Sign in with
Sign up | Sign in
Your question

prescotts tested by XS forum members, runs not hot

Last response: in CPUs
Share
February 9, 2004 9:51:04 AM

Okay, im beeing sarcastic, it runs hot as a 60 year old virgin.


<A HREF="http://forums.amd.com/index.php?showtopic=7710
" target="_new"> Open your eyes Intel Die HArd Fans ! , been reading to much Toms reviews eh ? </A>

And they have no 64 bit, so you must realy be an idiot to buy a P4 richt now.
2004 is the year of Athlon64, Prescott is slower and much much much hotter !

So you must be a real Die hard Intel fanboy if you still prefer Intel.


THG : The last 5-6 reviews have allways either had skewed results, or just somehow strangely come to a completly diferant conclusion then all of the other sites around.<P ID="edit"><FONT SIZE=-1><EM>Edited by coop on 02/09/04 08:38 AM.</EM></FONT></P>

More about : prescotts tested forum members runs hot

February 9, 2004 11:07:41 AM

ore an Idiot ofcourse...


THG : The last 5-6 reviews have allways either had skewed results, or just somehow strangely come to a completly diferant conclusion then all of the other sites around.
February 9, 2004 11:27:38 AM

remove the 2nd http:// in your link :) .

System on the way:
Antec Sonata Case w/380W PSU
Athlon 64 3200+
MSI K8T Neo-FIS2R
1 GB Mushkin PC3500
Sapphire Radeon 9800PRO
WD Raptor 74 GB
Related resources
February 9, 2004 11:40:40 AM

Okay, the adjustment is done, now all go read the user experiences.

Warning : this tread is not intend for Intel lovers with a weak hart !


THG : The last 5-6 reviews have allways either had skewed results, or just somehow strangely come to a completly diferant conclusion then all of the other sites around.
February 9, 2004 1:26:30 PM

Your point is what?
Anonymous
a b à CPUs
February 9, 2004 2:29:51 PM

Ya we already know that you can cook and egg on modern cpu...
Anythin else?
February 9, 2004 4:16:19 PM

Although you do have a valid point there, there is no reason to get so emotional about these things!
February 9, 2004 4:36:53 PM

We hate big companies like Intel and Microsoft because we just love to hate them. ;) 

A fine day!
February 9, 2004 4:45:31 PM

Let me know how hot your Athlon 64 is when it runs at 4 GHz, ok?
February 9, 2004 5:57:53 PM

Quote:

Let me know how hot your Athlon 64 is when it runs at 4 GHz, ok?

Your an intel lover, aint you ?
Well, try to understand what i`m saying : Intel chooses for MHz, AMD chooses for Performence.
Overclock a A64 a little bit and it will rape the Prescott so it aint funny to see anymore.
And you dont need a vapo chill to cool it down to 25 degrees !

Get real, anyone with a brain would not even consider a Prescott ! Prescott is a joke !
If AMD would make such a CPU there would be no more AMD in a few months !





THG : The last 5-6 reviews have allways either had skewed results, or just somehow strangely come to a completly diferant conclusion then all of the other sites around.
February 9, 2004 6:14:44 PM

Quote:
Intel chooses for MHz, AMD chooses for Performence.


Coop, you're an idiot, and that statement proves it.

You're right that more MHz doesn't necessarily transalte into more performance. However, (on the other side of the equation) increased IPC doesn't necessarily lead to increased performance either.

<b>In order to increase performance, IPC AND clock must be increased. AMD relies more on the IPC side, while Intel leans towards higher clocks.</b>

Now, if MHz truly meant nothing, then Intel would be a lot further behind AMD than they currently are, wouldn't you say? If your IPC is weak, higher clock will boost your performace. (Intel) If your clock is lower than your competitor, increasing IPC will give you extra performance. (AMD)



<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>
Anonymous
a b à CPUs
February 9, 2004 6:54:11 PM

A few thing
Quote:
Well, try to understand what i`m saying : Intel chooses for MHz, AMD chooses for Performence.

I think both chose the perfomance way. Intel higher mhz ; AMD higher IPC. Both ways can work but high MHZ number are sexier to martket...
Quote:
Overclock a A64 a little bit and it will rape the Prescott

Maybe because the prescott is more positioned as a mainstream product...

Quote:
Get real, anyone with a brain would not even consider a Prescott !

Right now this is absolutly true if some 3.8/4ghz version are descently priced, it might be another story...
February 9, 2004 7:25:50 PM

Quote:
Your an intel lover, aint you ?

Not a lover, just a happy costumer. But I'm a programmer so stability and new instruction sets mean the world to me. I guess I just have a different point of view. Oh, and my desk drawer is full of older Intel ánd AMD chips. Strangely all Intel chips still work and all AMD chips are broken...
Quote:
Overclock a A64 a little bit and it will rape the Prescott so it aint funny to see anymore.

And how exactly are you going to overclock an Athlon 64? It already runs at the limit of 130 nm technology. When you overclock it, it consumes more power than Prescott at the same performance level. And while Prescott can be overclocked 25% easily and still has a great future ahead, Athlon 64 on 130 nm is at its end.
Quote:
And you dont need a vapo chill to cool it down to 25 degrees !

Since when does temperature completely determine performance? Laptop chips are often specially designed to run at 100 degrees, stable! That's just to have a bigger temperature gradient so heat is dissipated more easily. I once even had a desktop Pentium III 800 MHz passively cooled, at 120 degrees. Made the socket color brown, but that chip still lives in my server. Try that with an AMD chip.
Quote:
If AMD would make such a CPU there would be no more AMD in a few months !

In a few months, AMD won't be able to keep up with Prescott. And by the time they get 90 nm technology working, Intel is running Tejas with improved 64-bit and Hyper-Threading at 5 GHz.

Just face it, exactly what future does AMD have with their current chip? Maybe 100 MHz more, but then they quickly have to come up with something new and I'm afraid their box of tricks is empty right now...
February 9, 2004 8:32:19 PM

I believe you have just rightfully won yourself an "Intel Fanboy" award. So coop exaggerates quite a lot but you are not any better the way I see it!

