's going and gets a little skittish at times and then it doesn't do so hot. Rather on the expensive side, too.

The Athlon XP is more of a draft horse. It's not very fast, but you can throw damn near anything at it and it'll never flinch, just keep on going.

Remember the fable? Slow and steady wins the race.

-David

DDR vs. RDRAM

[Henry V .009]

Why the hell does using DDR memory even the results? The Pentium 4 was designed for RDRAM for God's sake. A DDR platform for the P4 is just an economy decision.

To make things really even, you could take a hammer and drive the P4 into a socket A Motherboard. Then you could really compare them, couldn't you? Perfectly even field.

Re:DDR vs. RDRAM

[larien]

Yup, PIV's are memory bandwidth hungry (which is why I cringe when I see P4s with SDRAM). Fact is, this 'economy' decision is what drives consumer PCs; the average sensible consumer should accept the lower clock speeds of the Athlon (even if he sees behind the marketing numbers) if he saves a few hundred bucks.

If RDRAM can get its prices down to closer to DDR, it might actually compete properly. Until then, AMDs lower prices and the lower price of DDR ram is going to wipe Intel's ass on value for money.

Re:DDR vs. RDRAM

[fishebulb]

i thought RDRAM was not a very good memory. Correct me if im wrong, but isnt the latency really high. I cant remember exactly what it was. Or was it just the early versions?

Anyone with some insight on RDRAM

Re:DDR vs. RDRAM

[autocracy]

From the parent to this post: Moderation Totals: Funny=1, Total=1.. Note at this time that the score is currently 5!

Re:DDR vs. RDRAM

[Henry V .009]

That's why I build AMD machines, not Intel. But they were doing a hardware comparison.

Re:DDR vs. RDRAM

[Glock27]

Why the hell does using DDR memory even the results? The Pentium 4 was designed for RDRAM for God's sake. A DDR platform for the P4 is just an economy decision.

Except that RDRAM hasn't been accepted well in the marketplace, and systems based on such RAM are more expensive. Most people buying a P4 are getting either PC-133 or DDR.

I'll grant you that it's silly, given that (according to Pricewatch) RDRAM is commanding only a 15% premium these days. I guess Rambus Inc. really did annoy just about everyone in the industry...RDRAM seems to be anathema.

It's pretty amazing that the Athlon so regularly beats higher-clocked P4s using RDRAM, given DDR's inferior bandwidth. Intel is ripe for the plucking! (Anyone happen to catch Jerry Sander's statement at the last AMD earnings conference call? "Intel is over", quoth Jerry. Now that's what I call confidence!)

299,792,458 m/s...not just a good idea, its the law!

Re:DDR vs. RDRAM

[larien]

It has relatively high latency, but it has much higher bandwidth which is what the PIV wants.

Re:DDR vs. RDRAM

[RedWizzard]

It's a stupid attempt to remove a variable in order to isolate the CPU performance. The problem is that the benchmark community seems to think that CPU comparisons are relevant to normal buyers. They're not. The only thing relevant to 99% of buyers is overall system performace at a given price. Yet I've never seen one attempt to compare systems that way.

Re:DDR vs. RDRAM

[RedWizzard]

Unfortunately most of the magazines I've seen (though I don't really read them) tend to only give star ratings or whatever rather than actual benchmarks. What I'd like to see is a comprehensive benchmark comparison between different systems with different CPU, RAM, HD, graphics card, etc, at the same price point. That would be useful to me.

Re:DDR vs. RDRAM

[Alien Being]

Right.

Makes me think wonder about what AMD could do in their marketing dept to combat Intel's incumbent position. The bang-for-the buck argument is what they've got to get across to the public.

Maybe...A superbowl ad that analogizes computer performance with a football game, sorta like the Bud Bowl. AMD could call it the Bit Bowl.

Re:DDR vs. RDRAM

[Glock27]

Was that the same earnings conference where AMD announced that they lost money yet again?

Sure, under extremely difficult circumstances (many great companies lost money over the last quarter), predatory pricing from Intel, and taking into account some large infrastructure investments.

I'm sure they'll handily beat estimates this quarter, just as they did last quarter. And yes, AMD will turn a profit this quarter (barring unforeseen disasters etc.).

I'm not sure what you mean by Intel being ripe for the plucking - AMD is selling 100% of fab capacity right now.

Haven't you heard about the 0.13 micron fab coming online this quarter?!? Also, AMD can farm out as much production as it wants to other fabs. This will first happen in the Far East as a way to get more Durons into the marketplace. We'll see later on...IBM has some very advanced fabs these days, and there has been cooperation with AMD on things like copper technology.

For the near term, they can only raise prices, not gain marketshare. Which considering their poor profitability, is probably something they should consider.

Not only are they considering it, they've already gotten a good start on it! Average selling prices were up 50% over the previous quarter, from $60 to $90.

Regardless, the real reason to buy AMD stock is to bet that Hammer will sample this year and ship in volume next year. If that happens the whole competitive landscape will be dramatically altered, and AMD stock will go through the ceiling. Shorting Intel will be the way to go then. =)

299,792,458 m/s...not just a good idea, its the law!

Reason for DDR

[FileNotFound]

The reason DDR was used is because there have been COUNTLESS tests done with RAMBUS.

The whole goal was to see how well it'd do with DDR now that it supports it.

Tom's Hardware Did This 2 Weeks Ago

[LuxuryYacht]

Take a look at

http://www6.tomshardware.com/cpu/02q1/020107/ind ex .html

They posted the results of their showdown 2weeks ago.

Re:Tom's Hardware Did This 2 Weeks Ago

[glazed]

That comparison isn't at all the same. Tom's uses an Asus P4T-E, a RAMBUS board.

Re:Tom's Hardware Did This 2 Weeks Ago

[ryusen]

but tom also did a p4 chipset test:

http://www6.tomshardware.com/mainboard/01q4/0112 17 /index.html

it should give you numbers in around the same ratios if you just adjust for the cpu clock speed correct?

Re:Tom's Hardware Did This 2 Weeks Ago

[keeg]

Just for the record, the correct link is http://www6.tomshardware.com/cpu/02q1/020107/index .html

Here is anouther one at anandtech...

[Neck_of_the_Woods]

Anandtech match up

Interesting at least.

Re:Here is anouther one at anandtech...

[Neck_of_the_Woods]

http://www.anandtech.com/cpu/showdoc.html?i=1574

sorry about that... above link is correct.

In an effort to remain accurate...

[Restil]

I would check the website so I know what I'm talking about when I comment. But since I can't access the site due to the (I assume) brutal slashdotting, I feel almost compelled to comment without any supporting information to base my wildly inaccurate opinions on. If at least the article summary had summarized (probably incorrectly) the article content beyond saying it was "interesting" then at least we could get the debate rolling, at least until the page became accessible again. At which time, everyone else would join the fray complaining that nobody reads the articles.

-Restil

Re:In an effort to remain accurate...

