Sign in with
Sign up | Sign in
Your question

Very strange Athlon review on THG

Last response: in CPUs
Share
August 24, 2002 4:04:31 PM

I don't know what's going with THG. But some ppl are obviously living in a strange world of where fiction become facts.

In the latest review about the new Athlon XP 2600+ they compare this AMD CPU with an old 2.53.

(a) the tested XP 2600 is not available for weeks, you cannot buy it.
(b) the now availbale and retailing top CPU from Intel are the 2.66 and 2.8. They are on sales since 1 week. Why didn't they test them?
(c) the tested 2.53 was the new one with the samller die? I doubt!
(d) why was the 2.53 not tested with the chipset which shows it's fully potential? Either a 850 or a 850E? Like driving a Ferrari with the the feet on the break?

Very poor, biased approach!

<P ID="edit"><FONT SIZE=-1><EM>Edited by chainbolt on 08/24/02 12:07 PM.</EM></FONT></P>
August 24, 2002 4:18:57 PM

Quote:
the now availbale and retailing top CPU from Intel are the 2.66 and 2.8. They are on sales since 1 week. Why didn't they test them?


2.66 and 2.8 GHz P4' aren't officially released yet.

Quote:
why was the 2.53 not tested with the chipset which shows it's fully potential? Either a 850 or a 850E?


THG tested P4's with i850E chipset and PC1066 RDRAM, currently the best platform for P4

Quote:
Very poor, biased approach!


Very wrong!
August 24, 2002 4:26:13 PM

The 2.6 and the 2.8 are RETAILING since one week, I have one in my box, as I write here and so do 100 of other ppl.
OC results are posted, they are MUCH better than this XP 2600+.

The THG review is a very bad piece of deception to be honest. How can test a CPU with a chipset which is known not
to utilize it's full potential? As I said, driving a Ferrari with the the foot on the break and then complaining about missing speed.
Related resources
August 24, 2002 4:31:11 PM

This is WRONG! The Pentium 2.6 and 2.8 HAVE already been released this week and they are selling in ASIA. I have one
and have already yesterday posted my results.

The THG tested XP 2600 is NOT selling, it's not availbale for weeks. Therefore this review is a preview, that's all, and a bad one, because it is comparing a CPU which is not yet available with an outdated Intel CPU on a slow performing board. What is the purpose if this other than deception?

You can buy a P4 2.66 or 2.8 at ANY retailer in Japan, which is the biggest PC makrte in the world. And who or what is newegg?

<P ID="edit"><FONT SIZE=-1><EM>Edited by chainbolt on 08/24/02 12:32 PM.</EM></FONT></P>
August 24, 2002 4:35:24 PM

as for your theory the only thing i see is that they didn't use intel latest CPU's. Maybe they didn't receive them at the time they made the article.


this is what they used for the intel p4 in the test.
"Asus P4T533-C (Intel 850E Chipset) Revision: 1.01
Bios: 1006 BETA 002 (07.08.2002)
2x 256 MB RDRAM, PC800, 533 MHz, 40ns, Infinion
2x 256 MB RDRAM, PC1066, 533 MHz, 32ns, Kingstone"

they used the 850E chipset which is the highest performing chipset out there. So i'm not getting you there.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A><P ID="edit"><FONT SIZE=-1><EM>Edited by xxsk8er101xx on 08/24/02 12:39 PM.</EM></FONT></P>
August 24, 2002 4:35:32 PM

this is WRONG! The Pentium 2.6 and 2.8 HAVE already been released this week and they are selling in ASIA. I have one
and have already yesterday posted my results.

The THG tested XP 2600 is NOT selling, it's not availbale for weeks. Therefore this review is a preview, that's all, and a bad one, because it is comparing a CPU which is not yet available with an outdated Intel CPU on a slow performing board. What is the purpose if this other than deception?

