Sign in with
Sign up | Sign in
Your question

THE SUM OF ALL AMD FEARS!!!

Last response: in CPUs
Share
June 11, 2002 12:29:07 AM

well, there's no question about it, but imho (out of the gate) Intel's 13µ process is much better than AMD's.

can you point out a reason why I would be wrong?

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"

More about : sum amd fears

June 11, 2002 12:31:23 AM

have you some detail specifications for this new processor?
June 11, 2002 12:41:30 AM

Sure, I could point to the yields and crappy performance of the first P4s. Which means that Intel sucked at the .13 micron process for a while. Perhaps AMD will get better at it, just like Intel did :) 

Benchmarks are like sex, everybody loves doing it, everybody thinks they are good at it.
Related resources
June 11, 2002 12:54:08 AM

i think this will be certainly the case.
June 11, 2002 1:01:48 AM

Well, if you were to say that Intel's mature .13 process is performing better than AMD's not so mature process you might be closer to saying something true, maybe. However, you're comparing to very different cores so this would further muddle the statement.
June 11, 2002 1:43:29 AM

My first 2.0A that was purchased on Northwood release was able to reach 2.8Ghz, 800Mhz overclock is not "sucking" at launch.

The first available 1.6a reached 170FSB, that sucks too?

You are limited to what your mind can perceive.
June 11, 2002 2:14:38 AM

Again, I shall restate that since when is a chips overclockability a metric of its quality. All it means is Intel are morons and didnt label the chip at the correct speed. If they had better Q/A, they would have known the 1.6 would reach 2.8 gigs. And then sold it for a much healthier profit. All they did was lose money.
So let us explore if the original Northwood was better than the Williamette:

First link:
http://www.digit-life.com/articles/northwood22ghz/

** doesnt look that much better to me ***

This site got 400mhz out of their first P4, not bad:

http://www.overclockersonline.com/index.php?page=articl...


Another link showing the 2.0A wasnt much better than the Williamette:
http://www.ocprices.com/index.php?command=view&ID=1172&...
** If you look at the later benchmarks in the above link, you will see the athlon 1.6 beating the Intel 2.0A.. ***

Link:
http://www.active-hardware.com/english/reviews/processo...

*** not a whole lot of difference there, course I need new glasses though ***

My point is the Northwood didnt have a stunning release. It was a little better in the benches than Williamette. Sure it overclocked pretty good. But see my first statement. Which proves my point that the .13 micron process for Intel didnt go too well either. If it went well, how did a athlon 1.6 beat it?







Benchmarks are like sex, everybody loves doing it, everybody thinks they are good at it.
June 11, 2002 5:28:31 AM

well I am a very avid AMD fan.. Though AMD cannot beat intel in speed.. right now.. they offer a cpu with a better cost/MHz ratio.. witch will help them capture more market share.. Personally.. AMD is kickin Butt.. they have so much less money than Intel but can keep up with them as they are...
June 11, 2002 5:34:21 AM

Last bech show that 2.2 pc 1066 is equal to a XP 2200 so maybe the last speed of t-bred.

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
June 11, 2002 5:36:57 AM

I dont were you get that hardware.fr as been the 1 to test a early test sample and been able to break the 3 GHZ bar with a arir cooling

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
June 11, 2002 5:46:31 AM

a question...
Wasnt the Northwood initially meant to be the 'first' P4 out of the gate?
i thought that intel had to bolt the gate considerably to keep in touch with the XP at the time.


<font color=blue>Pants Down! Turn Around! Bend Over! You're about to Experience Telstra broadband! :lol: 
June 11, 2002 5:53:44 AM

The 1 P4 was featuring 128kb L2

northwood is a upgrade

cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
June 11, 2002 5:55:50 AM

indeed, which means they had a very long time to perfect it

<font color=blue>Pants Down! Turn Around! Bend Over! You're about to Experience Telstra broadband! :lol: 
June 11, 2002 6:48:42 AM

meltdown: I should hope Intel's 0.13µ processing is better than AMD's, they've only had an extra year to work all the kinks out.

-= This is our wading pool.
Stop pissing in it. =-
June 11, 2002 6:59:16 AM

At the high end everyone should agree that the P4 with a 533FSB is overall the more powerful processor and will be almost certainly be so through the next 6 months. Kudos to Intel for making a strong comeback after the disappointment of the willy core.

However on the lower end a system built around a new Tbred XP1700+ compared to a NWood P4 1700Ghz will probably provide a much better bang for the buck.