[Cardhore]

I agree. Google can cache its sites. Why can't slashdot?

Re:In an effort to remain accurate...

[Anonymous Coward]

Google runs off a cluster of 10,000 servers and employs 250 people. Slashdot does not.

Re:In an effort to remain accurate...

[Lyrrad]

I agree. Google can cache its sites. Why can't slashdot?

How did this get modded up?

See the FAQ for the answer: Here

Re:In an effort to remain accurate...

[fliplap]

Money

Mirror?

[Archie Steel]

I don't think so, however there are other benchmarks here.

can't get to article...

[night_flyer]

related?

But...

[Matt2000]

... they're not really levelling the playing field because DDR memory is a mature option for AMD whereas it's brand new on the Intel boards, and apparently has some problems.

If you're going to compare just CPU power then use synthetic benchmarks that test just that, otherwise if it's system performance you're going after why not compare AMD DDR to Pentium 4 RDRAM, at least those are two mature configurations.

Processor 101 (Re:DDR vs. RDRAM)

[autocracy]

The P4 wasn't designed for any particular memory - but the initial chipsets designed for it were. Newer chipsets now existt hat are designed for DDR rather than RDRAM, and that's what matters for memory format (besides, try to switch the two and THEN see what happens!).

Re:Processor 101 (Re:DDR vs. RDRAM)

[hobit]

the P4 wasn't designed for any particular memory

Not quite true. When you design a processor you do consider your memory type. For example, RAMBUS gives better potential bandwidth and slightly better latency. This will impact the expected memory read time. As modern computer architecture can be summarized as "finding useful things to do while waiting for memory accesses to finish" the memory speed certainly does impact the design of the processor.

That said, the slight differences between the two memory types probably is in the noise margins for this issue. But I promise you that someone on the processor design team was thinking about memory types when designing the P4.

Mark

Re:Processor 101 (Re:DDR vs. RDRAM)

[Henry V .009]

The P4 wasn't designed for any particular memory - but the initial chipsets designed for it were. Newer chipsets now existt hat are designed for DDR rather than RDRAM, and that's what matters for memory format (besides, try to switch the two and THEN see what happens!).

Wrong. The P4 was designed for high memory bandwidth. In fact, that may even be why it performs better on RDRAM chipsets than DDR chipsets. Who wudda thunk it?

Goddamn slashdot moderators. My orignal post gets modded down as a troll for pointing out a valid hardware issue, and this piece of cluelessness gets modded up.

Re:Processor 101 (Re:DDR vs. RDRAM)

[Junta]

Well, RDRAM is crap in single-bank configurations. High-latency, and low bandwidth (relative to DDR). True, RDRAM runs at effective rates of 800 MHz, but the data path is reduced to 16 bits over the SDRAM 64 bit typical bus. Granted, RDRAM can allegedly be more easily configured in configurations where modules can be run in parallell, providing effective 32-bit bus, but I haven't seen this implemented in any consumer grade PC motherboards, and even then it wouldn't be clear cut that RDRAM would be a winner (especially with it's high latency). The only reason why RDRAM could even win is because RDRAM is much more efficient in a way, it can come much closer to its theoretical bandwidth than SDRAM can come to its theoretical limit.
In essence, Rambus memory is a complete bust, and Intel is *finally* giving up at least in part.

Re:Processor 101 (Re:DDR vs. RDRAM)

[Pulzar]

A 2-channel 266MHz DDR memory controller has the bandwidth of 2x266x64/8 = 4.2GB/s. P4 front-side bus has the bandwidth of 4x100x64/8 = 3.2GB/s.

Even a single-channel of the new 333MHz DDR has the bandwidth of 2.7GB/s, almost enough to keep up with P4. Once the 400MHz DDR shows up, a single channel will have the same bandwidth as the P4 bus.

So, in a well designed DDR system, P4 will not have any bandwidth problems -- if anything, RDRAM systems will be slower because of RDRAM's lower latency.

Re:Processor 101 (Re:DDR vs. RDRAM)

[autocracy]

You forget speed. Bandwidth and speed have a lot to do with this together. First you must understand that the faster the processor (not in raw megahertz, but actual performance), the faster it can take data from the RAM. Your goal is to make sure that the RAM can provide it fast enough. RDRAM works by pushing a lot, but it does it rather slower. This means you end up waiting 3 clock cycles (that's NOT a valid number, it's an example) to get the data, but you get a shitload of it once it comes. DDR, however, uses a lower bandwidth (and I should note here that bandwidth is probably NOT the right term - though I can't think of what is), but a faster speed. It's akin to sending food by semis vs. sending it by cars. The semis are slower and take longer to load, but the cars are faster and take less time - even though they don't hold as much. Get cars that load and drive fast enough and you won't have to worry about capacity (and fuel economy and operator salaries don't really matter here either - computer's don't care about that).

Another person who replied to your was right - stop spreading FUD and KNOW before you speak (yeah, I've been guilty of it too - but learn!).

heh

[Joe the Lesser]

the Intel vs. Athlon war... "Breaking news from the front today. The armies of Athlon continue to push against the Intel positions. However, Intel's unique 'chime' technology his inspired their troops to hold the line..."

Re:Slashdotted... Why no /. mirror?

[renehollan]

He he.

Mine involved encoding it in white space between words and having the words describe the encoding.

Oh! No! Gotta ban spaces now!!

Yeah but did they

[afidel]

apply the AMD no Large Page Table processor bug patch?

Re:Yeah but did they

[Magila]

As this highily informative post points out, AMD has already fixed the problem in the latest revision of the Athlon XP.

Re:It always cracks me up...

[ender-]

Perhaps /. can warn sites that they're gonna link to them?

The funny part is that the "Augustus" that submitted the article is from linuxhardware.org . Which leads me to believe that they knew what was coming...

It's amazing that my little K62-350 stood up to the /. effect on 09-11-01 with no problem. Yet a fairly major web site can't handle it. Hmm...

Ender

Pretty irrelevant

[Ryan Amos]

I don't see why hardware sites insist on seeing which chip "is fastest." I'd be more interested in an acceptable price/performance ratio. The Athlon XP 2000+ (which can still hold its own fairly well against a P4 2200) costs LESS THAN HALF of a P4 2200. Why anyone would spend the extra $350 on a P4 for the minimal performance gains (relative to the cost) is beyond me. And for those who want absolute, unforgiving, raw performance.. For the same price as a P4 2200 with a decent motherboard, you can buy a Tyan Tiger MP with a pair of Athlon XP 2000s and a bunch of DDR memory (AMD reccomends you use Athlon MPs but there's no reason the XPs won't work.) Sure, graphs and kernel compile times are pretty and all, but eventually you have to think about what is practical..