You can buy at ANY retailer in Japan, which is the biggest PC makrte in the world. And who or what is newegg?
August 24, 2002 4:39:42 PM

My bad, i did not see the 850E, although the highest performing platform the P4T533 with RIMM 4200-32 bit and not the P4T533-C with the older type PC1060, but I damit the difference is marginal and not relevant.

It remains that this preview is grossly misleading, beause they compared a not retailing XP 2600+ with an outdated, "old" P4 2.53. The new 2666 and 2800 with the smaller die are retailing/selling and they should have been compared.
August 24, 2002 4:45:15 PM

Post deleted by chainbolt
August 24, 2002 4:45:26 PM

? You are not the author of the review.They could have very easily obtained a P4 2.66 or 2.8. Intel is producing them since beginning of August, that is the production date of my 2.66. Instead they preview a XP 2600, which is announced to be available for OEM in September??
August 24, 2002 4:45:47 PM

? You are not the author of the review.They could have very easily obtained a P4 2.66 or 2.8. Intel is producing them since beginning of August, that is the production date of my 2.66. Instead they preview a XP 2600, which is announced to be available for OEM in September??
August 24, 2002 4:45:58 PM

Post deleted by chainbolt
August 24, 2002 4:55:56 PM

Why you are posting 6 same post in a row?
August 24, 2002 4:59:39 PM

He just beat AMD_Man's multiposting, and Crashman's too, lol!

--
And now, an advice from your friendly Nike shoes slogan: JUST DO HER!
August 24, 2002 5:14:34 PM

Tomshardware made a very goo review showing that even a 5999999 mghz K7 will not be able to beat P4 due to is slower FSB and VIA bottleneck

At the end i have speak with a horny lady
August 24, 2002 5:16:22 PM

Explain how the 2.53GHz is so old and outdated? What are you smoking man? It's no different than the new ones minus the larger multiplier.

"Opinions are like assholes, everyone has one"
August 24, 2002 5:19:46 PM

for those who are interested to see how the already retailing P4 2800 fairs againts the not yet retailing XP 2600+ have a look here: the P4 is running on a P4T533-C

It's from the Japanese consumer site AKIBA
http://akiba.ascii24.com/akiba/news/2002/08/24/638116-0... and honestly it looks a little more competent and faire than what I can see from this particular THG comparo.
And it gives even undue credit to AMD, because the XP 2600+ is not available to us for weeks, the P 4 2660 and 2800 are since last week.


http://akiba.ascii24.com/akiba/news/2002/08/24/images/i...

http://akiba.ascii24.com/akiba/news/2002/08/24/images/i...
August 24, 2002 5:20:54 PM

The Intel P4 2.66 and 2.8GHz cpus arent for sale where the review was being written plus, if intel wanted THG to review 'their' latest cpus, they should have sent them to THG as i'm sure AMD did.

Secondly, this is only a preview, a good preview too. As soon as THG get a couple of the new p4's you'll see a review of them too.

Who's General Failure and why's he reading my disk?
August 24, 2002 5:51:33 PM

that does not make sense:

(a) the XP 2600 + was also not on sales when the review was written, it is not even on sales now, not even in a few weeks.
(b) If I write a comparison and intend to make valid statements about competing hardware, I try to get what is necessary, or I mention at least that non-comparable stuff has been compared.
(c) taking previews stuff from one manufacture should always be done with the greatest care and reluctance. That AMD offered a CPU for a preview does ot allow to compare a new, previewed XP 2600+ with an outdated 2.53, because everybody knew that an Intel 2.66 and 2.8 would be available long before AMD is in the market with the XP 2600+. It's grossly misleading. At least a more cautious stance would have been prudent.

The conclusions in the review were wrong and are wrong.

This is what the THG reviews says: "Intel P4's seemingly permanent performance lead has been broken."

This is and was wrong. At no time had AMD a faster CPU, not at the time when the preview was written, and not now.