1. Let me know how hot your Athlon 64 is when it runs at 4 GHz, ok?
2. But I'm a programmer so <b>stability</b> and new instruction sets mean the world to me.
3. When you overclock it, it consumes more power than Prescott at the same performance level.
4. I once even had a desktop Pentium III 800 MHz passively cooled, at 120 degrees. Made the socket color brown, but that chip still lives in my server. <b>Try that with an AMD chip.</b>
5. In a few months, AMD won't be able to keep up with Prescott. And by the time they get 90 nm technology working, Intel is running Tejas with improved 64-bit and Hyper-Threading at 5 GHz.
6. Just face it, exactly what future does AMD have with their current chip? Maybe 100 MHz more, but then they quickly have to come up with something new and <b>I'm afraid their box of tricks is empty right now</b>.

And btw, you said you are a programmer? I am also studying to become one, this is my first year in university, but I am sorry to say that for a person that is supposedly this close to computers, you are very ignorant when it comes to computer hardware.

I just don't get it why you get so emotional about these things. This question includes coop as well.
February 9, 2004 8:47:56 PM

>And how exactly are you going to overclock an Athlon 64? It
>already runs at the limit of 130 nm technology

blah blah blah. true, it doesnt overclock well. not true "its already at its limit". First Opterons @1.6 GHz didnt overclock squat either, I guess they where at the limit of 130nm SOI as well ?

> When you overclock it, it consumes more power than
>Prescott at the same performance level

Utter nonsense.

> And while Prescott can be overclocked 25% easily

Define "easy". If you don't mind your motherboard catching fire, I guess it is. See here: <A HREF="http://www.hardocp.com/article.html?art=NTg3" target="_new">http://www.hardocp.com/article.html?art=NTg3&lt;/A>.
Quote:
We used two ABIT boards that we had available. The AI7 and the IC7-MAX3. Both boards passed stock CPU speeds stress tests at 3.2GHz. We were working on stressing these boards before we got our 2.8E so we had days of stress testing done when the 2.8E arrived. All of the motherboards we have used with Prescotts run hotter than we are used to. <b>It is not uncommon to see components on the mother at 250F/120C degrees</b>.

The IC7-MAX3 did do something that we have never seen before. If you have noticed, we have used square blocks of plastic foam to elevate our boards for years now while we are testing them. <b>The IC7-MAX3 got so hot that it actually melted the plastic block into a liquid form </b>


You can love intel all you want, but there is just no denying Prescott runs incredibly HOT. Too hot for confort, and too hot to overclock without watercooling.

>In a few months, AMD won't be able to keep up with
>Prescott. And by the time they get 90 nm technology
>working, Intel is running Tejas with improved 64-bit and
>Hyper-Threading at 5 GHz.

And another nonsense statement. "improved 64 bit" ? Got a link ? Hyperthreading ? big deal, seen that for years now. 5 GHz ? Someday, you think AMD will still only be selling 2.4 GHz parts by then ?

>Just face it, exactly what future does AMD have with their
>current chip? Maybe 100 MHz more, but then they quickly
>have to come up with something new and I'm afraid their box
>of tricks is empty right now...

No you're not afraid.. you hope it is empty. Let's just hope it for you, but you may be in for a dissapointment. I'm sure you where dissapointed by Prescott, better get used to it.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 9, 2004 9:52:05 PM

Quote:
I believe you have just rightfully won yourself an "Intel Fanboy" award. So coop exaggerates quite a lot but you are not any better the way I see it!

Please clarify because I can't see where I was exaggerating. And no, I'm not an Intel fanboy. In fact I advise anyone to buy an Athlon 64 if they need a new system these times. But the way Prescott is said to have no future is just not correct. So in a few months I'm quite certain that I will be advising people to buy a Prescott or wait for Tejas.
Quote:
1. Let me know how hot your Athlon 64 is when it runs at 4 GHz, ok?

No exaggeration here. I'm very serious that Athlon 64 simply can't run at 4 GHz, or let's be fair, at the performance level of a Prescott at 4 GHz without consuming <i>at least</i> the same power.
Quote:
2. But I'm a programmer so stability and new instruction sets mean the world to me.

Not exaggerated either. No Intel system ever failed on me because of processor instability. As another example of AMD instability, my neighbor has an Athlon XP 2500+ for a while. Great system with great performance, but it has just instantly ceased to run a few times, when he was working on his thesis about sattelite picture processing. The rest of the system has good components so it can only be due to the processor, and 24 hour stability tests confirmed this. Coincidence? Maybe, but I've heard too many other problems with AMD systems where the only solution was to replace some expensive components. Even the cheapest Intel systems I've helped building never experienced such problems even in conditions like bad ventilation hot summers.
Quote:
3. When you overclock it, it consumes more power than Prescott at the same performance level.

Like I said above...
Quote:
4. I once even had a desktop Pentium III 800 MHz passively cooled, at 120 degrees. Made the socket color brown, but that chip still lives in my server. Try that with an AMD chip.

Seriously, I'd love to hear anyone claim that with an AMD chip, it would immediately gain a lot of my respect. But these burned AMD chips in my desk's drawer indicate the opposite.
Quote:
. In a few months, AMD won't be able to keep up with Prescott. And by the time they get 90 nm technology working, Intel is running Tejas with improved 64-bit and Hyper-Threading at 5 GHz.

Still can't see why I would be exaggerating. 90 nm for AMD won't be expected before the end of 2004, while Tejas is to be expected spring 2005. So by the time AMD really does bring non-prototype 90 nm chips to the market, it will have to compete against 5 GHz Tejas.
Quote:
6. Just face it, exactly what future does AMD have with their current chip? Maybe 100 MHz more, but then they quickly have to come up with something new and I'm afraid their box of tricks is empty right now.

With still nearly a year to go for 90 nm technology, I can't see how they would magically pull off an entirey improved core before that time. Or do you know details I haven't heard of? So, don't get me wrong, the current Athlon 64 is an excellent chip but it won't be able to compete long against Prescott. Only 64-bit technology and some lower power consumption will keep it attractive (and price of course).
Quote:
And btw, you said you are a programmer? I am also studying to become one, this is my first year in university, but I am sorry to say that for a person that is supposedly this close to computers, you are very ignorant when it comes to computer hardware.