Re:Pretty irrelevant

[agurkan]

well,

for one thing CPU is not the only part of hardware you buy. eg. if you are building a cluster, and need a certain computing power having a performance increase of 10% might be significant since it will save you 10% of nodes which might include a gigabit ethernet card or a myrinet card which cost ~$1500 (i am not sure about the price, but it is in the right ballpark). and, this is not counting all other things that go in a cluster node. that is why dual processors are not such a bad idea in clusters :)

so, IMHO it is pretty relevant.

Re:Pretty irrelevant

[Trogre]

I would be more interesting in reliability, as well as raw performance.
What is the MTBF of these units, and how many errors per billion calculations are encountered?

Run some cpu-hogging rendering benchmark (eg povray) over a *long* period of time and do a bitwise comparison of the final images.
Time how long it takes each to crash when running Linux Quake3 benchmarks.
I'm sure someone can come up with better reliability test than these, but it's a start...

Re:Pretty irrelevant

[ImaLamer]

I'm looking at RAM prices now, and for the $350 difference, that means about a gig-o-ram.

I think I would save the cash and rather have the gig-o-ram than have no RAM at all.

Of course you can also go the dual-cpu route, and that seems nice too. But for someone who wants a computer to do whatever on, get the AMD and gig-o-ram.

I just like thinking gig-o-ram. A few [3-4] years ago I upgraded to 64MB and was like "wow! this is smoking!"

gig-o-ram :-)

Re:Pretty irrelevant

[Black Parrot]

> I would be more interesting in reliability, as well as raw performance.

For my next purchase I will be looking at power/heat as well. I don't subscribe to the "1G is enough for anyone" school of thought, but I am starting the think I'd trade down some raw speed to keep from running a heater in my room during the summer.

Re:Pretty irrelevant

[RedWizzard]

so, IMHO it is pretty relevant.

How many people build clusters in their bedrooms? Anyone serious about a cluster will be doing their own benchmarks using the apps they intend to run on the cluster. So IMHO it is not relevant, at least for anyone interested in clustering.

Re:Pretty irrelevant

[Anonymous Coward]

I've used both the fast Athlon XP's and the new Northwood P4's. They're both plenty fast.

The reason I *did* pay double for the P4 northwood is that the support chipset is not a piece of garbage like many of the Athlon support chips.

I really *like* the price / performance of the Athlon chips I've had, but the VIA support chips suck really bad. Hard drive corruption, PCI bus noise, miscellaneous other PCI bus problems. Just not worth the half-price savings IMHO.

Re:Pretty irrelevant

[chihowa]

Maybe I'm missing some point here, but why don't you save the difference and buy a mobo with a chipset that doesn't suck.

Maybe I'm just not using mine the same way you're using yours, but my VIA chipset hasn't given me any problems so far.

Re:Pretty irrelevant

[rgmoore]

There are reasons that you might want to consider spending the extra on a really hot processor. The main one I can think of is if there's something else very expensive as part of the setup. The biggest, most obvious one I can think of is software licenses. I've encountered software with licenses of as much as $8000 per processor, which obviously places a very large premium on having the fastest possible processor running it. I'm sure that there are other applications with even more outrageous prices, too. When you're forking out that much for your software, hardware costs start looking pretty tame.

Re:Pretty irrelevant

[BeazleyR]

just to nitpick, I own a Tyan Tiger MP, and in the users manual it only supports up to the Althon 1600.

Re:Pretty irrelevant

[aussersterne]

It's down to mainboard manufacturers. Some of them are very cheap.

I paid extra for an Asus board. It's got a VIA Athlon chipset, and I'm running a 1GHz T-bird with five PCI slots and the AGP slot full. Three of the PCI cards have PCI bridges on them; this machine has:

2x 10/100 Ethernet ports
3x LVD SCSI channels
2x PCI Audio
1x Conexant NTSC video
1x Philips MPEG hardware encoder/decoder
4x USB channels (3 being used)
768MB RAM at PC133 2-2-2
and GeForce2-Pro in the AGP slot

This running with the Nvidia drivers under Linux. I'm running kernel 2.4 and getting uptimes in weeks, with at times very heavy I/O and multimedia IRQ load. The root filesystem is using the VIA 686b UDMA-100 support, while the data filesystems are on LVD SCSI.

The Athlon chipsets don't have to suck... It's all down to the quality of hardware you buy. No-name import == bad for stability, in my experience, and P4 boards are no different! Some of the current no-name P3 and P4 boards on the market are total crap, but not because of the chipset... because of shoddy design.

Re:Pretty irrelevant

[MonkeyBoy]

I own a Tyan Tiger MP, and in the users manual it only supports up to the Althon 1600.

Well, at least you read the manual. That's more than I can say for most people putting PCs together!

Anyhoo, manufacturers typically add support for newer/better/faster processors through BIOS updates. Obviously miracles are impossible, but if the chipset can support the processor it should be possible for the manufacturer to add support. Witness those Socket370->Slot1 adapters from a couple years ago - the processors were much faster than originally spec'd, but with a BIOS update (and the slocket) you could usually use them.

You should be able to find your latest BIOS here.

Only thing I don't like about Tyan is they never seem to keep on top of updates like some other manufacturers. Granted they're much better than most no-name places, but...

Re:Pretty irrelevant

[VAXman]

If you were building a central database server, and one option delivered 5,000 transactions per day, at a cost of $1 each, and another delivered 500,000 transactions/day at the cost of $1000 each, and you had demand for 1,000,000 transactions a day @ $1,000,000 each in revenue, which server would you choose?

If you choose the one with better price/performance you will get fired in a heartbeat.

Get a clue: price/performance is a bogus benchmark because it assumes the cost of lower performance is $0. Period.

Re:Pretty irrelevant

[ghmh]

How many people build clusters in their bedrooms?......

Quite a lot of people I assume - people are certainly imagining it all the time at least, particularly the Beowulf variant.

For their sake, I hope they all have large bedrooms... - In fact you could take it to the next level:

"Imagine a cluster of bedrooms of clustered Athlon XP's"!!!!! (blah blah blah)

Re:Pretty irrelevant

[tshak]

I've encountered software with licenses of as much as $8000 per processor, which obviously places a very large premium on having the fastest possible processor running it.

We aren't comparing Xeons or even IBM's latest supercomputer. All these tests show is, "hey look, I have the fastest consumer level chip on the block regardless of cost". How does a $550 chip's speed have any relevance in a ~$175 chip market? It's like comparing a $2000 3D graphics board in a SGI workstation to a Geforce 3 ti500.

Re:Pretty irrelevant

[Idaho]

Why anyone would spend the extra $350 on a P4 for the minimal performance gains (relative to the cost) is beyond me.

Well, minimal performance gains IF THAT!

It turns out the Athlon was faster in many tests. When you use RDRAM on the Intel platform, as you really should, they will probably be a bit faster, but not even in all benchmarks (we have seen these tests too, maybe not with exactly this clockspeed but the idea is the same).

So, reasons I can think of are Stability (not really an issue as long as you stay away from VIA chipsets as much as possible), Brand name recognition, and the law of Highest Number Must Be Fastest.