Fact is that Intel has a 2.66 and 2.8 out and selling, and AMD has currently nothing else than a XP 2200+ selling.
August 24, 2002 5:52:18 PM

Actually you are right about the biased review.. There was another thread about this un-fairity. People said, P4 should to be overclocked with the watercooling to see what's the real limit of that CPU. But they didn't, and i don't know why. Strange things are going in these days.. AMD's latest products are expensiver and worse than Intel's, and THG began to lose its reliablity.. I hope this is not real, just my wrong opinions. :frown: I hope...

<b>Before getting angry to the (d)evil, just think about who CREATED it!</b> :mad: 
August 24, 2002 5:53:18 PM

that does not make sense:

(a) the XP 2600 + was also not on sales when the review was written, it is not even on sales now, not even in a few weeks.
(b) If I write a comparison and intend to make valid statements about competing hardware, I try to get what is necessary, or I mention at least that non-comparable stuff has been compared.
(c) taking previews stuff from one manufacture should always be done with the greatest care and reluctance. That AMD offered a CPU for a preview does ot allow to compare a new, previewed XP 2600+ with an outdated 2.53, because everybody knew that an Intel 2.66 and 2.8 would be available long before AMD is in the market with the XP 2600+. It's grossly misleading. At least a more cautious stance would have been prudent.

The conclusions in the review were wrong and are wrong.

This is what the THG reviews says: "Intel P4's seemingly permanent performance lead has been broken."

This is and was wrong. At no time had AMD a faster CPU, not at the time when the preview was written, and not now.

Fact is that Intel has a 2.66 and 2.8 out and selling, and AMD has currently nothing else than a XP 2200+ selling. That is the reality, and this THG preview is painting a totally different picture from this.
August 24, 2002 6:07:42 PM

Quote:
AMD's latest products are expensiver and worse than Intel's


AMD's products are not more expensive than intels. AMD's top of the line processor out, the 2200+ is as little as $142 on pricewatch.com. Intel's 2.8 $540, even the 2.4 is still $250 - you have to go down to a 1.8 P4 to reach the same price range as AMD's top of the line chip, hardly more expensive. As for worse, I don't think so. An athlon XP at 2.13GHz put up a very decent fight against the 2.53GHz pentium 4 that has double the FSB speed, bandwidth and double the cache. Yes, overall the P4 still performed better, but not by too much for having all the advantages it did.

As for why a 2.8GHz wasn't in the benchmark, who cares? They did overclock a P4 system as reference, plus with the models numbers being 2400/2600, it makes since to bench them against a 2.53GHz processor since it falls in between as reference. How much sense does it make to have a 2.8Ghz in the test? It would only show how much faster it is than the 2.53GHz and the athlons. This wasn't a bench testing P4's, but athlons. I am sure they will release a new bench soon on the P4 2.8 and future P4's and pit them against the new athlons.

So why don't we all calm down and get over it. This isn't the first time THG has shown bias and undoubtly won't be the last, don't let it interfere with your lives... sheeesh :)  smile

"Opinions are like assholes, everyone has one"
August 24, 2002 6:13:31 PM

THG is the best neutral hardware reviewing site till now. They are more neutral than any hardware reviewing site.
August 24, 2002 6:20:40 PM

I think you should take a chill pill and relax urself.............

At the time of the review...the P4 2.6 and 2.8 were NOT released.....
As for includign them in the test...mayeb ur blind, but he DID Include the benchamrks of a P4 2.8Ghz chip..and OC'd 2.53......which should realisticaly have better performance than a P4 @ 2.8GHz....bigger FSB/MEMORY Bandwidth....

<A HREF="http://www.anandtech.com/mysystemrig.html?id=13597" target="_new">-MeTaL RoCkEr</A>
August 24, 2002 6:25:06 PM

Which one is the "top of line" CPU of AMD? XP2600, right?
Which one is better? P4 2.53 or XP2600? P42.53, right?
Which one is expensiver? 250$ or 299$? 299, right?

XP2600 may be not at shops now, it doesn't matter.

