Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

AMD Releases Athlon XP

By Ialdabaoth in MLP
Wed Oct 10, 2001 at 09:46:44 PM EST
Tags: Technology (all tags)

AMD's Athlon XP is out, and it looks shiny.

Advanced Micro Devices (AMD) unleashed the Athlon XP yesterday.

According to the propaganda at AMD's Athlon XP site, the Athlon XP outperforms the Pentium IV when equipped with DDR RAM despite running at a lower clock speed. AMD claims that this is possible because of their QuantiSpeed Architecture, which supposedly allows the Athlon XP to execute more instructions per clock cycle.

While I'm not about to rush out and trade my 1GHz Thunderbird for an Athlon XP, I do plan to buy one eventually. Furthermore, AMD's propaganda is relatively plausible to me, despite being a cynic and studying for A+ certification.


Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure


AMD or Intel
o I want my, I want my, I want my AMD! 36%
o Intel invented x86, and AMD's just an upstart. 1%
o I vote with my wallet, and Athlon's cheaper. 39%
o I use a Mac. 16%
o I use an Alpha. 1%
o I use a Cray (Luck bastard!!) 4%

Votes: 61
Results | Other Polls

Related Links
o Advanced Micro Devices (AMD)
o Athlon XP
o Also by Ialdabaoth

Display: Sort:
AMD Releases Athlon XP | 19 comments (14 topical, 5 editorial, 0 hidden)
Well, I'll give them one thing... (4.66 / 3) (#2)
by RareHeintz on Wed Oct 10, 2001 at 02:17:29 PM EST

...they're right on their money with their new pitch, which is that MHz is not the primary determinant of processor speed. It's an open question whether your average mouth-breathing consumer will ever get around to rating processors by something other than a single, scalar number, but I wish AMD luck in trying to induce that kind of market sea-change. It would be for the better if it worked.

As for the XP... Well, I guess I'll see what Tom's Hardware and Anandtech have to say. I know my next box will have an Athlon, not sure how hot I am for the XP yet.

- B
http://www.bradheintz.com/ - updated kind of daily

AMD Rock! (3.33 / 3) (#6)
by HereticMessiah on Wed Oct 10, 2001 at 04:11:51 PM EST

Why? They don't use the Mhz Myth, nor do they drastically overprice their hardware.

Disagree with me? Post a reply.
Think my post's poor or trolling? Rate me down.
Hah! (4.66 / 3) (#7)
by twodot72 on Wed Oct 10, 2001 at 04:23:43 PM EST

I don't remember them trying to debunk "the MHz myth" when they beat Intel to 1 GHz. I stongly doubt they're doing the right thing now because of sudden moral qualms, it simply does not work in their favor any more.

The Athlon definitely gives you more bang for the buck than the P4 though. I'm getting one for my next box.

[ Parent ]

The reason doesn't matter. (4.00 / 1) (#15)
by Ialdabaoth on Wed Oct 10, 2001 at 10:31:38 PM EST

The fact is, somebody's exposing the megahertz myth for the nonsense it is. If AMD's doing it for money, more power to them.

Reaching 1GHz was a big deal. Going to 2GHz is relatively trivial. If Intel were to belt out a 100GMz chip, or a 1 terahertz (THz?), then that would be news, and AMD might not get away with saying that clock speed isn't everything.
"Act upon thy thoughts shall be the whole of the Law."

--paraphrase of Aleister Crowley
[ Parent ]

They've never resorted to myths (3.00 / 1) (#16)
by HereticMessiah on Wed Oct 10, 2001 at 11:15:56 PM EST

And that's what's important.

Disagree with me? Post a reply.
Think my post's poor or trolling? Rate me down.
[ Parent ]
More Reviews (4.00 / 1) (#8)
by no carrier on Wed Oct 10, 2001 at 05:05:30 PM EST

There is a nice review here. The benchmarks are very good for this chip, although intel still has the fastest quake3arena server.

I stab people.
Why is that? (3.50 / 2) (#11)
by Elendale on Wed Oct 10, 2001 at 09:42:08 PM EST

although intel still has the fastest quake3arena server.

Not surprising, considering Intel probably built the P4 to specifically do well in Q3. For a while, your Q3 benchmark was basically the whole deal for proc speeds so it's only logical. Of course, the Athlon was being touted as superior to the P3 quite often because it was better in Q3 (Pentium 3 had more DirectX strength, however).


When free speech is outlawed, only criminals will complain.