It's indeed real strange to see people buying P4's that will be beaten hands-down by AMD systems that cost about half as much. When will people finally learn? Much as I hate the AMD '2200+' (but actually really 1.67 Ghz) marketing ploy, they finally have a clue how they should market their products!

But as long as they don't put the highest number on their products, they probably won't win this battle against Intel.

DDR only doesn't make much sense.

[macinslak]

Given that the P4 costs more than twice as much as the Athlon ($548 versus $263 on PriceWatch), why would they bother with only DDR? Just by including the P4 they've pretty much thrown price/performance ratios out the window anyway.

A better question to ask of the P4 might be whether it could beat the Athlon with any kind of memory, and if so, by how much?

New Intel Processor

[Mignon]

In other news, Intel released their Pentium "Morningwood" processor which is optimized for displaying porn.

Just killing time while my program compiles and the site becomes available again.

Re:New Intel Processor

[og_sh0x]

Originally they were going to call it the "Peter Northwood"

RTFM!

[meggito]

Read the damn faq once and shut the hell up. I see this quesion so many damn times. This is EXACTLY what the fact says: Slashdot should cache pages to prevent the Slashdot Effect!

Sure, it's a great idea, but it has a lot of implications. For example, commercial sites rely on their banner ads to generate revenue. If I cache one of their pages, this will mess with their statistics, and mess with their banner ads. In other words, this will piss them off.

Of course, most of the time, the commercial sites that actually have income from banner ads easily withstand the Slashdot Effect. So perhaps we could draw the line at sites that don't have ads. They are, after all, much more likely to buckle under the pressure of all those unexpected hits. But what happens if I cache the site, and they update themselves? Once again, I'm transmitting data that I shouldn't be, only this time my cache is out of date!

I could try asking permission, but do you want to wait 6 hours for a cool breaking story while we wait for permission to link someone?

So the quick answer is: "Sure, caching would be neat." It would make things a lot easier when servers go down, but it's a complicated issue that would need to be thought through in great detail before being implemented.

Re:RTFM!

[baptiste]

commercial sites rely on their banner ads to generate revenue

Um, yeah and if you cache the page itself with all teh links back to their advertising provider, they'll still get all teh ad views and clicks they would have if you viewed their page on their server. The only thing a /. cache screws up is the stats for the content page itself. Seems to me you cache the HTML page linked to - it will still point back to ad banners and stuff. Sure folks use relative addressing - easily fixed with teh cache script - insert absolute URL for links/images without it, etc.

Re:RTFM!

[ColdGrits]

You mean like how Google cannot cache for the same reasons... oh, wait, it does cache.

Case dismissed, let the Slashdot cache commence.

Re:RTFM!

[Howie]

but do you want to wait 6 hours for a cool breaking story while we wait for permission to link someone?

No, I'd rather wait 6 hours after the story has been posted for the site to settle down enough that it actually works. :)

Re:RTFM!

[DaEvOsH]

Thanks for the excellent answer, SAVE for the first paragraph - loose the attitude, and just remember me to RTFM. I have read the FAQ, but that was just a few years ago. Thanks to you and all who replied with their thoughts.

Re:It always cracks me up...

[invenustus]

It's amazing that my little K62-350 stood up to the /. effect on 09-11-01 with no problem. Yet a fairly major web site can't handle it.

Um, maybe it was explained in another thread.... what was your computer doing on September 11th?

How does this benchmark really weigh in?

[KenSentMe]

With processing power up at the level it is today, how much of a difference in these two chips' performances is really notable? Even if the Athlon "XP" outperformed the intel (as I am told is often), it probably wouldn't be much, at least not enough worth talking about, and it sure wouldn't negate the fact that there have been several AMD "bugs" which notably inflict Linux users. They are, however, much cheaper. So I guess for me, the comparison isn't "price/performance", it's "price/functionality". Just my .02

Re:How does this benchmark really weigh in?

[glwtta]

several AMD "bugs" which notably inflict Linux users

I've heard of one, and it affected both Linux and Windows, what are the others?

Just wondering what I should watch out for...

Re:How does this benchmark really weigh in?

[s390]

Even if the Athlon "XP" outperformed the intel (as I am told is often), it probably wouldn't be much, at least not enough worth talking about...

How much faster is "worth talking about" huh? The Athlon XP 2000+ beat the Intel Northwood 2.2Ghz by 45% on the integer benchmark in this comparison!

Re:How does this benchmark really weigh in?

[peter]

Distributed.net's RC5 client is _not_ and integer benchmark. It was not written to give any kind of useful comparison of different CPUs. The Athlon RC5 core is hand-tuned specially for that CPU. There is no hand-tuned core for the P4, which is one reason it does so poorly. Search the rc5 mailing list: This has already been discussed. If you want an integer benchmark, stick with SPEC marks.

Non-slashdotted review of 2000+ at anandtech.com

[Anonymous Coward]

http://www.anandtech.com/cpu/showdoc.html?i=1574

http://www.anandtech.com/cpu/showdoc.html?i=1574

Google Cache ?

[Quazion]

Doesnt google cache site's also ? dont they get in trouble then ?

Quazion.

Re:Google Cache ?

[fishebulb]

but they also have the actual link as the prominent address. the cached link is below that, and they also have it display so the user understands its cached. something like this could be done on slashdot?

Re:It always cracks me up...

[ender-]

Um, maybe it was explained in another thread.... what was your computer doing on September 11th?

It was mirroring every bit of news/pic/vid I could get my hands on since all the news sites at the time were useless.

Ender

Use common sense people...

[Anonymous Coward]

You KNEW the site would get slashdotted, so there was NO reason to post a quick summary in the followup about the test results. Instead we've got people sitting here wondering whats up, and probably a lot of angry people over at linuxworld.com or where ever. Very POOR judgement.

Finally, a review NOT by Tom's Hardware

[kochsr]

It's nice to see a review that is NOT by tom's hardware on slashdot... nice to see a little variation in the works

Re:Finally, a review NOT by Tom's Hardware

[pctainto]

yea, but, once again, Tom's had the article on this out 2 weeks ago, so, therefore, tom's IS still the best.

verses??

[cswiii]

...the Pentium 4 verses Athlon XP war...

'Pentium 4 verses'? Are they anything like Spam Poetry?

It's 'versus', Mr. Editor Sir.

Re:verses??

[Mantorp]

vs. for the lazy

Re:verses??

[Figaro]

It's a little more like "Satanic Verses".

Re:verses??

[linzeal]

More like the satanic verses.

Re:verses??

[Dirtside]

More like the Satanic Verses. *rimshot* Thank you! I'll be here all week, unfortunately.

Re:verses??

[cswiii]

Not particularly.

The editor shoulda caught that one, and either fixed it, or marked it [sic].

Who cares?

[Anonymous Coward]

They all work faster than my P-150. They're all fast.