<b>Before getting angry to the (d)evil, just think about who CREATED it!</b> :mad: 
August 24, 2002 7:38:55 PM

Down boy... don't get so rabid over this. Yes, Intel has filled the retail channel with 2.6 and 2.8 chips, <i>but</i> Intel has not lifted the NDA (Non-Disclosure Agreement) on the reviews for those CPUs, because Intel has not <i>officially</i> released them yet (rumor sets that date as Aug. 26th, Monday). AMD conducted a paper-launch, something pioneered by Intel (and most video card makers) when the K7 reached 1GHz well before Intel was able to fill up the retail channel with their own 1GHz parts. But Intel went ahead and released a couple hundered samples of the 1GHz P3 and allowed reviewers to post reviews <b>even though you couldn't find a 1GHz P3 for a long time after it was "released"</b>. What AMD is doing is trying to steal the thunder of the release of the 2.8GHz P4. Simple as that. It isn't some vast conspiricy by THG and AMD to pull the wool over our eyes. The review clearly stated that the retail channels won't see these parts until the beginning of Sept. Same with the Anandtech review, TheInq, TheReg, Amdmb, [H], and others. Again, it was a paper launch, to try to stem the errosion of market share and confidence of investors and partners in AMD. That's all.

I wouldn't be surprised if Nvidia pulled this a month before the NV30 is actually in shipping quantities, and the Hammer by AMD will be the same way, as only about 100,000 will be ready by the end of the year, so a Christmas-time launch indicates that there will be very few retail Hammers, and most will be in OEM boxes. The same goes for the 2600+. Those that were ready by the 21st were probably sent straight to the OEM box makers, and when they are full, then the parts will trickle into the retail market, either by gray market channels (OEMs selling excess quantities to resellers) or the actual official processor-in-a-box (retail version directly from AMD).

-SammyBoy
August 24, 2002 9:09:35 PM

Quote:
It remains that this preview is grossly misleading, beause they compared a not retailing XP 2600+ with an outdated, "old" P4 2.53. The new 2666 and 2800 with the smaller die are retailing/selling and they should have been compared.

yeah, it's been brought up many a times before. the news is senstationalize, but it brought up some great discussions in the last week nonetheless. at the end of september or so, we'll see the p4 kick ass when all the real numbers are in.

also, the p4 3Ghz (with hyperthreading?) in the christmas season will make amd fanboys alot less jolly if they don't have any presents under the tree.

<font color=green> there's more to life than increasing its speed -Ghandi</font color=green><P ID="edit"><FONT SIZE=-1><EM>Edited by shallowbaby on 08/24/02 05:33 PM.</EM></FONT></P>
August 24, 2002 9:22:57 PM

first of all ...

"AMD ATHLON XP 2600 /266 FSB PROCESSOR CPU- OEM
Specifications:
CPU: 2.13 GHz
Type: XP 2600 Throughbred
Cache: 256K
BUS: 266MHz
Socket A

$299.00"

second of all ....

No one seems to understand that AMD had no research to do in making these chips. The only thing you are paying for is the cost it took to modify it and mass produce it. Where for intel your paying that as well as the cost it took to research it. The reason the pentium 4 is more expensive is because you are paying for the research it cost to produce the chip.

It's gonna be funny because when clawhammer is released everyone is gonna be like, "holy crap that chip is expensive" .. don't be surprised if the clawhammer is more expensive then the pentium 4. Because AMD has put a lot money into Research and Development (R&D).

I think money is irrelivent when compareing cpu's. Fact still remains, you get a cpu based on your needs.

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
August 24, 2002 9:52:03 PM

I can't agree with you.

Pentium 4 is priced higher not for it's brand new architecture. It's because P4 is an intel cpu. When Athlon classic appeared in the market at 1999, it was based on a brand new architechture. Pentium III's P6 architecture was older. But Athlon was priced lower than P3's.