I am pretty happy with this current state of affairs. Intel is still feeling the heat at the top end with the potential impact of hammer and there is a heathly price war at the low end.

This all leads to me (the consumer) being the winner, and I love it!

<font color=blue> The Opteration was a success... I'm now a full-wit</font color=blue> :eek: 
June 11, 2002 7:05:32 AM

Quote:
My first 2.0A that was purchased on Northwood release was able to reach 2.8Ghz, 800Mhz overclock is not "sucking" at launch.

The first available 1.6a reached 170FSB, that sucks too?

Keep in mind that Intel was able to work most of the kinks out of their 0.13mu fab process with the Tualatin, who's 1.26GHz release did not see such incredible overclocks, either. It did do better than AMD's dismal 5% OC, which shows AMD is probably having start-up problems with the new fab process, but it wasn't the 50% OCing we saw with the NW. This is AMD's first shot at 0.13 and so should not be compared to Intel's second.

Likewise, you're talking about a 2.0A, which was not the top-of-the-line model at the time and therefor a better OCer. Lower rated tbred XPs will likely see much higher OCing results than the 2200+.
June 11, 2002 7:09:11 AM

Quote:
This is AMD's first shot at 0.13 and so should not be compared to Intel's second.

Very good point

Gosh I'm such a nerd sometimes, but then again arn't we all. :smile:
June 11, 2002 10:17:22 AM

"However on the lower end a system built around a new Tbred XP1700+ compared to a NWood P4 1700Ghz will probably provide a much better bang for the buck."

There is no Northwood 1.7. Theres a 1.6ghz, and when OC'd that chip out performs all AXP processors. I would say the 1.6A is more bang for the buck.

I sold my sig for $50.
June 11, 2002 10:25:45 AM

Quote:
There is no Northwood 1.7. Theres a 1.6ghz, and when OC'd that chip out performs all AXP processors. I would say the 1.6A is more bang for the buck.


If given the choice between a 1.6a or a 1700+ tbred I would go with the tbred, the 1700+@1.9ghz@166fsb will outperform the p4@2.4ghz which is above average for that chips overclocks.



:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
June 11, 2002 4:06:19 PM

You seem to be getting better with your thread titles, and worse with your logic. Have you been doing drugs?

<font color=blue>Hi mom!</font color=blue>
June 11, 2002 4:09:45 PM

At least we hope to, we do want to see the chip in action first. I am pretty confident, though... It's another stepping and not a downbinned from what I read.

--
Meow
June 11, 2002 5:47:31 PM

"If given the choice between a 1.6a or a 1700+ tbred I would go with the tbred, the 1700+@1.9ghz@166fsb will outperform the p4@2.4ghz which is above average for that chips overclocks."

Can I see some benches for that or is this anther I know the future statement?

Jeff
June 11, 2002 7:05:10 PM

Go see the THG OC test of a 1.9GHZ XP vs 2.4GHZ P4.

--
Meow
June 11, 2002 7:13:30 PM

You mean you actually trust THG's benchmarks?

<font color=blue>Hi mom!</font color=blue>
June 11, 2002 7:17:32 PM

Let me rephrase that where can I buy a 1700 t bred that will over clock to 1900?

Jeff
June 11, 2002 7:21:40 PM

Well there has to be some results that make sense man!
The XP2400 is a 1.9GHZ CPU, no? Technically it should rival the PR2400 right on, with some more losses against 533MHZ FSB versions.

BTW Fat remember our Optimization debate, where I was saying that something isn't entirely optimized until it runs higher per clock by its optimizations? Well take a look at Anandtech's SSE 2 optimized Adobe test: <A HREF="http://www.anandtech.com/cpu/showdoc.html?i=1635&p=9" target="_new">http://www.anandtech.com/cpu/showdoc.html?i=1635&p=9&lt;/A>

Now this my friend is a perfect example of how SSE2 is finally pushign the P4 IPC in these programs way above Athlon's, and thus explains indeed that when they properly optimize, it CAN be better per clock. This proves here that Athlons cannot anymore compete future SSE2 optimized apps if they follow Adobe's wonderful work here. However as for Newtek's cheating Lightwave 7B, even then the opt isn't making it better per clock like the Adobe's optimizations, so even their cheating isn't that worth it. As for why I say cheat, I or Mat can tell you that no Celly 1.3GHZ is supposed to beat a 1.4GHZ XP.

