Processor Speed being misreported

[bum]

It goes without saying that I need a new Heatsink and fan, though. Normally my CPU runs at about 51º C, which I'm a little apprehensive about. I'd like to get that down by maybe another 5 degrees if I can, coz I dont know what the crash temperature of a Palomino core (i think) AMD Athlon XP 1800+ is. But if its chunking on a 43 degree day, more cooling is in order, methinks.

53 degrees ? ok, that is kinda low. i had a freind who's xp2100 was pupming out at 90 and was still stable fpr a few minutes at least (the guy has no idea how to put a fan on prperly) . ok basicaly, from what ive haerd, ya shouldnt get stressed until ya start going 65+

[the Black Monarch]

I recently (3 weeks ago) ordered a water cooling system from SwifTech, and it still hasn't come. Bastards. I'm considering getting a Prometeia Mach II vapor-compression phase-change cooling unit and regassing it with r404a, giving me maybe 250W of raw cooling power. That's f*cking COLD.

Oh and dw, I've been meaning to bitch at you about something... the Athlon 64 isn't a new architecture, you just took the old school Athlon and added some 64-bit code to the instruction sets (and added two pipeline stages and made a few minor tweaks). 8th-generation my ass, it's more like "seventh and a half" generation... Prescott is doing the same stuff, as if the P4 didn't already have enough pipeline stages... stupid 90nm process... you suck... uphill both ways... grumble grumble bitch bitch...

Basically the Athlon is more efficient than the P4 core and thus peforms better at lower frequencies

You once again conveniently left out the part about how they have to because they're unable to get to frequencies as high as the P4... Some guys used liquid nitrogen and overclocked a P4 to 5.25 GHz. Phase-change cooling had to be used on the fricking northbridge! For an Athlon to perform that well, it would have to attain 3.5 GHz, and so far the record is ~3.0 GHz. But the A64 3k+ ever so slightly beats a 3GHz P4 in most situations.

[dwchang]

AMDs official spin is that the performance ratings are based off of CPUs with the older Thunderbird core. So XP1800+ is equal to an AMD cpu with an 1.8Ghz Thunderbird core. Which the Thunderbird is no longer being made, being replaced by the palimino, thoroughbred and barton...

So how good these ratings are is kinda suspect and it's making my life difficult in finding the actuall clock rates of them :p.

Where did you hear this? I mean the original performance rating was based on that, but I assure you it changes with every revision. Now I don't work in that department and don't do the actual calculation, but I imagine they do comparisons and benchmarks.

Also if it were so inaccurate, why are the benchmarks on 3rd parties fairly accurate (i.e. a 1800+ ~ 1.8 Ghz P4 and so on)?

says some stupid stuff and tries to sound like he knows what he's talking about...*insert everyone laughing here*

You once again conveniently left out the part about how they have to because they're unable to get to frequencies as high as the P4... Some guys used liquid nitrogen and overclocked a P4 to 5.25 GHz. Phase-change cooling had to be used on the fricking northbridge! For an Athlon to perform that well, it would have to attain 3.5 GHz, and so far the record is ~3.0 GHz. But the A64 3k+ ever so slightly beats a 3GHz P4 in most situations.

I'll say a few words:

I have no idea what you're talking about in the first sentence...as for the second part, the best part about the overcloekd 5 Ghz P4 is that...an Opteron still beat it on *some* benchmarks. That in itself tells you exactly what I said about performance and speed.

And if the Athlon needed to be at 3.5 Ghz to perform well then...A) why are hundreds of thousands of people buying them and B) why do benchmarks and well anyone in the industry disagreeing with you. Maybe because you're wrong.

[the Black Monarch]

And if the Athlon needed to be at 3.5 Ghz to perform well then...A) why are hundreds of thousands of people buying them and B) why do benchmarks and well anyone in the industry disagreeing with you. Maybe because you're wrong.

Or maybe because I didn't say just "perform well," I said "perform as well as the P4 that got overclocked to 5.25 GHz"

And if you can point me to a non-overclocked Opteron that can beat a 5.25ghz P4, I'd REALLY like to hear about it.

Again, grumble grumble bitch bitch and the like.

[jonmartensen]

So you want a test where a non-overclocked AMD processor performs better than a highly overclocked Intel processor?

[the Black Monarch]

Not just highly overclocked, but THE most overclocked Intel chip EVER.

[dwchang]

And if the Athlon needed to be at 3.5 Ghz to perform well then...A) why are hundreds of thousands of people buying them and B) why do benchmarks and well anyone in the industry disagreeing with you. Maybe because you're wrong.