RDRAM, DDR RAM, and chipsets

[kenneth_martens]

Why the hell does using DDR memory even the results? The Pentium 4 was designed for RDRAM for God's sake.

Not entirely true. Although at first the only motherboards you could get for the Pentium 4 supported only RDRAM, but now Intel has released chipsets that support DDR RAM as well. You can check out the chipsets on Intel's website: the 850 chipset supports RDRAM, and the 845 chipset supports DDR RAM and PC133 RAM.

Same Old; Same Old

[Renraku]

As usual, a lower MHz rating on an Athlon easily competes with the higher MHz on a Pentium. Maybe if Intel starts renaming their processers, starting with maybe the 1800- and 2000- series.

At least they didn't...

[Wrexs0ul]

They should even things out even better and keep those power-guzzling volcano fans away from the Athlons. Then you can compare both in environments they weren't designed for :)

-Wrexsoul

Complicated?

[ergo98]

Overly complicated? I suppose for the slow it might be, but it seems incredibly simple to me. a) Set price point(s) b) Determine the purpose (Linux workstation, gaming, Office benchmarks, etc.) c) Develop a benchmark plan. d) Use your credibility as a hardware guru site to pick the best hardware that fits within each price point, for the applicable systems, and for the task set out. Document your choices so that critics (which there always are) can bitch that you didn't use the KT266a motherboard, etc. Earn your keep as a hardware site by making good choices. e) Benchmark. The reality (and home users) is that businesses don't say "I need a computer to do workstation tasks and I'm paying $1300 if it's an AMD system and $1600 if it's an Intel system!" : They buy based upon what they need to do, and the budget they have, and the performance that they get at that budget, otherwise there'd be a lot more options than just AMD and Intel there. Where's the IBM mini-computer options?

Re:Complicated?

[Chris Burke]

Well, to a first order you are right, except that those decisions involving hardware (eg DDR being cheaper than RDRAM, so the AMD system gets more ram, or a faster HD) will always be open to dispute, and then the results invalid. The question of which platform is better becomes obfuscated.

Another point is that while most everyone has a budget, those budgets are not entirely fixed. As in, you might want to only spend $1200, but if by spending $1300 you get something noticeable for your money, you might do it. Similarly, if you can get pretty much what you want for $1100, you might save the $100 for the beer fund.

I think the best thing to do is to compare a wide range of systems, and include the system price as one of the "benchmarks". I do -not- think that major changes to the hardware configuration should be made to account for more available cash. Things like the disks, the video cards, etc should remain constant or otherwise the relevant comparisons start to become meaningless. However, this would still give you a way to see how the motherboard, chipset, ram, and processor (the things that substantially vary between an Intel vs AMD setup) affects the price/performance, and let you make a reasonable decision as to what you want to do with your computer/beer money.
But at the same time, I still want to see who is the fastest, and that means throwing more or less equivalent systems together, but not throttling either based on price.

Tom Hardware Link

[leastsquares]

Based on the title, Tom's Hardware have a very similar review.

No bandwidth problems here.

Results

[Thorin_]

When posting about hardware comparisons please summarize the results in the front page post. Most of us don't care about the little details. Just say AMD won, Intel won or tie. Then I won't have to go reading through the 20 pages of tests.

Re:Results

[glwtta]

Hmm... it was a tie at 1.5 vs 2.0 GHz with both processors winning in the areas where they always win, what do you think it's gonna be at 1.66 vs 2.2?

[OT] Streamsicle

[TotallyUseless]

Followed your sig's link, and wow. Nice streaming server. You should consider getting it listed at VersionTracker. I spent a good day looking for a decent mp3 streamer that would run on OSX. After giving up in frustration, I ran across your link, and was saved.

Re:[OT] Streamsicle

[Matt2000]

Hey, thanks for the compliment, that's all we're hoping for with this thing - people who don't want to bother setting up insane mod_mp3 modules or shoutcast and who want a web interface.

Thanks for the tip on versiontracker too, we just got listed on Download.com and our downloads are through the roof, so maybe a versiontracker listing will really put us over the top.

Upper Headroom?

[4of12]

How do the next few months look in terms of the ability of either Intel or AMD to improve upon these products?

While I'm a fan of AMD's price/performance ratios, it looks as if they will be hard pressed to keep increasing the clock on the Athlon, while the Pentium 4 seems to have a lot more potential for higher clock rates.

Then, too, I'm wondering about the news reports that suggest that Athlons won't be paired up with the new DDR 333 MHz memory.

It may mean that the highest performance x86 architecture this summer will be from Intel and will be able to command more of a premium in price than if AMD were breathing down their necks, which has been the case over the past year and a half.

Re:Upper Headroom?

[Anonymous Coward]

Yes, but AMD has the appaloosa in their pipeline.

> It may mean that the highest performance x86 architecture this summer will be from Intel and will be able to command more of a premium in price than if AMD were breathing down their necks,

I think so...

> which has been the case over the past year and a half

Actually, for the last 3 years, since August 17th 1998 when the 500-650 mhz Athlon came out :D

Re:Upper Headroom?

[Sokie]

Actually Appaloosa is the next core in the Duron line of CPU's from AMD. The next Athlon core, due in the 1st half of this year (so by the end of June) is Thoroughbred, which will be a 0.13 micron process, basically a die shrink from 0.18 micron for the current crop of Athlon's. This will allow AMD to again ramp up clock speeds because of the reduction in heat/power consumption. Appaloosa is also a 0.13 micron process btw.

The roadmap further indicates that in the second half of this year, another incarnation of Athlon will appear using a 0.13 micron Silicon-On-Insulator (SOI) process.

AMD Processor Roadmap

-Sokie

Re:Upper Headroom?

[steveha]

How do the next few months look in terms of the ability of either Intel or AMD to improve upon these products?

The Pentium 4 is now being made with a .13 micron process, vs. the .18 micron process still being used to make the Athlon. When AMD starts selling Athlon chips made with the .13 micron process, they will once again be beating the Pentium 4 in benchmarks.

That is, assuming AMD gets the .13 micron process version out in a reasonable amount of time. If AMD takes too long getting their act together, Intel may be able to push the Pentium 4 to some ridiculous level (5 GHz or something) and win.

However, the Athlon will continue to rule in price/performance. Those of us who pay for their own computers will likely keep buying Athlons.

The latest I have seen on the AMD website is that the .13 micron chips should be out in "1H 2002", which presumably means June this year. That shouldn't be too late by any means; I doubt that Intel can do much with the Pentium 4 by that time.

P.S. Who among us really needs more performance than current Athlons? Even when Doom III comes out, current CPUs will have adequate performance (it's the 3D graphics card you will need to drop money on, not your CPU). I'm looking forward to buying a .13 micron Duron chip, to use in low-heat/low-noise computers.

steveha

OK, I'll bite...

[S. Allen]

Did you do it? How well does it work? I was thinking of doing the same thing, but thought better of it since it would be a major fn() hassle if it didn't work out.