--
Meow
June 12, 2002 2:41:30 AM

Quote:
Let me rephrase that where can I buy a 1700 t bred that will over clock to 1900?

Coming out soon, next time think before you make mistakes(like asking for benchmarks that exist on thg's own article we are discussing).

Id just like to say, your attitude has been annoying, I have never met you or read your posts before today, but you comment towards me(watching me squirm) as if you are fugger or meltdown and I have personally humilated you time again on the boards.


I have not, but if you wish to make me an enemy do so at your own peril, even if my prediction was wrong, you have no right to attack/flame me.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
June 12, 2002 2:48:26 AM

What is worse, is I have a hard time understanding how a Tbred 1700+, with a 0.2V less than its counterpart 0.18m, could not OC as well. It is now obvious those chips are not perfected 0.13m silicons, since the 0.18m, your sys was OCing much higher.

--
Meow
June 12, 2002 2:53:47 AM

My cpus never been over 1.7ghz, I havent unlocked it, im guessing from my 104f max temp under load I have some room, I just dont care enough to open her up and unlock it.

But there are 1.9ghz axps out there, I think somethings fishy, waiting for more results.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
June 12, 2002 3:00:53 AM

Well whether or not it did, it still means that the 0.13m version should do better, but according to Anand, it didn't. Your system, whether 1.7GHZ or a tad below, proves how the new chip has a silicon problem.

--
Meow
June 12, 2002 3:48:27 AM

Quote:
Id just like to say, your attitude has been annoying, I have never met you or read your posts before today, but you comment towards me(watching me squirm) as if you are fugger or meltdown and I have personally humilated you time again on the boards.


<scary>oooh, be afraid, be afraid of the big bad matisaro, oooooh, don't you dare dis the big bad matisaro! ooooooh.</scary>

I think jeffs right on the money, <i>We</i> are watching you squirm.

everyone is in disagrement with you, yet you continue in making blatant excuses for AMD, if that's not AMD bias I don't know what is.

maybe dhlucke should put you the list(if he's not looking the other way).

"<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"<P ID="edit"><FONT SIZE=-1><EM>Edited by AmdMELTDOWN on 06/11/02 11:50 PM.</EM></FONT></P>
June 12, 2002 3:56:14 AM

Alright bitch, you want to flame Mat, fine.

Take this for example, the infamous Pentium4. Before it was even released, the Pentium3 existed in .13m already, meaning the Pentium4 could, or couldn't it?

Nope! The stupid hoe was released in .18 as the weak P4 Wilamette, with only 256kb L2, which had a hard time competing with the Thunderbirds...namely the 1.4Ghz.

Now, are you expecting AMD to have a flawless release of .13m on their first try? Intel didn't with the Pentium3 tually, it worked, but didn't oc beyond a couple hundred Mhz. Given time, it has matured, and picked up an additional 256kb cache. Picture the Barton, an improved T-bred, and you have a better idea of AMD's new weapon.

"you continue in making blatant excuses for AMD"
so all of a sudden, AMD needs an excuse to make a product comparable to the 1.6a? On the first try?

You are nothing more than Intel progaganda and FUD




"When there's a will, there's a way."
June 12, 2002 4:17:33 AM

Matt has admitted that he might have made a bad prediction based on the info he had at the time. That's more than I can say for others. Matt puts up a good debate.

<font color=red>I have a computer and it does weird stuff. please help.</font color=red>
June 12, 2002 4:21:04 AM

Are you referring to me? If you want to flame me, say it

"When there's a will, there's a way."
June 12, 2002 4:35:56 AM

Note that I responded to AMDmeltdown after he mentioned me by name.

<font color=red>I have a computer and it does weird stuff. please help.</font color=red>
June 12, 2002 4:40:16 AM

Oh *embarressed look* i'm terribly sorry to jump to conclusions

I apologize

"When there's a will, there's a way."
June 12, 2002 7:07:51 AM

No problem. Don't worry about it. :smile:

<font color=red>I have a computer and it does weird stuff. please help.</font color=red>
June 12, 2002 7:22:58 AM

Nope! The stupid hoe was released in .18 as the weak P4 Wilamette, with only 256kb L2, which had a hard time competing with the Thunderbirds...namely the 1.4Ghz

Ok let do some bench I got a willi get a thunder bird 1.4.