[ Parent ]
Was it Mark Twain... (5.00 / 2) (#14)
by Ialdabaoth on Wed Oct 10, 2001 at 09:53:29 PM EST

Was it Mark Twain who first made the comment about lies, damned lies, and statistics? I wonder what he would have made of the average computing benchmark.

No doubt he'd have turn around and made a comment about lies, statistics, and benchmarks..

I wonder, though, what does Q3 stress more, integer math or floating point math?
"Act upon thy thoughts shall be the whole of the Law."

--paraphrase of Aleister Crowley
[ Parent ]

Indeed (4.00 / 1) (#17)
by Elendale on Thu Oct 11, 2001 at 02:21:28 AM EST

Any look at the Apple-sponsored benchmarks (certain Photoshop tests are all they compare with) would probably back that up. Still, Q3 benchmarks (and others) are often useful as real-world tests of processor speed, if nothing else: because most people who buy processors this rediculously fast are going to be playing Q3. In reality, the AXP at 1.533 Ghz appears about as "fast" as a P4 at 2Ghz. Not too bad, but we'll see what else comes into play. I know where my money would be going, were i to build a new computer today.
As to your question, Q3 stresses floating point more (IIRC) than math.


When free speech is outlawed, only criminals will complain.

[ Parent ]
Why "XP" ? (4.66 / 3) (#10)
by MugginsM on Wed Oct 10, 2001 at 09:37:49 PM EST

I really wish people would label their products distinctly.

It's bad enough searching for eXtreme Programming with that MicroSoft disaster clogging up the search
results. Now this.

Maybe it's just me. My next computer will likely
be PowerPC (of some kind) based anyway. I tested
the non-Intel(& clone) waters a little while back
by buying a Mac. I discovered that I don't need
to torture myself with the broken PC design any
more. AMD or Intel, it's all the same to me.

- MugginsM

Athlon XP. (4.00 / 1) (#13)
by Ialdabaoth on Wed Oct 10, 2001 at 09:48:30 PM EST

According to AMD's propaganda, XP is supposed to stand for e<u>X</u>treme <u>p</u>erformance. Their FAQ also claims that it will run Linux and FreeBSD, not just Windows. However, Athlon XP is supposed to be optimised for Windows XP.

Which is strange to me -- I'm using to optimising my code to fit my hardware, not optimising the hardware to fit my code.
"Act upon thy thoughts shall be the whole of the Law."

--paraphrase of Aleister Crowley
[ Parent ]

Benchmarking (5.00 / 1) (#18)
by twodot72 on Thu Oct 11, 2001 at 03:05:08 AM EST

Of course it works with Linux and FreeBSD, it's an "x86 compatible" chip, and furthermore the changes from the existing Athlon are small. That it would be optimized from Windows XP is likely just marketing fluff. To wit, Windows XP did not exist when the Athlon XP was being developed, making it very hard to do any performance profiling for it.

But yes, the CPU guys do use profiling of common applications to determine how to improve their designs.

[ Parent ]

Athlon XP (buzzword) Question... (3.50 / 2) (#12)
by 0x00 on Wed Oct 10, 2001 at 09:46:46 PM EST

Does the lookahead cache adversely affect the superscalar pipelined FP unit if there is a cache-miss; i.e. the pipeline has to throwaway existing units which have been prefetched into the superscalar structures? This appears to be a big problem with the P4, will it also occur in the Athlon XP?

I've read through most available information (to me) on the new Athlon XP but have been unable to find an answer.



clown. lol.

It seems not.... (5.00 / 2) (#19)
by evanbd on Thu Oct 11, 2001 at 11:19:41 AM EST

First off, there are no benchmarks to suggest that Thunderbird performs better than Palomino at the same clock. The Palomino does no additional speculative excution, only speculative fetch to cache. So there isn't anything extra in the pipes to throw away, ever, it's just that when it wants to start something all the stuff is there sooner (one hopes). The prefetch just tries to hide latencies to memory even more.

Also, the reason it is a problem with the P4 actually has very little to do with prefetching or any such. It has to do with the long pipeline -- with close to 20 instructions wandering through, if the pipeline is flushed, that's a big performance hit. The P4 needed those instructions to work in order for everything to go well. If you want good, technical info, I'd suggest aceshardware.com, especially their technical articles (not news blurbs; those are good, but the meat is in the articles) and the tech message board. They have really smart industry people posting there...

[ Parent ]

AMD Releases Athlon XP | 19 comments (14 topical, 5 editorial, 0 hidden)
Display: Sort:


All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!