Re:OK, I'll bite...

[Ashran]

I'm running a server with 2 Athlon XP 1700+ and havent noticed any problems yet.
I'm using some Tyan MB, but too lazy to look up which ;)

Re:OK, I'll bite...

[NumberSyx]

I don't know about the Athlon XP, but I am running a pair of older 1.2 Ghz (266 FSB) Athlons on my Tyan board and they work just fine. The board detects them and Linux uses them perfectly well. Redhat 7.2 detected them and installed the SMP kernel, though later I recompiled my own kernel to optimize it for Athlon processors.

web sites need to start caching more

[hanulec]

why is it that linuxhardware.com is unaccessable right now? did the server crash? or is its perl script too busy rebuilding the page every time it is requested. web sites need to start caching more! use a apache http accel cache.. anything!

It could be close...

[Chazmati]

Think about that.

Ten nodes with $300 Athlon 2000+ and $1500 Gb NIC = $18,000.

Nine nodes with $600 P4 2.2 GHz and $1500 Gb NIC = $18,900.

Where is that savings?

Re:It could be close...

[crisco]

Are those $1500 Gb cards quite a bit different from the $35 Gb cards (copper) or the $175 cards (fiber) listed on Pricewatch?

Re:It could be close...

[sweetooth]

Gee, you get one less node in your P4 cluster and still spend $900 dollars more... I'd say that shows savings right there.

Ten nodes with $300 Athlon 2000+ and $1500 Gb NIC = $18,000.

Ten nodes with $600 P4 2.2 GHz and $1500 Gb NIC = $21,000.

Re:It could be close...

[nosferatu-man]

Considering that the original poster was refering to myrinet and not
gigabit ethernet, I'd hazard a "yes" there.

Peace,
(jfb)

Re:It could be close...

[autocracy]

And mobo, and RAM, and blah blah blah. Take into consideration something more relevent: Athlon's have better floating point operations - something that would be more important in most cluster-utilizing apps. Getting that for less cost makes it worth it. So yes, you're right to a point - but Gigabit nics can come for much less than $1500 and you purchase more than just that and a processor...

Results

[Derkec]

For those of you wondering what the results are, here's the summary of what Tom's did, I still can't get to LinuxHardware

A comparison of the two top products from AMD and Intel reveals the astonishing: although the processors are as different from one another as apples and oranges, the difference is much less obvious in the benchmark results, when taken from an absolute standpoint.

In any case, one thing is visible: in the majority of performance tests, the new Pentium 4/2200 is ahead. After all, the top AMD processor has to make do with 1666 MHz, while its archenemy steps in with 2200 MHz. A closer look at the comprehensive benchmarks reveals that in Office performance as well as Linux Kernel compiling, the Athlon XP still takes the lead, despite its 32% clock speed disadvantage!

Ultraslow? What are you & the moderators smoki

[Kjella]

ultra-slow MP processors

Hello, the fastest MP processor is 4% slower (1600 compared to 1667MHz), probably due to the extra stability they want there as the server marked will drop anything unstable faster than lightning. Stick with the facts (they're expensive and don't add any real value add-on beyond certification) and don't FUD.

Kjella

question

[jon_c]

While reading through the results an idea came to me. Is it possible that the reason the P4 generally does better on 'one algorithm type of tasks is because its long pipeline wouldn't get busted as much, meaning that the branch prediction worked, which is based of past branch statistics (right?).

This makes since to me actually, in speaking with my friend about this last night I was asked 'well, what do you need a fast CPU for, when does it matter?', I replied 'Well, games, anything to do with multimedia, like Photoshop effects, ray tracing, mpeg encoding, but ya for general use, the CPU doesn't as much'.

but wait, lets look at that list in how it relates to the pipeline idea:
games: probably a good deal is going on here, AI, 3d pipelines, IO, networking probably not something a branch predictor would excel at

Photoshop effects ray tracing, mpeg encoding: all relatively contained algorithm that (if I'm right) would work well with the brand prediction.

So actually maybe having such long pipelines isn't that bad of a thing, because the majority of your day to day would doesn't care that much anyway, and most of the time when you need something as fast as possible its a small repetitive algorithm that could be predicted.

no?

Re:question

[jon_c]

wow, 3 hours of sleep + sloppy spell checking = embarrassing post.

my bad.

Re:question

[glwtta]

I would love to believe that the long pipeline was an engineering decision, but for some reason I can't shake off the feeling that it's just:

Long Pipeline == High Clock == Layman Buy!

And hey, it worked.

(sure call it flamebait if it so seems to you, it is nonetheless what I think; not that I am original in any way about it)

Re:question

[pclminion]

but wait, lets look at that list in how it relates to the pipeline idea: games: probably a good deal is going on here, AI, 3d pipelines, IO, networking probably not something a branch predictor would excel at

Actually, most of what you mentioned do involve small, tight loops over many iterations -- perfect fodder for a branch predictor.

Long pipelines do have drawbacks. A stall on a long pipeline, especially if it keeps happening over and over, can really hurt your performance. I'm more excited about this "hyperthreading" stuff, personally. I think CPU makers should start putting more resources toward these new ideas instead of just extending and re-extending the old. Longer pipelines and more cache can help to a point, but I think there's a barrier that needs to be broken through. Once manufacturers start to do it, I think we'll start seeing all kinds of scary-cool CPUs.

Re:question

[jon_c]

right, but those together and in different threads, like: T1: AI task/game logic, T2: vector math/rendering pipline: T3: network polling, switching between those 3 constantly. i would think means that the pipeline gets screwed. but maybe the OS and compiler we're smart enough to somehow know to do X amount of each so the pipeline wouldn't getting flushed all the time, i'm not sure.

anyway i find the suject interesting and would like to know more.

-Jon

Re:question

[pclminion]

When you say "switching constantly" you are talking about simple threading. In this case, pipeline length does not matter much because the time it takes to finish a pipeline's worth of instructions is much shorter than the time between task switches. The biggest problem for this kind of heavy threading is keeping all of the appropriate data in cache. It's tricky, but not impossible.

A hyperthreaded processor actually has multiple instruction streams. Each of these has its own pipeline. An instruction executes whenever the necessary chip circuits become available. This isn't quite the same as having two processors, because the sum of the clock rates (roughly the same as instruction throughput) for the two hyperthreads must add up to at most the chip clock rate -- perhaps less if stalls and cache misses are common. It's an elegant idea, but you could probably spend much less money and get far more processing power by simply buying two processors.

OT, cool sig

[pyros]

I've been Ippy Dippied to death.

But Sir, DDR-SDRAM chips DO perform

[amccall]

Uhhhhhh errrrrr Ummmm..... Benchmarks show otherwise. A SiS 645 chipset running PC2700 DDR-SDRAM will in most cases beat a P4 running PC800 RDRAM.

Not to trusting of Tom's Hardware? Have another set of benchmarks.