To all lemmimng you have screw up ok give up it just going worse.K7 have reach it limits even the Pr it no good anymore.Let talk about K8 without overhype.


cheap, cheap. Think cheap, and you'll always be cheap.AMD version of semi conducteur industrie
June 12, 2002 1:21:49 PM

I just sick of seeing your high and mighty I’m always right attitude. Have a nice day!

Jeff
June 12, 2002 1:47:08 PM

We got a labdog fanboy over here people, get the forum police!!!

--
Meow
June 12, 2002 3:24:23 PM

Let me tell you what I see here jeffg.

I see a guy, hes an enthusiast, great, maybe hes posted a hundred times or so in the printer forum, I certainly havent seen you in the cpu forum, and if you have posted here that tells me your posts dont stand out from the rest of the newbie garbage.

I see a guy who seems to be taking a pleasure in attacking someone in a gang, because that guy is too much of a pussy to disagree with me on his own, if you are tired of my "high and mighty" replies why didnt you say something to them, perhaps its because you are a spineless nobody who silently disagrees with my points and posts. You know, I dont like fugger, meltdown and the rest of the troll wastes, they have nothing good to add to any conversation, much like you have nothing good to add to this one. However, you know what? At least they have the BALLS to debate me when they think im wrong, unlike you, who resorts to attacking people when they make a mistake or you percieve them to be vunerable.

I take it from your unknown status coupled with your dislike of me that you are a long time lurker who apparently has disagreed with me in the past, but being too chickenshit to say so untill now only makes you a spineless oppurtunist and a nameless flamer and while you may enjoy imagining I am squirming now, I assure you I am not.

I have no problem admitting I am wrong, and I have no problem telling people I disagree with I think they are wrong, and apparently if you had one tenth the resolve and honesty I have you would have attempted to call me wrong when it was prudent, instead, you lurk and try to gangbang someone when you incorrectly percieve they are down. That is what makes you pathetic.

Well, little man, I am not down, and you are still spineless....whos squirming now.

:wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
June 12, 2002 3:52:13 PM

Quote:
<i>Eden says:</i>
where I was saying that something isn't entirely optimized until it runs higher per clock by its optimizations?

To which I heartily disagree.

Quote:
<i>Eden says:</i>
Now this my friend is a perfect example of how SSE2 is finally pushign the P4 IPC in these programs way above Athlon's, and thus explains indeed that when they properly optimize, it CAN be better per clock.

Or it explains that they didn't optimize well for Athlon. If they decided to only feed two FPUs and include SSE2, it'd be simple to make the P4 look faster per clock.

Quote:
<i>Eden says:</i>
This proves here that Athlons cannot anymore compete future SSE2 optimized apps if they follow Adobe's wonderful work here.

Perhaps, but I have a hard time believing that SSE2 makes that much of a difference in a 2D graphics application.

Quote:
<i>Eden says:</i>
However as for Newtek's cheating Lightwave 7B, even then the opt isn't making it better per clock like the Adobe's optimizations, so even their cheating isn't that worth it. As for why I say cheat, I or Mat can tell you that no Celly 1.3GHZ is supposed to beat a 1.4GHZ XP.

And no 1.4GHz P4 should be able to beat a 1.4GHz Athlon, as you're saying would happen with Photoshop. So why is one cheating and the other not?

Quote:
<i>Quetzacoatl says:</i>
Take this for example, the infamous Pentium4. Before it was even released, the Pentium3 existed in .13m already


There was a year or more between the P4 being released and the P3 going to .13.

Quote:
<i>Quetzacoatl says:</i>
it worked, but didn't oc beyond a couple hundred Mhz


Which is more than the Tbred is showing, percentagewise. AMD also has a more advanced fab that what the Tullies came out of, so I of course expect more from it.

Quote:
<i>Quetzacoatl says:</i>
so all of a sudden, AMD needs an excuse to make a product comparable to the 1.6a? On the first try?


I agree, I don't expect non-downbinned Tbreds to OC 50%. The down-binned ones should, according to Mat. We'll see. It doesn't look like it now, but who knows what'll change in the next couple of weeks.

Quote:
<i>Quetzacoatl says:</i>
Oh *embarressed look* i'm terribly sorry to jump to conclusions


Maybe we should get you the 'Jump to Conclusions Mat' from Office Space :wink:

<font color=blue>Hi mom!</font color=blue>
June 12, 2002 8:58:54 PM

Quote:
Or it explains that they didn't optimize well for Athlon. If they decided to only feed two FPUs and include SSE2, it'd be simple to make the P4 look faster per clock.