To be more precisely I'm a civil engineering student in computer sciences, but I've been programming professionally for several years now to pay for it all. And before you call me ignorant, I suggent taking some processor architecture classes, and that's nothing personal. All I see is that in the current situation AMD has a really nice chip, but it won't keep the performance crown for long. And I'm not saying Intel will suddenly race past it and AMD is doomed. All I'm trying to do is make people realize that these are alternating situations but in the end Intel will remain the overall winner. It's just ridiculous that I'm called a fanboy for just being realistic and not sharing the opinion that AMD rulez just because that happens to be the case right this moment.
February 9, 2004 10:40:07 PM

> So in a few months I'm quite certain that I will be
>advising people to buy a Prescott or wait for Tejas.

I'm sure you adviced them to wait for Prescott a few months ago as well.

> or let's be fair, at the performance level of a Prescott
>at 4 GHz without consuming at least the same power.

Yeah let's be fair. The A64 currently roughly consumes *half* the power of a Prescott for equal performance. That is on 130nm versus 90nm for intel. By the time Prescott hits 4 GHz, AMD will have shifted to 90nm. If all goes as well as IBM's very similar 130->90nm process shrink, it will cut power consumption again in *half* (cf PPC970FX). If intel doesnt fix its process, A64 may consume only <b>1/4</b> the power of a equivalent Prescott, and this time, I am hardly exagerating.

> As another example of AMD instability, my neighbor has an
>Athlon XP 2500+ for a while. Great system with great
>performance, but it has just instantly ceased to run a few
>times

If it where his CPU, I'm sure he got a RMA. But I bet you a fortune it was not his cpu, *if* you didnt make up the story, it was more than likely his PSU, or his motherboard. I've worked in this industry about as long as you walk on this earth, and I have yet to see *any* cpu die for no reason.

> But these burned AMD chips in my desk's drawer indicate
>the opposite.

They only indicate you're a complete moron, or more likely, lying through your teeth. Why not post a quick photo of all those burnt cpu's ?

>Still can't see why I would be exaggerating. 90 nm for AMD
>won't be expected before the end of 2004

Says who ? AMD says second half '04. could be august as well as december.

>while Tejas is to be expected spring 2005.

Yeah, at 4 GHz. At a time AMD will be selling 4400+ cpu's.

>So by the time AMD really does bring non-prototype 90 nm

You think that will be before or after intel releases non prototype 90nm cpu's ?

>chips to the market, it will have to compete against 5 GHz >Tejas

5 GHz is not gonna happen before the end of 2005. Keep dreaming mate..

>With still nearly a year to go for 90 nm technology, I
>can't see how they would magically pull off an entirey
>improved core before that time.

Why should they? Its not like Prescott is a serious threat, even to old process 130nm A64's. A die shrink is all AMD needs to remain at least competitive, and IMHO, maybe not even a die shrink.

>So, don't get me wrong, the current Athlon 64 is an
>excellent chip but it won't be able to compete long against
>Prescott

Based on what, besides pipedreams of future 5 GHz chips ?

> All I'm trying to do is make people realize that these are
>alternating situations but in the end Intel will remain the
>overall winner.

In an alternate universe perhaps ?

>It's just ridiculous that I'm called a fanboy for just
>being realistic

you're not !! You are *hoping* (expecting/thinking/) intel will leapfrog AMD, but there is just no evidence whatsoever to base this upon. Face the facts: intel has just shrunk their top of the line processor, introduced a new core, and still lags behind AMD's chips. AMD has a shrink and redesign (Newcastlle) ahead of it, what exactly makes you think Intel will magically pull this rabbit out of their hat, lower power consumption significantly, increase clock speed, while at the same time, expecting AMD to stumble on their die shrink, being stuck at 2.2 GHz, .. ?

>and not sharing the opinion that AMD rulez just because
>that happens to be the case right this moment.

Exactly. and no data I am aware leads me to believe it will change much in the next year. The fact you are convinced, and state as fact the oppposite, with no data, not even rumours, leaked roadmaps, or intellegent speculation.. *nothing*.. that is what makes you a fanboy. The fact you throw in unprovable and very unbelieveable claims about mysterious deaths and instability of AMD cpu's, makes you a (lying) troll as well.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 9, 2004 11:13:42 PM

Erm... Sorry, P4Man, but I'll have to correct you...
Quote:
AMD has a shrink and redesign (Newcastlle) ahead of it, what exactly makes you think Intel will magically pull this rabbit out of their hat,

Newcastle is not 90nm, it's 130nm. It's a value line that will be introduced in Q2 2004; check out<A HREF="http://www.amd.com/us-en/Processors/ProductInformation/..." target="_new">AMD's own roadmaps.</A> The correct name you're looking for - which is more a shrink than an actual redesign - is San Diego. It's a 90nm shrink of the current A64, and it is indeed slated for the 2nd half 2004; it's not a true redesign, however, as far as I know. Not much will really change in the architecture.

I'm more inclined to think that 90nm chips from AMD will only be around by Q4, but that doesn't make AMD any less of a threat to Intel. I must agree that there is the possibility that they eventually will manage to get it out by Q3, though I still think it's unlikely...

<i>Edit: I think the 3000+ already is a Newcastle core: it has half the cache. I'm not completely sure on that one, though. In any case, Newcastle is supposed to give a boost to mainstream A64, not state-of-the-art ones.</i>

<i><font color=red>You never change the existing reality by fighting it. Instead, create a new model that makes the old one obsolete</font color=red> - Buckminster Fuller </i><P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 02/09/04 11:17 PM.</EM></FONT></P>
February 9, 2004 11:47:00 PM

I was glancing through the pictures at work earlier and thought "What the #*($ is that?" and then of course had to actually do some reading. Right now, I could set my case out my window to keep in nice and cool, but in a few months that will be no good. Maybe they could make it into a new add campaign. "Type documents and boil water for tea, all at the same time."
February 10, 2004 5:56:59 AM

>Newcastle is not 90nm, it's 130nm.