We can expect same things from Clawhammer. It will be cheaper or at least equal priced compared to P4. And keep it in mind that Clawhammer will have integrated memory controller. So Clawhammer would require no northbridge chipset. As a result, clawhammer motherboards will be cheaper too.
August 24, 2002 10:15:55 PM

ok you're spinning my words... thats not what i said ... now read very very carefully ... as it seems complicated.

i said the pentium 4 is priced higher because you pay for the R&D ... the Research and Development. AMD had nothing to research. All of the technology came from Intel. The only thing AMD has done as of now is develop 3dnow which no one uses. They haven't even researched proper thermal protection. Because it cost a lot of money and it would raise the cost of chips dramaticly. They can't just go use Intel's design because thats patent infringment. So they would have to ask Intel, "can i please use your patents for thermal design" and INtel will say, "ok but it'll cost 15 bucks per chip" and then amd raises there chips up 30 bucks. It's not free!

these are the facts of business. Perhaps enlightenment is needed? Or down to earth realization that not everything is magic?

<A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
August 24, 2002 11:27:12 PM

Digitimes and other reputable news sources (read non-hardware sites, but professional business IT sites) peg the price for the initial Clawhammer to be around $400 per proc in a 1,000 unit order. Now, given AMD's reputation for keeping prices low, I tend to believe that they will be selling them at a price that will allow them a 5% profit margin per proc at most. They will probably try to establish the Hammer technology with the Clawhammer first, and then charge a Xeon-like prices for the Sledgehammer/Opteron, since that is the supposed market for the Opteron, and probably with a much large profit margin per proc.

As to the price of Intel CPUs vs AMD CPUs... Intel is relying on brand name to support the inflated prices. It's not really about R&D so much as Intel knowing that they can safely get away with higher prices because the majority of people out there buy on name rather than performance. Also, Intel sells 4 times the volume of procs that AMD does, meaning that they are able to get more from less... they can sell a proc at a 5% profit margin instead of 10%, and still end up with twice the money that AMD would... but they know that they can charge that 10% (theoretical, since I really don't know the actual price per proc) and still sell 95% of the procs they would if the price was only 5% more than the cost of production. Intel makes plenty of money on licensing agreements to cover much of the R&D costs. They license almost everything they make, and since they get a take on not only every proc produced, but chipset produced (well, forgetting the VIA situation), as well as many "standards" out there that they created, and probably got royalties for, it adds up to probably a good portion of the R&D costs. The P4 R&D was a fraction of the Itanium R&D, and Intel maintains that the price of Itanium chips includes very little to cover R&D costs. So, I'm sure Intel operates the same way with the P4. If you passed on all R&D costs to the consumer, rather than absorbing them, and *gasp* charging them as an expense, things would cost so much, no one would buy them, even for the Intel name.

And AMD does plenty of R&D. Being part of a consortium (sp?) such as HyperTranspot carries cost, as well as core layout, optimizations, etc. To actually believe that just because the K5, K6, and much of the K7 used technology from cross-licensing agreements, AMD didn't have to spend money on R&D is ridiculos. If I gave you all the parts of a car, and told you to do whatever you wanted with it, do you think that I'll also give you instuctions on how to put it all together, and hold your hand while you construct the car? No, I'll just take the money to the bank, provide you with details on what each part is on its own, and leave the rest to you.

If we applied this high cost comes from R&D costs, then Apple systems should cost near nothing, since all they really do is design an OS that works with whatever processor they feel like using that year. Right now it Motorola that's doing the grunt work, all Apple is doing is letting Motorola power the Macs, and paying Motorola a small fee, probably based on systems sold. But, Apple, since they own 100% of their market, still are able to charge high prices for their systems, even though they only do the R&D for the software aspect.

In conclusion, Intel charges what it does, because through shrewd business practice, they are charging the max amount they can while still being able to hold off AMD and not lose much market share, or, such as this year, actually gain market share. For every shmuck out there who balks at the price of a top of the line P4, there are 3 more shmucks who will pay that price. And Intel will keep those prices up there, not to pay for R&D, or other capex, but to keep investors happy. As long as there are more shmucks willing to pay that price than not, it'll stay up there. I think this is called supply and demand....