Stop spreading FUD about DDR chipsets and do a bit of research first. Any set of benchmarks I've seen has shown Intel's i850 w/ RDRAM and SiS 645 chipset in a dead heat - and most of the time SiS comes out on top.

Re:But Sir, DDR-SDRAM chips DO perform

[Henry V .009]

Stop spreading FUD about DDR chipsets and do a bit of research first. Any set of benchmarks I've seen has shown Intel's i850 w/ RDRAM and SiS 645 chipset in a dead heat - and most of the time SiS comes out on top.

You notice how it was pointed out that SiS 645 is overclocked as a factory default on your link?

Re:But Sir, DDR-SDRAM chips DO perform

[amccall]

You mean the OCing default with the SiS chipset(13Mhz) compared to that done in the Intel i845(17Mhz). Or only 5Mhz over the i850 chipset, which is itself overclocked by a entire 8 Mhz.

This is LESS than a 1% difference from the default clock. Given that the Intel chipset does the same with both its RDRAM and DDR-SDRAM configurations, I really don't see your point.

Further, if you compensate for the adjusted clock speed, I think you'll find that the results are still very similar. (Unless my math is wrong, SiS still comes out ahead on the first Q3 benchmark.)

He's talking about net performance, though

[Chazmati]

Agurkan was saying (I think) that if you design a cluster with a certain Gflop budget, you can achieve the same net performance with approximately 10% less P4 nodes if those P4 nodes are 10% faster.

So maybe there's other specialty cluster hardware that drives up the per-box cost, but if you start talking about dual-cpu boxes, it gets even worse...

Ten nodes with $600 (2 x Athlon 2000+) and $1500 Gb NIC = $21,000
Nine nodes with $1200 (2 x P4 2.2 GHz) and $1500 Gb NIC = $24,300

for comparable net performance.

I guess if you're trying to break a speed record, a 10% speed gain might be important even if the price doubles, but if you have to be on the 'bleeding edge' then you're probably not concerned with $$$ anyway.

Same level should be the two at prime level

[GdoL]

It should be fair if if both processors were to use their optimum RAM types,the one they were initially designed for? The Pentium 4 was designed for RDRAM. Now they are designed for DDR rather than RDRAM.

The Athlon as a fist advantage because it is more high IPC than the P4. Memory access is certainly a factor in this.

The P4 wins ordinary benchmarks because it can use the higher bandwidth RDRAM. Other benchmarks show otherwise. See tomshardware.com.

It should be more of more interest a price/performance ratio, for the consumer point of view.

Thinking that this kind of processors will be of more use for clusters, the price/performance ratio is an issue.

Re:Same level should be the two at prime level

[glwtta]

Isn't there very little difference in the performance of the P4 between RDRAM and the new DDR chipsets? At least that's the impression I got from reading a few reviews (not that I've dedicated much time to this).

I don't know, I still would not even consider buying a Rambus based system - but I guess this whole righteous outrage thing lasted longer with me than most others.

Linux Hardware...

[Mr. Piccolo]

obviously needs more hardware. Dead already.

I knew it.

[MisterBlister]

I just knew it.

What about stability?

[azrix]

I've not read the article because it's still /.'ed, but...

I was wondering about the stability of P4 vs. Athlon platforms. The awsome overclock-ability of the P4's has gotten me very interested in them, so I've been reading up on the recent reviews. Reading them carefully, I've caught a few interesting lines about the Athlon systems locking up some when running benchmarks, but the P4 system running like a rock. Makes me wonder if a P4 wouldn't be a better choice when you want performance and stability, with the Athlon being a better choice when you want performance and low cost. While this doesn't sound like a big deal, it is if you're like me and like to leave your computer running 24/7.

Just wondering if anyone had any comments on this as it's not something that I've seen anybody mention anyplace and it seems very relevent.

Re:What about stability?

[glwtta]

Hmm... I seem to recall a few places mentioning some stability issues with the P4, sorry I couldn't be more specific, but just saying I didn't get the same "like a rock" impression.

On the other hand, I've heard plenty of bitching about Athlon stability, all I can offer here is my personal experience - during the year or so that I've had this system (T'Bird 1.2 + ABIT KT7) it's locked up twice - once when I naively tried to get my Vortex 2 working (I run Linux) and the other was one of my Maxtors acting up with the Promise controller - seems to be pretty common too.

In other words, in a year of heavy use - pretty much on during the day, and frequently left on over nights - I've never heard a peep of instability from anything CPU/Platform related. Maybe I got lucky, maybe the good RAM makes a difference (not like you have to pay for it nowadays), but that's been my experience.

Re:What about stability?

[azrix]

Yeah. I may have overstated the "like a rock" part. Plus, I was refering to most of the benchmarks that I've read, which seem to pit Intel P4's w/ Intel chipsets against AMD Athlon's w/ Via chipsets. I think it may be a Via issue more than anything, but I thought it was worth mentioning.

And just to prove I'm not making this up, heres what Tom's Hardware had to say in this article:

Another factor is the stability and product quality of a system: while all Athlon processors suffered from occasional instability in our tests, the Pentium 4 platform ran without a glitch. Reasons for this behaviour might not lie in the processor itself, but rather in the motherboard design and the chipset used.

So, there you go.

Re:Pentium 4

[Mr. Piccolo]

Oh really? Let's clock the P4 at 3 GHz (it's been done with water cooling) and the Athlon XP at 1 Hz. Now who's going to win???

It's worse than that...

[Rimbo]

It's not just that DDR is a mature tech for AMD but not so much for Intel. It's that the two memory types work very differently, and how the processor expects data to arrive is going to be affected by the memory. P4's are designed to expect the higher bandwidth and latency of Rambus -- they expect to wait longer, and to get more in one chunk.

So no, I think that this would skew results in AMD's favor. I actually think that a fair comparison would require Rambus on the P4 system, and DDR on the AMD system.

Re:It's worse than that...

[Skuld-Chan]

I thought thats what the northbridge chip did on the K7 series - IE abstracted that interface to allow you to run at any fsb on one side and use any other technology memory on the other side. I thought it was originally developed by dec to reduce the cost of expensive memory (by not having to use it) in axp machines.

Re:It's worse than that...

[javiercero]

Actually Rambus acts in the opposite way, it has very little latency but a smaller data width per transfer (narrower). So the P4 should spect small pieces of data very fast, whereas the athlon will spect large pieces of data in a slower fashion. But this is pointless because claiming that the P4 is designed with RAMBUS in mind is nonsense...

When are the .13 micron Athlons coming?

[glwtta]

Is there any sort of realistic forecast? And I mean not just the chips, but good mobos to go with them.

Was hoping to upgrade this summer...

Oh and so we are on topic: it sure is interesting about all them Athlon vs. P4 comparisons!

Re:When are the .13 micron Athlons coming?

[yobbo]

Expect them to be out around April.

Re:Let's try and grow a brain next time

[s0l0m0n]

How exactly do you do that? Either 'grow a brain' or test two disimilar pieces of hardware to the same standard.