[edit]
Newcastle is a redesign; smaller core,less cache, double bandwith (for A64). San Diego is shrink. AKAIK, a straight shrink, but I don't think that is a given.[/edit]

>It's a value line that will be introduced in Q2 2004;

Not correct; its the successor to the current A64, the value line is Paris (likely to be called Duron).

> it's not a true redesign, however, as far as I know. Not
>much will really change in the architecture.

Well.. A64 will gain dual channel, but lose half its cache.FX will gain non registered memory support. Both are supposed to be an improved stepping with an improved memory controller and better SSE performance.

> I think the 3000+ already is a Newcastle core:

No, its still the same old clawhammer die, 193 mm², single channel where NewCastle will be smaller, and have dual channel. 3000+ is just a 3200+ with half its cache disabled.

= The views stated herein are my personal views, and not necessarily the views of my wife. =<P ID="edit"><FONT SIZE=-1><EM>Edited by P4Man on 02/10/04 03:51 AM.</EM></FONT></P>
February 10, 2004 7:37:14 AM

Quote:
Not exaggerated either. No Intel system ever failed on me because of processor instability

Lucky you ! I'm sure this is your first intel system ? You know, I open my deskdrawer, and I see:
-<A HREF="http://x86.ddj.com/errata/feb97/bugs.htm" target="_new">2 buggy Pentium 66 that can't properly do math</A>
-<A HREF="http://www.ebnonline.com/story/chipwire/OEG20000510S001..." target="_new">1 i820 motherboard that has a memory corruption problem</A>
-<A HREF="http://www.sharkyextreme.com/hardware/reviews/cpu/penti..." target="_new">3 P3 1.13 GHz that won't even compile Linux.</A>
-<A HREF="http://www.theinquirer.net/?article=1557" target="_new">1 Itanium 733 that is not working properly</A>
-<A HREF="http://www.itworld.com/Comp/4211/030512itanium2bug/" target="_new">2 Itanium 1 GHz that have to be underclocked to 800 MHz to work stable</A>
-<A HREF="http://www.ocforums.com/vb/vb/showthread.php?s=&threadi..." target="_new"> Half a dozen Northwoods that have died from sudden death syndrome </A>
-<A HREF="http://www.tech-report.com/news_reply.x/3244/" target="_new"> a bunch of Apollo Pro 266A and P4X266 based Pentium 3 and Pentium 4 motherboards that have very serious PCI issues and data corruption</A>
-<A HREF="http://www.intel.com/support/graphics/gaming/82845.htm" target="_new">A few i845G based motherboards than won't run most current games</A>

it looks to me like even VIA based AMD K6 systems are a safer bet than intel.

I am severly limited in what my mind can perceive.
February 10, 2004 7:46:39 AM

>I'm very serious that Athlon 64 simply can't run at 4
>GHz, or let's be fair, at the performance level of a
>Prescott at 4 GHz without consuming at least the same
>power

Bleh.. to achieve the gaming performance level of a 4 GHz prescott, you'd need to underclock the A64 to ~1800 MHz, and which speed it consumes less than 50W on 130nm compared to 120+W for the PresHOT. On 90nm that may be closer to 25W.

I am severly limited in what my mind can perceive.
February 10, 2004 7:51:27 AM

That is obviously not true, and it will not ever be true, except in games optimized for 64 bit.
February 10, 2004 8:07:26 AM

Oh but it will be true, even without 64 bit games. A 3.2 GHz prescHOT is already running at 70°C or more, and that is temp throtteling starts. A 4GHz will only run at 4 GHz while surfing web, and will have to throttle back for anything else.. or melt through its socket and blow up motherboard.

I am severly limited in what my mind can perceive.
February 10, 2004 9:38:39 AM

you may not like what he write for any reason but you dont need to reply with stupid thing

Just to show dad
February 10, 2004 10:51:52 AM

Quote:
I'm sure you adviced them to wait for Prescott a few months ago as well.

I didn't. But I did tell people to wait a few months for Athlon 64 to become available.
Quote:
Yeah let's be fair. The A64 currently roughly consumes *half* the power of a Prescott for equal performance. That is on 130nm versus 90nm for intel.

Half? All I see is an Athlon 64 3400+ consumes 90 Watt. That's hot either way you look at it. And even though Prescott is hotter than hot, it's not 180 Watt. So who's being fair?
Quote:
But I bet you a fortune it was not his cpu, *if* you didnt make up the story, it was more than likely his PSU, or his motherboard.

I did not make up any story, what would be the purpose of that? The name of my friend is Bert Bossyns and he's in his last year of geographical sciences. His case and power supply are the same as a stable 2.4 GHz Intel system I know. His motherboard is based on the nForce 2 chipset.
Quote:
I've worked in this industry about as long as you walk on this earth, and I have yet to see *any* cpu die for no reason.

It didn't die at all. The system just shuts down instantly for no known reason. It has happened three times as far as I know, the last time while running a stability test but there was no indication of what really caused it.
Quote:
They only indicate you're a complete moron, or more likely, lying through your teeth. Why not post a quick photo of all those burnt cpu's ?

I didn't burn those CPUs myself. And stop calling me a moron or a liar it only shows your true nature. I can't post a picture of these CPUs because I don't have a digital camera and because they are at home, not in my university dorm.
Quote:
Says who ? AMD says second half '04. could be august as well as december.

By august Intel will have a 3.8 GHz Prescott on LGA-775 socket. By december it will be 4.0 GHz and above. The best desktop chip AMD will have by that time is the 3700+, which without a doubt will also consume more than 100 Watt.
Quote:
Yeah, at 4 GHz. At a time AMD will be selling 4400+ cpu's.

Strange how the 4400+ was first scheduled for 2003 still... Anyway, don't forget that Tejas will most definitely have higher IPC than Prescott. Probably 4x Hyper-Threading with double the execution units.
Quote:
Why should they? Its not like Prescott is a serious threat, even to old process 130nm A64's. A die shrink is all AMD needs to remain at least competitive, and IMHO, maybe not even a die shrink.