I mean, how else can you explain the difference between the 1.8GHz and 2.8GHz prices... we all know that most 1.8s can get 2.6GHz, so really, it's just a promise by Intel that the proc will run at that speed. It's just a little extra step of validation, clearly not $400 worth. But they know that for every proc they release, a certain number will be sold at a certain price level, so they just find the best procs sold vs price per proc point, and charge that. Also, Intel has shown in the past that it can drop the prices way down in a hurry if they have too, and can still make a profit, which tells me that they are getting way more then they need to in order to cover production costs.

Oh, and R&D money generally comes from sales in the past, so whatever profits are being reaped by the P4 are being applied to future products (more future than Prescott). Intel really doesn't have to spend it's profits on R&D if it doesn't want to.... but that's where R&D money comes from, profits and sales. It's an extra expense, though pretty much required in the IT world.

-SammyBoy
August 25, 2002 12:06:12 AM

AMD bought real R&D when they took over Nexgen. Before that they were just cloning Intel. Sanders wanted to gain 30% of the processor market from the time they merged in 95 and set the 30% goal to make in 98. That's from the K6 line, and up. AMD scrapped their future Roadmaps (their own K6 and up) and took NexGens. Nexgen didn't have a FAB, they were pure R&D. Used to sub out the manufacturing. Even passing Intel in performance, and some design innovation. NexGen (Now AMD's Microprocessor division) hasn't changed much. Still Innovative and resourceful as ever.......And still in the shadows of Intel

This sig runs too hot.
August 25, 2002 1:53:41 AM

Quote:
I tend to believe that they will be selling them at a price that will allow them a 5% profit margin per proc at most.

LOL!!!

That's a bold statement. Intel has 50% profit margins....

Intel can sharge whatever the consumer is willing to pay, and since it appears like they're doing well, then what's the point in lowering prices?

<i>Past mistakes may make you look stupid, but avoiding future ones will make you look smart!</i>
August 25, 2002 2:41:07 AM

" Intel is relying on brand name to support the inflated prices"

Intel has good reason to rely on just their brand name. They've worked hard to mkae sure that the Intel name is associated with quality.

"In conclusion, Intel charges what it does, because through shrewd business practice, they are charging the max amount they can while still being able to hold off AMD and not lose much market share, or, such as this year, actually gain market share."

Actually, I think that Intel is perposely doing things like that so they can point to AMD and say "Lookie! We're not a monopoly!". Intel needs AMD to do relativly well so they can keep the DoJ off their backs.

Knowledge is the key to understanding
August 25, 2002 5:28:24 AM

Oh man... it's a Van's review... and the only benches I know of that Intel would not be able to best the 2600+ are the SPEC veiwperf tests. Some AMD has a lead by as much as 40%. Otherwise, though, they are equal, at least 2600+ vs. 2.53GHz P4. The P4, I'm afraid, will probably blow the 2600+ out of the water.

-SammyBoy
August 25, 2002 5:36:44 AM

Seriously... I can't imagine that the Clawhammer will be able to sell for less than $350 initially without a loss per part. The reason is that it's a completely new core (well, a overhauled K7 core, at least), meaning that new testing proceedures had to be designed, new inventory systems, lines taken away from the T-bred B cash cow, etc. Also, with AMD publicly stating (well, telling it's investors) that they would only be able to ship 100,000 or so procs by the end of the year means that either the yields aren't up (raising the cost per part) or that they don't plan on starting until real late in the year. Regardless, the initial intro of the proc will be expensive, and still not net AMD much cash. But, when you intro a new product, you're not looking for cash always, but quick acceptance and strong demand. So, AMD wants to keep the prices as low as possible, and I'd imagine that they wouldn't want to tack on more than 5-10% to the price per proc, for the reason of making even recent computer buyers and enthusiasts reconsider their current system, and getting a Hammer-based one instead. After it's estabilished, I'd imagine they would charge near 50% for the flagship proc at initial release.