No because Athlon does not use SSE2. They are probably optimizing based on the multimedia extensions a processor use, and the P4 has an upper hand now with SSE2. We'll see how much the ClawHammer will benefit when they test it with SSE2.


Quote:
Perhaps, but I have a hard time believing that SSE2 makes that much of a difference in a 2D graphics application.

Well it can if they work it right. PhotoShop is FPU intensive, no? Well one can suppose that the full use of SSE and SSE2, made it shine. AMD commented themselves that SSE 2 is a wonderful and powerful MM Xtension, and that they wanted it. If anything, Adobe did a good job on optimizing. However no P4 at stock use, under 2GHZ is really worth buying for these tasks, even if it is optimized to give a few seconds less per clock agains the XP. Above 2.2GHZ, yes it is worth to get it for Adobe since the gap widens a lot.

Quote:
And no 1.4GHz P4 should be able to beat a 1.4GHz Athlon, as you're saying would happen with Photoshop. So why is one cheating and the other not?

Now really you are playing the devil's advocate. I can see through you that you couldn't find a better argument.
Again I mentioned Adobe's opts were awesome, but Newtek's weren't, but Newtek's ARE cheated. Because a Celery 1.3GHZ is not better in any way, regardless of using SSE to compete a XP1600. How on earth you try to find arguments in favor of Newtek, I simply don't know why. It is obvious they have not included some AMD string or something. Many people have found this theory to be credible, and THGC has had a debate once on the liability of this program.


--
Meow
June 12, 2002 9:54:14 PM

Quote:
<i>Eden says:</i>
They are probably optimizing based on the multimedia extensions a processor use, and the P4 has an upper hand now with SSE2.


My point is that you don't know that, it's an assumption. That may be the case, but I question why SSE2 suddenly makes such a dramatic appearance.

Quote:
<i>Eden says:</i>
Now really you are playing the devil's advocate. I can see through you that you couldn't find a better argument.


I am completely serious. You are basing two separate conclusions on the exact same facts. In both cases, one chip runs faster than it should. In both cases, little is known about the true optimizations. How you are arriving to opposite conclusions is beyond me.

Quote:
<i>Eden says:</i>
but Newtek's ARE cheated.


Why? Tell me exactly what makes these two cases different.

Quote:
<i>Eden says:</i>
How on earth you try to find arguments in favor of Newtek, I simply don't know why.


I'm not, I am merely pointing out the fact that you are making conclusions based on assumed facts. How do you know that SSE2 is what makes the difference? Do you have proof that those are the only optimizations put into Photoshop? Do you have proof that equal time was spent optimizing for Athlon as well as Intel?

<font color=blue>Hi mom!</font color=blue>
June 12, 2002 10:07:16 PM

Because I remember seeing in some benchmark, a Celeron 1.3GHZ beating an AXP 1600. Can YOU point out a reason for that?

--
Meow
June 12, 2002 10:08:48 PM

BTW I took this conclusion, because Anandtech said so. They make more knowledgeable reviews than THG, so I will take what they say so far, to be true. Adobe is making full use of the SSE2. When you check out non-SSE2 optimized Photoshop you will some bad P4 results, IIRC, but I am not sure.

--
Meow
June 12, 2002 10:42:06 PM

Quote:
Because I remember seeing in some benchmark, a Celeron 1.3GHZ beating an AXP 1600. Can YOU point out a reason for that?


Because Newtek optimized much more for the Celeron architecture than the Athlon XP architecture. Whether they did so unethically or not is another issue.

Quote:
When you check out non-SSE2 optimized Photoshop you will some bad P4 results, IIRC, but I am not sure.


But is that the only thing that changed? Unless you know that for sure, it's just an assumption on SSE2's power.

<font color=blue>Hi mom!</font color=blue>
June 12, 2002 10:57:17 PM

What's so much impressive about the Celeron's architecture to actually be THAT dumb, and optimize for it than Athlon?
This just shows Newtek being very pro-Intel.

As for the SSE2, I simply don't know what else would Adobe invest in optimizing, if they themselves (hell even Meltdown) said it was purely SSE 2 opt. Don't beleive it if you want, but to me, this looks like SSE2 and the P4 showing some real flexing.

--
Meow
June 12, 2002 11:11:47 PM

So it's not just good cause "I like it"? I go w/ anand also, unless Fat got a benchmarkin website of his own....

I sold my sig for $50.
!