They really need that die shrink to reach over 2.4 GHz, but that will still take many months and by that time Intel will have clearly taken the lead again.

Just look at how things went with Athlon XP vs. Northwood. At first everyone laughed that it was clocked at only 2 GHz like the last Wilamette, but in barily ten months it went to 3 GHz and Athlon didn't scale all that well.
February 10, 2004 11:14:37 AM

89 Watts is the maximum thermal dissipation for all A64's. In reality, the A64 doesn't consume nearly that much, that's just an overly high figure that leaves AMD with a lot of headroom.

DFI i815EP, Intel Pentium III (Tualatin) 1.26GHz, ATI Radeon 9100, SoundBlaster Audigy 2 Platinum, Samsung 512MB SDRAM, Western Digital 80GB 7200RPM 8MB Cache
February 10, 2004 12:10:52 PM

>Half? All I see is an Athlon 64 3400+ consumes 90 Watt

That is not what you see. What you see, is that AMD has given their entire K8 family a TDP of 89W. Even the 1.4 GHz Opteron. You really believe opteron 1.4 consumes anywhere *near* 89W ?

here is a reality check for you:
<A HREF="http://reviews.zdnet.co.uk/hardware/processorsmemory/0,..." target="_new">http://reviews.zdnet.co.uk/hardware/processorsmemory/0,...;/A>. A similary configured prescott machine consumes about the same as the A64 when iddle (cool and quiet disabled), but almost <b>80W</b> more than the equivalent AMD system when under load. Try and explain that with TDP values. TDP is meaningless, its a design specification, and an arbitrary at that.

80W includes PSU ineffiency, guestimated at 50% by some EE guys here before. That means, prescott in real world apps, consumes 40W more than A64, which has to mean Prescott has roughly twice the actual powerconsumption as a A64 (probably around 40W versus 80W) running the same intensive code at comparable speeds. Are you shocked ? you should be. Its only the 130nm hammer you are comparing for now.

>I did not make up any story, what would be the purpose of
>that?

Fudding.

>It didn't die at all. The system just shuts down instantly
>for no known reason.

then its a PSU issue, not a cpu issue. Send me his "unstable" cpu, I'll gladly use it.

>And stop calling me a moron or a liar it only shows your
>true nature

Making wild, unsusbtantiated claims shows your nature. Im not surprised you can't post pics. And if you would have several burnt K7's, it only shows you can't even properly install a HSF. hence, I call you a liar or a moron. Choice is yours. Show me your pics, and I'll retract the liar, and stick to moron if that makes you happy.

>The best desktop chip AMD will have by that time is the
>3700+,

Try a 3900+, which may or may not be leaps and bounds faster than a 4 GHz P4, especially running 64 bit OS/code.

>which without a doubt will also consume more than
>100 Watt.

Try 50W. IBM cut its power consumption in half when migrating from 130nm SOI to 90nm SOI. Even if AMD would not be nearly that succesfull with its very similar 90nm SOI process, I can not believe power consumption would be higher than its current cpu's, which are not even near 89W in real world apps. But you expect AMD to double its power consumption, right ?

> Anyway, don't forget that Tejas will most definitely have
>higher IPC than Prescott

That is what people claimed about Prescott with its bigger L2 cache, bigger L1 cache, enhanced branch predicition, improved HT, SSE3, etc. You got flamed if you wouldnt take for granted a 10 to 20% IPC increase. Lets just wait and see what Tejas brings, I'm not holding my breath.

>They really need that die shrink to reach over 2.4 GHz

Says who ? K7 reached 2.33 GHz with a shorter pipeline, an older design, and without SOI. What makes you think K8 is limited to 2.4 GHz on 130 nm ?? Overclocking results don't tell you a thing with K8, even 1.6 GHz hammers didnt overclock. I see no evidence K8 could not scale to 2.6 or beyond on 130 nm if required. AMD needs a dieshrink to reduce diesize, not to solve thermal issues.

> and by that time Intel will have clearly taken the lead
>again.

Says who ?? You are guessing/hoping, not stating facts here. Not even arguing your reasoning.

>Just look at how things went with Athlon XP vs. Northwood.
>At first everyone laughed that it was clocked at only 2 GHz

No one laughed. northwood performed rather well at 2 GHz, unlike Willamette, nor did it run hot, and it overclocked very, very well.

>like the last Wilamette, but in barily ten months it went
>to 3 GHz and Athlon didn't scale all that well.

Which proves what exactly ? that it *could* happen ? sure, AMD could stumble, sure intel might pull magical rabbits out of its hat, sure both could happen simultaneously, but that doesnt make it either a fact, or even likely. The odds seem firmly against you.

= The views stated herein are my personal views, and not necessarily the views of my wife. =<P ID="edit"><FONT SIZE=-1><EM>Edited by P4Man on 02/10/04 09:19 AM.</EM></FONT></P>
February 10, 2004 12:16:13 PM

Quote:
2 buggy Pentium 66 that can't properly do math

First of all, AMD was no competition at all at that time. Secondly, it's only a <i>slight</i> precision loss for a very limited range of numbers. I'm sure you personally never experiences any problems with it.
Quote:
1 i820 motherboard that has a memory corruption problem

AMD doesn't make chipsets, does it? Should I sum up all the defects on motherboards for AMD? Besides, Intel did replace these faulty boards.
Quote:
3 P3 1.13 GHz that won't even compile Linux.