-SammyBoy
August 25, 2002 12:05:58 PM

Quote:
Seriously... I can't imagine that the Clawhammer will be able to sell for less than $350 initially without a loss per part. The reason is that it's a completely new core (well, a overhauled K7 core, at least), meaning that new testing proceedures had to be designed, new inventory systems, lines taken away from the T-bred B cash cow, etc. Also, with AMD publicly stating (well, telling it's investors) that they would only be able to ship 100,000 or so procs by the end of the year means that either the yields aren't up (raising the cost per part) or that they don't plan on starting until real late in the year. Regardless, the initial intro of the proc will be expensive, and still not net AMD much cash. But, when you intro a new product, you're not looking for cash always, but quick acceptance and strong demand. So, AMD wants to keep the prices as low as possible, and I'd imagine that they wouldn't want to tack on more than 5-10% to the price per proc, for the reason of making even recent computer buyers and enthusiasts reconsider their current system, and getting a Hammer-based one instead. After it's estabilished, I'd imagine they would charge near 50% for the flagship proc at initial release.

lol... I think you misunderstood me. When you said 5%, I was thinking a lot higher, not lower. :smile:

<i>Past mistakes may make you look stupid, but avoiding future ones will make you look smart!</i>
August 25, 2002 12:09:14 PM

Van's is probably the <b><i>WORST</b></i> review I've ever seen in my whole life! :mad: 

First off, they used a Dell system with PC800 only, therefore hurting the P4. (PC800 is probably as good as DDR333 in real life apps) Then, he used the 8K3A for the AMD system, and that's arguably the best platform for any AMD proc. With the AMD system, he could have tweaked the RAM settings and everything, and with the Dell you obviously can't.

Then he said there were signs of thermal throttling. What did he do? run it without a HSF?...

BS review.

<i>Past mistakes may make you look stupid, but avoiding future ones will make you look smart!</i>
August 25, 2002 1:28:20 PM

Quote:
Seriously... I can't imagine that the Clawhammer will be able to sell for less than $350 initially without a loss per part. The reason is that it's a completely new core (well, a overhauled K7 core, at least), meaning that new testing proceedures had to be designed,


You really dont know how much semiconductors really cost do you lol.

350 per part a loss???

Only if you only could fit 50 per wafer.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
August 25, 2002 3:21:53 PM

What size wafer and at what fab?


<A HREF="http://www.btvillarin.com/phpBB/index.php" target="_new">A better place to be</A> :wink: <P ID="edit"><FONT SIZE=-1><EM>Edited by scotty35 on 08/26/02 01:25 AM.</EM></FONT></P>
August 25, 2002 3:26:04 PM

For that 2.8GHZ Van review, I clicked the link and I got a 2.53GHZ vs XP2600 review, where is the 2.8GHZ?

As for your comments on the CH and how to sell it, I must say Sammyboy, this proves you are well fit into making a small PC selling franchise! You really know some nice strategies by yourself, and I am confident this can help you should you decided to go for it.

--
And now, an advice from your friendly Nike shoes slogan: JUST DO HER!
August 25, 2002 4:02:05 PM

It's amazing that some of you Intel fanboys are so blinded by Intel marketing and biased benchmarks that you can't read the "writing on the wall". If we're debating which processor is superior, and I'm talking about the PROCESSOR ONLY, wouldn't it make sense to put the P4 and Athlon on equal footings and benchmark the two using similar memory; as in PC2700? Because only when benchmarks are heavily SSE2 optimized or memory bandwidth dependent does the P4 look good. The best processor is the one which balances IPC and clock speed, something the Athlon does amazingly well. The P4 sacrifices IPC for raw clock speed, a marketing gimick that some of you seem to think makes it better. Which is why Intel designed the P4 the way they did.