Or maybe because I didn't say just "perform well," I said "perform as well as the P4 that got overclocked to 5.25 GHz"

And if you can point me to a non-overclocked Opteron that can beat a 5.25ghz P4, I'd REALLY like to hear about it.

Again, grumble grumble bitch bitch and the like.

Main story on Slashdot found here:
http://slashdot.org/article.pl?sid=03/1 ... ad&tid=137

Most relevant comment found here:
http://slashdot.org/comments.pl?sid=91015&cid=7840687

More noteably, Tom's Hardware took down the benchmarks where the Athlon-64 beat the P4 at 5.25 Ghz. This, obviously, brought about the argument about how Tom's Hardware is a bunch of Intel fanboys, but that's something that has been known for years.

Regardless, it's pretty dumb to be comparing a Pentium nearly doubled by overclocking to something that isn't. In fact, the liquid nitrogen cooling alone would cost more than just buying a dual-Opteron system (which I imagine would perform well). You're basically saying "Oh yeah! Well if I get this abnormally rare cooling solution, I own your consumer-level desktop! So there!"

You also don't reply or respond to anything else that you make broad generalizationd and assumptions about. Oh wait, you always do that! . Get some credible information or a degree specializing in said field...maybe then you'll say something relevant and come up with your arguments based on YOUR knowledge and not someone else's.

[bum]

a p4 at 5.25ghz ? my god, seriosly. first of that thing would have required a shitload of cooling, including water cooling that uses nitrogen instead of water, a vantec cpu cooler case (which costs about $1000 itself, and its just a fansy case that cools the cpu, and nothing els) and it (the case) was probably filled with some ice cold liquid that resists electricity and isnt flamable, and whj knows what els. oh and, despite all that costing a shitload, id give that cpu a lifespan of 1month max. seriosly.

oh and, i ran a benchmark called testlab03 yesterday, and it turns out that an athlon1600xp actualy has the same FPU performance (floating point calculations, as in decimals, floading point is just a fansy pc tech word for them) as a p4 2.5ghz, which strangely enough my 2600xp beat the crap out of ([results]1600xp = 186 - p4 2.53 = 284 - 2600xp = 234)

[dwchang]

a p4 at 5.25ghz ? my god, seriosly. first of that thing would have required a shitload of cooling, including water cooling that uses nitrogen instead of water, a vantec cpu cooler case (which costs about $1000 itself, and its just a fansy case that cools the cpu, and nothing els) and it (the case) was probably filled with some ice cold liquid that resists electricity and isnt flamable, and whj knows what els. oh and, despite all that costing a shitload, id give that cpu a lifespan of 1month max. seriosly.

oh and, i ran a benchmark called testlab03 yesterday, and it turns out that an athlon1600xp actualy has the same FPU performance (floating point calculations, as in decimals, floading point is just a fansy pc tech word for them) as a p4 2.5ghz, which strangely enough my 2600xp beat the crap out of ([results]1600xp = 186 - p4 2.53 = 284 - 2600xp = 234)

I believe on the Tom's Hardware site, they list what the equipement they used. And yeah, it'd cost a pretty penny. It's pretty much a moot point to try and compare it to "normal" CPU's and cooling solutions (like TBM was trying to do).

I am actually a bit surprised that the FPU performance is that good, but still good news .

[the Black Monarch]

Regardless, it's pretty dumb to be comparing a Pentium nearly doubled by overclocking to something that isn't.

Weren't you the first of us to do that? I only compared the 5ghz P4 to an extremely overclocked (3.5ghz) Athlon FX... you were the one who brought stock Opterons into the equation.

In fact, the liquid nitrogen cooling alone would cost more than just buying a dual-Opteron system (which I imagine would perform well).

Yeah, a simple copper pipe with a bit of foam taped to it would cost three thousand dollars. Sure.

You're basically saying "Oh yeah! Well if I get this abnormally rare cooling solution, I own your consumer-level desktop! So there!"

No... I said "This Intel user's abnormally rare cooling solution owns this AMD user's equally rare cooling solution in most cases". Then you said "Oh, yeah? Well, my consumer-level desktop owns your abnormally rare cooling solution in one or two benchmarks!" to which I am now replying "WTF do you mean by 'my' abnormally rare cooling solution? I don't have any liquid nitrogen. I don't even overclock! WAAAAAH mommy!"

You also don't reply or respond to anything else that you make broad generalizationd and assumptions about. Oh wait, you always do that! . Get some credible information or a degree specializing in said field...maybe then you'll say something relevant and come up with your arguments based on YOUR knowledge and not someone else's.

Again I have to ask what the hell you're talking about.