The problem only occured on a very small number of 1.13 GHz processors, and only with specific code sequences. So it was hardly detectable with automatic checks, but you're right it shouldn't have occured. They did take them back though.
Quote:
1 Itanium 733 that is not working properly

Once more, where is AMD's completely new chip? Intel did identify and solve the problem in a matter of days. It's really a small bug compared to the gigantic work that the Itanium systems were.
Quote:
2 Itanium 1 GHz that have to be underclocked to 800 MHz to work stable

Again it only happened at specific code sequences, and Intel offered to replace any faulty chips.
Quote:
Half a dozen Northwoods that have died from sudden death syndrome

Overclocking is alwasys a risk. At least as many AMD chips suffer from similar problems when extreme overclocking.
Quote:
a bunch of Apollo Pro 266A and P4X266 based Pentium 3 and Pentium 4 motherboards that have very serious PCI issues and data corruption

Entirely VIA's fault.
Quote:
A few i845G based motherboards than won't run most current games

All these games request hardware T&L, even though they would run perfectly with software. Bad game programmers.
February 10, 2004 12:32:40 PM

FUDDER is a fudder, but he has a point, and its ironic he links facts, while you spread fud. you claim AMD chips are "instable", and post some vague unsubstatiated "proof", which even if true, have nothing to do with the chips, but everything with a bad PSU or leaking capacitors on the mainbaord (your neighbour), and user error (your overheating).

He does post several factual links to a lot off known Intel related issues and bugs, and all you can do is say "only very a limited range of numbers, specific code, small error, ..". Now, please, do show us some evidence of AMD chips having stability issue. Anything.. credible. Even with "specific code paths or small ammount of numbers". Show us !

We know early Athlons had no thermal protection, but that hardly qualifies as a bug. We know early VIA chipsets sucked (for either intel or AMD), but you can't blame AMD for that. I only seem to recall a glitch in an AMD chipset that would not allow you to run AGP 2x, only 1x. And that is it. Not bad compared to intel's trackrecord

Now if you have anything else but misinterpreted hear-say stories and a claimed drawer full of dead AMD chips, please share it. Otherwise, stop fudding.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 10, 2004 1:24:37 PM

Quote:
First of all, AMD was no competition at all at that time

that makes the bug not a bug ? No stability problems because amd did not have as fast chip ? Or you mean, if you want fastest intel chips, its okay to have some bugs ?

Quote:
Secondly, it's only a slight precision loss for a very limited range of numbers.

oooh.. a *slight* precission loss.. you mean, the end result was still somewhat correct ?

Quote:
AMD doesn't make chipsets, does it?

It does

Quote:
Should I sum up all the defects on motherboards for AMD

Yes, please do list all stability errors with AMD chips and chipsets. Verie curious !


Quote:
Besides, Intel did replace these faulty boards

good ! good ! so its not an issue, they replace all their bugged equipment. very happy customers I'm sure !

Quote:
The problem only occured on a very small number of 1.13 GHz processors,

Intel only made very small number of these processors.

Quote:
and only with specific code sequences. So it was hardly detectable with automatic checks,

Very, very specific code sequance. No one ever compiles Linux right ? Normal intel did not test that, I agree.

Quote:
Once more, where is AMD's completely new chip?

In the shops, and no bugs so far.

Quote:
It's really a small bug compared to the gigantic work that the Itanium systems were.

LOOOOL.. yeah, VERY small bug in VERY big cpu.. HUGE cpu LOLOL. not so bad no, unless you need huge and hugely expensive cpu to keep factory running.. small bug may mean very big money loss.

Quote:
Overclocking is alwasys a risk

Yes, so is running your athlon with no heatsink.. big risk..

Now, you have some proof of unstable AMD cpu's ? as bad or worse than intel ?

I am severly limited in what my mind can perceive.
February 10, 2004 1:59:51 PM

Are you on Intel's payroll or something? Ok, you are right, Prescott rules, AMD sucks. Intel will dominate the world by the end of 2004, AMD will declare bankruptcy by the end of 2005. Do you feel any happier now? Geeshh!
February 10, 2004 2:47:04 PM

Good war here... :smile:

--
Would you buy a GPS enabled soap bar?
February 11, 2004 11:20:02 AM

Quote:
FUDDER is a fudder, but he has a point, and its ironic he links facts, while you spread fud.

Yeah he links facts, but nothing that would indicate Intel is crap and AMD rules the world. It's not because Prescott is currently not an interesting chip and Athlon 64 has superiour performance that things can't turn around. And looking back at history I don't think it's unrealistic that Intel will produce better results in a matter of months.

In Belgium we have this saying: Tall trees catch a lot of wind. In this context it would mean that every tiny bug or performance loss for Intel is regarded as a huge mistake. But let's not forget Intel does a lot more and has done a lot more than producing Prescott.

Call me an Intel fanboy, I say I'm not. I do see AMD fanboys take advantage of situations like this to claim the end of Intel is near. And that doesn't mean I'm here to claim the opposite...
February 11, 2004 11:34:45 AM

Quote:
Are you on Intel's payroll or something? Ok, you are right, Prescott rules, AMD sucks. Intel will dominate the world by the end of 2004, AMD will declare bankruptcy by the end of 2005. Do you feel any happier now? Geeshh!

See, this is eactly what I mean. It's not because I think Prescott isn't too bad as an intermediate step towards Pentium 5 that I immediately think it's the best chip on the market. Intel has taken steps like this before.

Just think twice. What would be my benefit of claiming Prescott is superiour in every way to Athlon 64? Nothing. And if you take off those green AMD sunglasses you'd see that's not what I'm doing.

And no, I'm not on Intel's payroll, but I do might be on its payroll, just as much as I could be on AMD's payroll, after I get my degree. But that doesn't influence in any way how I think about these companies right now.

And last but not least if I had to buy a new system today, it would most probably be an Athlon 64. Hurray for AMD! But unlike many others in this thread I do realize that situations might change a lot in a few months.
February 11, 2004 12:02:30 PM

>Yeah he links facts, but nothing that would indicate Intel
>is crap and AMD rules the world.

No one even claimed AMD produced "more stable chips" than intel, or intel is crap, <i>YOU</i> claimed the opposite, and are now even forced to deny the complete opposite of what you claimed, and without any proof whatsoever. Therefore, I think this myth has been debunked rather convincingly, wouldnt you agree ?

> It's not because Prescott is currently not an interesting
>chip and Athlon 64 has superiour performance that things
>can't turn around.

I never claimed the opposite, and did anyone ? but I won't accept someone posting this as a given when there is neither evidence, nor even likelyhood.

> And looking back at history I don't think it's unrealistic
>that Intel will produce better results in a matter of
>months.