Comparing the P4 directly to the athalon would be well, weighing them, measuring them, etc. The motherboard is one of the requisite tools for a processor test. Trying to compare them directly just doesn't work.

maybe you should go back to playing with your kernal, instead of ranting about something you have no authority with?

Okay. Kernel compiles may be bitchy, but not nearly so annoying.

What Was Tom Smoking?

[Glonk]

That article always left a funny taste in my mouth.

Why was he comparing next-gen DDR (DDR333), which isn't officially out yet, to the OLD PC800 RDRAM? Wouldn't it make more sense to compare PC1066 RDRAM (see the AcesHardware benchmarks)?

PC1066 RDRAM and DDR333 will both come out officially around the same time in official chipset support.

In other words, next-gen DDR performance for the P4 is about 1.5 years behind the RDRAM performance. Tom didn't mention that part...

In other news, Samsung is sampling PC1200 RDRAM now, too. 4.8GB/s in a dual channel config.

you people are missing the point

[john_uy]

most of the sites that i have seen benchmark the two processors head to head using generic applications.

however, intel does not concentrate on applications that do not require more cpu power like office apps. who cares if i run office app with benchmarks slightly slower? as if when i type i eat up the cpu. i mean i bet it will be idle 99% (and one percent for other system processes)

intel is now concentrating on multimedia (which shows their lead over amd most of the time.) the good thing about intel is that they are concentrating on new technologies such as the sse2 and hyperthreading.

on the sse2 side, the performance of the pentium 4 2.2ghz beats athlon by as much as 64% in specially programmed applications like lightwave 7. and more than that if you enable hyperthreading, you can get a rough gain of 30% on certain applications including server apps. so if your program has been properly programed and compiled you can have gains of around 100% over the fastest amd.

and here is the catch. when you buy a pentium 4 2.2ghz, most people will overclock it to 3ghz. so that is worth the high price. and amd won't be having any athlon xp 3000+ soon.

intel has very good fabrication technology in microchips (not just processors.) they have upgraded their fab plants to 0.13 micron and 300mm wafer. so actually the hefty price of intel cpu gains them more profit than amd (and in the latest financial results, intel has earned $1.8b compared to amd loss of $61m. intel's gains would be bigger if not for their network and wireless division's loss)

intel has many cards to show. i mean a processor having roughly 55m transistors in it (around 4.5 times more than amd) what are they all doing in there?

just wait and see. intel will slowly activate more features in pentium 4 that are already built in to it or things that in can handle. like release a pentium 4 with 133fsb (over the current 100mhz) will surely make it fly.

by the way, i will report to you guys after we get our engineering sample of pentium 4 (without any multiplier locks!!!!!) let's see how fast the pentium 4 is over amd. note than pentium 4 scales very well.

Nope.

[himi]

The problem with this analogy is that it looks at the wrong thing. The performance of a CPU isn't a matter of how quickly it can do one thing (ie, how fast it can finish a race), it's a matter of how much it can get done in a given period.

A much better analogy is water flowing through a pipe. You want to get some volume of water out the far end, so you have two choices: you can pump the water through a thin pipe really quickly (the P4), or you can pump the water at a slower speed, but through a much fatter pipe (the K7).

The K7 core can retire, on average, about 1.6 instructions per clock cycle. At 1.667GHz, that means that your XP1900+ can complete about 2.667 billion instructions per second.

Now, assuming the equivalent performance on benchmarks and the like indicates that the P4 2000 can complete about the same number of instructions (since it gets the same results at that clock speed (this is a /big/ assumption, but for the purposes of this argument it's reasonable)), that means the P4 is retiring about 1.33 instructions per clock cycle. (Note that the 1.6ipc figure for the K7 is from an old RealWorldTech article by Paul DeMone that I can't find right now - it's somewhere in the Silicon Insider archives . . . )

/That/ is the fundamental difference between the two chips: the K7 completes more instructions every time it's clock ticks. That's what people talk about when they talk about "brainiac" versus "speed demon" processors: the P4 gets it's performance because it completes lots and lots of clock cycles in a given period; the K7 gets it's performance because it does a lot in every clock cycle, even though it completes fewer cycles in the same period.

CPUs aren't a horse race - they're a production line, where what matters isn't how fast an individual thing is done, but how many things get done in a given time period.

himi

Re:Nope.

[DemiKnute]

Aye, that's what I said. Perhaps I wasn't verbose enough.

The Athlon/draft horse is still correct. It pulls a hell of a lot, just not as far as fast.
The P4, on the other hand, is blurringly fast, but it doesn't pull as much.

It's a handicapped horse race. The Athlon has the larger handicap (although in this case it works out to be a good thing), in that it does more every clock cycle. The P4 is going faster, but its handicap isn't as large, so it's not _actually_ doing as much.

More of an indictment of the i845

[mpsmps]

Although the review says that using the i845 chipset with DDR memory gives "a nice level playing field when it comes to pure processor verses processor benchmarking," the i845 chipset is actually known to be a dog and the Pentium 4 performs much better with other DDR chipsets. The LinuxHardware review may be a valid comparison of an all Intel solution against a multi-vendor AMD solution, but its not a reasonable "processor versus processor" benchmark until they pair both processors with their best DDR chipsets and motherboards.

Where's the results then??

[egriebel]

Where's a summary of the results for this? Wouldn't it have been nice so I don't have to click thru? Not that I'm lazy or anything :-)

Memory Bandwidth

[BurkeChowdah]

There is another clock speed issue to examine as well when comparing these technologies. There has been frequent mention of comparing RDRAM and P4 with Athlon with DDR, for various reasons. One of these is that the RDRAM is "faster" and more mature, so therefore will "level the playing field." The problem with this is the addiction to the clock speed rating. PC800 RDRAM does indeed run at 800 Mhz, but it only runs at a 4 bit bus width. DDR on the other hand, only runs at 266 Mhz FSB, but runs on a 64 bit bus width. Because of this, RDRAM chokes when you are doing RAM intensive tasks, where the DDR has plenty of additional bandwidth.

Just my $.02
Ed

Not just for supercomputers

[HiThere]

Standard animation programs are quite expensive on a per processor basis. Once you get past blender, things get expensive FAST.

Actually, any specialized field has expensive options that real professionals need. Except programming. And that's because of the FSF. GNU/FSF tools allow programmers to get away from the hyper-expensive tools. But animators, circuit board designers, secretaries, architects, etc. can't write their own applications. KDE, OpenOffice, etc. are working on the problem for secretaries (there's a lot of them, and middle management generally uses the same toolset, plus a few). Blender exists for animators, but it isn't (doesn't appear to be?) up to professional use. Gimp is challenging Photoshop, but only Kontour is challenging Illustrator, and it needs a LOT of development. etc.

So per processor licenses effect a lot of folk. (Well, few word processors need multiple processors, but you know what I mean.)
.