<A HREF="http://www.vanshardware.com/reviews/2002/08/020822_Athl..." target="_new">http://www.vanshardware.com/reviews/2002/08/020822_Athl...;/A>
August 25, 2002 4:45:58 PM

I recall someone saying that Van's is very biased a little bit ago...

Knowledge is the key to understanding
August 25, 2002 5:51:21 PM

I would hardly call a Dell platform vs. a good AMD platform, which is designed for tweaking, equal footing. The P4 was choked back for this review. Additionally, what's wrong in using benchmarks that are SSE2 optimized. If SSE2 gives a performance increase in an application and people use that application then they would want to know how a processor with SSE2 does on that application. Benchmarks shouldn't be for fanboys to root for their team, they should give potential buyers some information when deciding what to buy. And no, I'm not an Intel fanboy. I've owned a K6-3, a Thunderbird, and a Palomino since 1998.
August 25, 2002 6:11:23 PM

Nor am I a fanboy of Intels. I prefer AMD parts, and am currently running only AMD boxes. I'm just telling you that Van's is the absolute worst in AMD fanboys, and does his best to make anything from AMD look it's best while crippling the Intel. The fact that he wouldn't even build an Intel test bed himself speaks volumes. If he had used both OEM boxes, it'd be a little more believeable, but with the AMD box being a tweaked out, made from scratch box, while relying on a mass produced, untweakable POS like a Dell for the Intel tells me that all he cares about is making AMD look good. While I disagree that using PC800 is a bad idea (only a 3-5% increase in performance with PC1066), such a biased POS site such as Van's is not something to base anything on.

Regardless, the 2600+ is about equal to a 2.53GHz P4. The 2.8GHz will probably increase the lead in many apps to the 7-11% barrier, making the performance gains actually tangible.

-SammyBoy
August 25, 2002 6:24:32 PM

A chip alone won't do a thing. You need a system built around it in order to have a computer. And only a computers' performance is measurable, not the chips.
And if u want to know my opinion, (you probably don't care, but I'll give it anyway ...) I think th P4 is the better chip. This is opinion is not based on benchmarks and stuff, but just on the technical beauty this processor shows. I've read some articles at <A HREF="http://www.arstechnica.com" target="_new">http://www.arstechnica.com&lt;/A> after someone (I think Eden, not sure) posted it in another thread. The P4 (even the Willametta) features many nifty and innovating things that makes me think it is superior to the Athlon XP. Though the Athlon has a very good IPC-ratio and is has a strong FP-unit, it feature (as far as I know) no really innovating and really new techniques, AMD just optimized the hell out of their K7-core. Nice work, but I prefer the beauty of a P4 to the raw power of an AMD.
If u want a non-opinion based comparison of the two, however, you will have to put them in a system. And then you have to choose a platform. Taking the best available one to get the most out of the chips is a logical thing to do, since you want the platform to hinder the performance of the processor as little as possible. The P4 has the feature of supporting a 533 MT/s FSB, the XP only a 266 MT/s one. That last thing is hindering the XP to perform better than it is now, but it is a CPU-bount feature, so it is AMD's fault the performance is lagging, not the platform makers. So, you <i>could</i> conclude that the XP is worse than the XP, since it has a worse FSB.
BTW, people are complaining about the XP having a too slow FSB and the P4 having a bigger L2 cache, but the XP has a very big advantage in it's L1 cache ... It's got 128 KB of L1 cache, the P4 only 8KB of L1-data-cache and an equivalent of, let's say, 24 KB L1-instruction cache. Imagine how a P4 would perform with a 128 KB L1-cache ... (Yes, I know P4's L1-cahce is faster than XP's, but I don't want to spoil my reasoning with that thing ...)

Anyway, my point being ... Chips aren't comparable without a platform to use it on.

"Dixi."

Greetz,
Bikeman

<i>Then again, that's just my opinion</i>
!