If you consider only (recent) history, it may not be unrealistic. If you dig a bit deeper, examine roadmaps, and *think* for a change, it does look at least *very* unlikely. A matter of months hu.. ?

And if you want to go by history, look up what happened to IBM, SGI, 3DFX to name just three ex-giants you might remember. If you try and predict the future by ignoring the facts and prospect, but by just looking at the past, you are bound to be wrong.

>In Belgium we have this saying: Tall trees catch a lot of
>wind

In Belgium you also have a saying:"wie hoog klimt laag valt". It seems spectacular and unlikely mighty big Intel would not be able to outperform and outcompete little AMD, but its not all that unlikely at all. In fact, its pretty much a certainty it will happen at some point, if its not by AMD, it will be another company. If not this decade then the next, or the one thereafter.

>Call me an Intel fanboy, I say I'm not.

I say you are, for reasons explained above: blind faith in intel's "resurrection", no proof, evidence or even logical arguments other than "cause its intel", but most of all, this tiresome, typical, "AMD=unstable" fudding. Dreaming up a drawer full of unstable or broken AMD chips doesnt help your credibility either.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 11, 2004 12:09:54 PM

The prescotts a POS deal with it ;) 

(not specifically directed at the above post, just a general comment :-) )

---
"I thought i could see the light at the end of the tunnel, but it turned out to be some ba$tard with a torch and a load more work"<P ID="edit"><FONT SIZE=-1><EM>Edited by ITSALLBS on 02/11/04 09:11 AM.</EM></FONT></P>
February 11, 2004 3:56:34 PM

"...it's only a slight precision loss for a very limited range of numbers. I'm sure you personally never experiences any problems with it."

From this I assume that you only use your computer to play games and web surfing, because for serious work, the processors that had this bug are USELESS.

In 1993 I worked for Oracle Corporation, in an implementation project at a bank. The old NCR dual 486 server was replaced a a new NCR with dual Pentium 66 Mhz one. Compared to the old one, it was faster. Then came the end of the month and the interest rate calculation process. All hell went loose. Interest calculations were incorrect, everyone blamed the software. After countless sleepless nights trying to find an error in the code, we found out that the error was in the Pentium CPU, after comparing the results with those from a Sun SPARC powered server.

It's been more than 10 years since this, and today I still believe Intel only good is for PCs and workstations, but for servers I'll continue to choose any 64 bit RISC processor over anything made by Intel.

__________________________________________________
It's not important to know all the answers, as long as you know how to contact someone who does.<P ID="edit"><FONT SIZE=-1><EM>Edited by gaviota on 02/11/04 12:59 PM.</EM></FONT></P>
February 11, 2004 4:47:19 PM

Interesting gaviota
What you think of the Opteron ?

THG : The last 5-6 reviews have allways either had skewed results, or just somehow strangely come to a completly diferant conclusion then all of the other sites around.
February 11, 2004 6:41:53 PM

It is interesting how excited about all of this everyone gets.

I, generally, tend to sit back and watch what developes.

Often it is fruitless to try to "predict"/"anticipate" what some company that one has not direct contact with or first hand knowledge of.

Last summer, everyone anticipated Prescott as an AMD killer, and that it would set a new standard for CPU performance.

That really hasn't turned out to be true.

No everyone is saying, "Wait until Prescott gets it's speed turned up, AMD won't be able to compete.".

It may well indeed be true, but it would seem unlikely that what we currently call the "Prescott" will ever have its speed raised very much. It already needs to dissipate too much heat, so how much more can that be pushed?

There are many that claim AMD is unstable, but I have never personally found that to be true. (Having purchased, installed, and closely supported several hundred AMD CPU's over multiple years I am comfortable with that conclusion.)

Generally what I have found is those that often buy AMD CPU's are looking for something cheap. They end up slanting everything that their PC is made up of to the cheap end of things. (Motherboard, Video card, fans, power supplies, cases, etc.....) All of these little things together add up to less reliability, and it's the overall affect that is often the cause of percieved stability issues.

Then, of course, it's blamed on the CPU.

I will freely admit that Intel CPU's quite often have proven to be on the tougher end of chip scale. They are, normally, much more forgiving than competitor chips of adverse environments.

I wouldn't, however; call AMD chips unstable though.

Currently, I think everyone would agree, that AMD appears to have the "best available overall CPU".

Will that continue?

We will all just have to wait and see.

These are interestings times, and it may turn out that the MBA's at Intel have morgaged their future so deeply that they are going to slide for awhile. (Remember when Xerox was the scientific computer of choice, or Perkin/Elmer? Remember when IBM's research and developement was more than the profit of the several of the next largest computer manufacturers? Or when Sun had 85% of the workstation market?)

Will Intel fall?

All of that is incredibly complex, and likely even difficult to predict if one had detailed knowledge of their inner workings.

Even at this time of, apparent, decline Intel has vast resources at it's disposal.

We also know that Intel is worried, otherwise they wouldn't have announced at 64-bit Pentium type CPU? What exactly it is remains to be seen. Hopefully ,as some predict, we will see much of this on February 17th.

Depending of how they go about this could well seal the future, or lack there-of, of the Itanium line.

AMD had better be careful, and they should keep a few aces up their sleeves if possible.

I am quite happy to see AMD come up in the world, but like I said, I basically just watch and see what comes out.

Results, products, are always proof of the pudding.

Additionally, I support AMD by buying their products, defending their products, and pushing their products when appropriate.
February 11, 2004 7:23:38 PM

I think you summed that up quite nicely. I agree with every word. i just want to caution against this line:

>Will Intel fall?

I don't think even an AMD hardliner would expect/believe/dream this would happen. Even IF intel would drop Itanium at some point, even IF they would not manage to keep up performance wise with AMD in the high end desktop and x86 server market, the best AMD can hope for for the foreseeable future is becomming a less distant second in marketshare and revenue. AMD simply doesnt have the production capacity to replace intel's hegemony, nor will it have it for the next 10 years or so. Being highly profiteable, getting near 50% of the x86 market and being dominant in the performance sensitive markets is the best they could wish for even in Hector's wettest dreams :) 

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 12, 2004 2:20:24 AM

I think a 25% market share would wet Hector's pants.
February 12, 2004 9:30:02 PM

Honestly this is stupid. I love an intel vs amd war anyday but at this time its not a good time for it. A64 kicks @$$ in gaming and a P4 will mow over an A64 in media encoding. 64 bit software is going to make A64 even better. Better for everyone then. Prescott will scale up and around 3.8 to 4GHz start to turn some heads. YAY! Everyones a fanboy of AMD or intel. You say one word supporting one or the other and ur a fanboy. Hype AMD, hype intel make little comments like 'presHOT' or diss AMD for instablity. Ive had both AMD and intel systems. None have failed. I own 20 computers and none have had a fried chip. AMD and intel both have bright futures. Why argue over which is better when ur comparing apples and oranges? Each is better for different uses. Prescott isnt a piece of sh*t because its 10 fps behind A64 in a game that runs at 150fps. You wont know the difference so whats the big deal there? A P4 will encode a video a few fps faster. Whether it takes 2 hours or 2 hours and 10 mins will you sit there and watch it the whole time? NO, ull walk away and do something else. AMD needs to worry about getting the clock speed up over 2.4Ghz which they can do on 130nm and intel needs to think about getting prescotts temperature down so higher speeds wont be a problem. Dissing a CPU because it runs hot is stupid, if the heatsink that comes with it keeps it running fine why does it matter? Dissing a CPU because it is on a 130nm process instead of a 90nm process is dumb too.

I owned an AMD K6-III 450 laptop for over a year. I'm 16 so this is 1/16th of my entire life and it never gave me a problem. It was slower than my PIII desktop but it kept goin. My PIII desktop I mentioned is still running perfect today since I got it in December 1999. My current desktop, P4 2.4B runs perfect and I have yet to see a problem. I had an Athlon XP 2400+ laptop which i sold because the laptop itself was build bad but the cpu never gave a problem, ever. Same goes for my new P4 HT 3.06GHz laptop. Never a problem. Whats the big deal, a debate is fun but calling people liars and fudders is a bit much. If they claimed A64 or P4 "THE FASTEST CPU YOU CAN BUY, MOWS OVER ALL OTHER DESKTOP CPUS!!!" then they are fudding but by saying they prefer it for specific reasons is not.

I prefer the Pentium 4 because I've used my 2.4B for over a year and its been great. I do video editing and not too much gaming. If Athlon 64 at 4600+ looks to be the best for me in 2005 when I get a new desktop I'll go for it. If P4 at 4600MHz looks good for me in 2005 I'll go for it.

I consider myself to be an intel fanboy. I will say that plain and simple. I spent loads of time ripping athlon xp back when a P4 2.53Ghz with PC1066 RDRAM was top notch. I enjoyed every minute of it. I still like to make jokes about AMD and AMD fanboys. However, I think what I said up there is really what it boils down to. Maybe we take sides because its more fun and lets us fight! I'm not sure but we'll see what happens in the next few months. I'm hoping to see what AMD can do with their 64bit and to see what intel can make of prescott!!!

Ya don't say? Is that so?<P ID="edit"><FONT SIZE=-1><EM>Edited by PCfreak15 on 02/12/04 06:42 PM.</EM></FONT></P>
February 12, 2004 10:28:22 PM

Nothing wrong with buying intel cpus! Espcially if 80% of the computer are selling with them in it :p  Just because you are poor fool with no $$ don't complain about intel cpus :p 
February 12, 2004 11:19:57 PM

anyone who is AMD fan boy is a poor fool. However, Intel Fan are super cool rich dude who get everything they want! BTW I only say "fool" because I don't want to resort to using cuss words :p 
February 13, 2004 9:04:03 AM

A 16 yr old with 20 computers, among which a 3.06 P4 laptop ? Some people...:rolleyes:

= The views stated herein are my personal views, and not necessarily the views of my wife. =
February 13, 2004 10:25:41 AM

What a waste of 5 pages of a once decent forum. I've never seen so much bloody b/s on a forum. I've owned a total of 7 AMD cpu's ranging from a duron 700 to my current A64 3200.

My 700 duron is now in my grandads possesion and after years of use its still going strong.

My 1700 pally XP is now with another person after 2 years its still going strong even though I installed the heatsink on backwards and almost cooked it.

My 1700 and 2100 tbreds are now with happy customers after I tested them for overclocking and sold them on.

My last 2 2500 bartons are now with happy customers after testing for overclocking and both run over 3200 speeds on under 1.8v. 1 even runs @ 2400mhz on just 1.73v.

My current A64 3200 is the best processor I have ever owned, especially since its fairly new tech with new mobos that have yet to mature. Currently running @ 213mhz fsb on my msi mobo and the cpu is sitting @ 35c. Overclocked just 130mhz the cpu performs faster than a 3400 because of the higher fsb and ht speeds etc.

<A HREF="http://service.futuremark.com/compare?2k1=7523662" target="_new">http://service.futuremark.com/compare?2k1=7523662&lt;/A>

Thats on a 9700pro clocked to just above 9800pro speeds.

Been running a few weeks now with not one crash or error in prime. Considering that when I got my 1700 pally a few years ago it wasnt even prime stable I think the new A 64's are very good....



<A HREF="http://service.futuremark.com/compare?2k1=7523662" target="_new"> MY A64 System </A>
February 13, 2004 10:30:33 AM

Just to put things into perspective I have seen a few overclocked P4 4ghz pc's get a slower 3dmark score than my slightly overclocked 3200 even though they have a 9800pro etc.

A guy posted yesterday about his Intel 2.8C @ 4.1Ghz and 9800pro 3dmark score. He got 20,200 where as I get 20,700 with my rig on a 9700pro....... Scroll down about half way for his screen shot and his rig in his sig...

<A HREF="http://forums.overclockers.co.uk/showthread.php?s=&thre..." target="_new">http://forums.overclockers.co.uk/showthread.php?s=&thre...;/A>

<A HREF="http://service.futuremark.com/compare?2k1=7523662" target="_new"> MY A64 System </A>
!