Sign in with
Sign up | Sign in
Your question

why don't accept the truth?????????

Last response: in CPUs
Share
October 7, 2003 11:49:24 PM

I just wanna now, why anyone of u is 100% intel, or 100% AMD, why? what's the deal???? intel or amd will pay u back??????
whats happening now or later, is that everybody don't accept whats going on.
We have to accept that when the P4 was first introduced, it sucked, and that AMD took a clear lead (for INTEL lovers), and then INTEL took the lead with theire HT technology (for AMD lovers), but now its also clear that AMD has taken the lead again, so just acept it, and don't be like fools to tell that yeah but sooon intel will..., we don't now, cos look at nvidia and ATI, ati was falling behind Nvidia, till the radeon 9700 was introduced, and nvidia is having bad times to get back, even it have more money.
So what im telling here, is not what u love, or the brand, that should make u buy this one or the other, is the price performance AND stability ratio that u have to look at.
All have UP's and DOWN's, so just think.
My first CPU was a CYRIX 33mhz, And for me it was the most stable processor i ever worked on, even when ive upgraded to a 486 intel cpu at 100MHZ, the cyrix was kiking ass, but u now marketing, can make u down.
SO i now the critics that i will have here from all of u, but if intel falls or amd falls, its gona be bad for all, cos witought intel and amd, there will be no competition, and no fast and better cpus like we'r seeing now.

More about : accept truth

October 8, 2003 2:11:22 AM

I'll have to agree with most of it.

I do believe AMD has a slight lead now, though not significant. It's a dilemma between gaming vs multimedia, really, IMO.

And I would not recommend an Athlon 64 or FX any soon yet. I just don't trust any early chips. I'll give em a few months so the bugs are worked and chipsets are more stable.

HEY!
d00d you're in Lebanon right now? I'm Lebanese! I'm in Quebec Canada though, but damn, it's rare as hell to find PC enthusiasts on the other side! I knew of one user from there who did, AMD_Flinkster. Didn't see him since a long time. But hey, pretty cool though!
Ya ahla!

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 8, 2003 3:32:12 AM

Just cause I get checks from Grove and drive a nice BMW doesnt mean I am anywhere bias'd to AMD... not at all...

~Jeremy
Unofficial Intel PR Spokesman.(nVidia fill in rep for CoolSquirtle)

:evil:  <A HREF="http://forumz.tomshardware.com/modules.php?name=Forums&..." target="_new">Busting Sh@t Up In My Buddies Face!!!</A> :evil: 
Related resources
October 8, 2003 3:43:56 AM

Do you think MMX was a great advantage to Intel marketing wise, regardless of its actual performance? Talking about Joe six pack here.
October 8, 2003 12:53:42 PM

As a dumb Joe sixpack back then, whenever I'd read MMX I'd tout it!
I was always reporting my P2 350 as 350MMX rather than 350MHZ MMX. It just sounded cooler.

Now I'm "just a tad bit" smarter. :wink:

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 8, 2003 1:26:56 PM

The Major problem this time around though, is that it's nowhere near as clear-cut as the times in the past you mention.. There is very little between them in the majority of benchies. It's this 'grey-area' combined with the strange, frankly hard to understand phenomenon of 'fanboys' that gives rise to all this commotion.

I think everyone needs to wait a few months really.. 'The Best Chip' is no longer a simple choice - it depends on what you want to use it for (as Eden said above)


---
<font color=red>The preceding text is assembled from information stored in an unreliable organic storage medium. As such it may be innacurate, incomplete, or completely wrong</font color=red> :wink:
October 8, 2003 5:50:39 PM

Quote:
I just wanna now, why anyone of u

Parhelia, it <i>seems</i> like you might have a good point somewhere in your post. Your grammar (or more importantly your lack thereof) however really makes it difficult for me to read.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 8, 2003 6:13:51 PM

Since he did have a good point I'll even goto the effort of translation just to see what you have to say:

I just want to know, why is anyone 100% intel or 100% amd? What's the deal? Will intel or amd pay you back? What is happening now and what has happened in the past is that people refuse to accept the facts. Everyone has to accept that when the P4 was first introduced it sucked, and that amd took a clear lead (For Intel Fanboys). Then intel took the lead with their hyperthread technology (For AMD Fanboys), but now it is clear that AMD has taken the lead again, so don't act like a fool and say that "yeah, but soon intel will..." Honestly we don't know, look at nvidia and ati... ati was falling behind nvidia until the radeon 9700 was introduced, now nvidia is having a rough time even though they have a larger portion of the market. So, what I'm saying here is it shouldn't be what you love or what the brand is that makes you buy a specific part, but rather the price/performance/stability ratio.

Everyone has their ups and downs.

My first cpu was a cyric 33mhz, and it was the most stable processor I ever worked on. Even after my upgrade to a 100mhz intel chip, that little (sniffle) cyrix was kicking ass, but you know marketing, it can take you down.

I know that I'll hear from the critics here, but if amd or intel fails, it's going to be bad for everyone across the board. Without either one there is no competition and then the invovation in the industry will end.

Note: THESE ARE NOT MY COMMENTS, Just translating the original poster...

Shadus
October 8, 2003 7:03:27 PM

Thanks Shadus. It's amazing how much of a difference simple grammar makes. Normally I would try to struggle through it, but I've got a cold and my brain is already fuzzy enough as it is. :\

As to my thoughts on the post itself:
For the most part I agree. I agree that it's silly to be 100% for any company, or even 0.00001% for any company. It is far more sensible to consider current products with current needs (or even future products with future needs) and ignore brand labels entirely.

That aside, I disagree with a couple of minor points:
1) I personally disagree that AMD has taken the lead in <i>desktop</i> performance. Why? Because I <i>still</i> very much disagree with any claim that Socket940 is for <i>desktop</i> use. I see the Socket939 A64FX as a very good future desktop product. I see the Socket 754 A64 as a good move forward for AMD's desktop line right now. And I see the Opteron as a very good current workstation/server release.

That aside however, AMD's renaming (but not even slightly repackaging) of an Opteron to an A64FX does not fool me in any way into believing that it is anything other than an Opteron. It requires a Socket940 mobo. It requires registered RAM. So even <i>if</i> the CPU itself didn't have the hyperthreading channels for multi-CPU boxes (which it does) it <i>still</i> takes a workstation/server <i>platform</i> to run that CPU.

And since it is the A64FX that falsely wins AMD the desktop performance 'crown', I personally refuse to give AMD that win just because their marketing team has managed to play the world for fools.

Don't get me wrong, A64FX is a top-notch workstation/server chip. Or more accurately <i>Opteron</i> is a top-notch workstation/server chip. They're not the absolute best when compared to Itaniums, but still great. They definately beat out Xeons and that's pretty sweet. But in any event, AMD didn't take the lead in <i>desktop</i> performance, and didn't take the lead in <i>workstation</i> performance either. They're strongly in second place, but that's still no lead position.

2) nVidia is having a rough time because they didn't have as much clout as they thought they had. They designed their GPUs around a 32-bit architecture. ATi designed their GPUs around a 24-bit architecture. nVidia thought that they were big enough to push the world into 32-bit. The world refused to budge.

nVidia designed their programmable shaders around their own Cg standard. ATi designed their programmable shaders around Microsoft's DX9 standard. nVidia thought that they were big enough to get the industry to openly accept their standard. The industry didn't. They stuck to Microsoft.

Now nVidia is suffering. They tried to push the industry in two directions and failed on both efforts. Either success would have won them a considerable amount. Instead they're now trying to claw their way back up from the hole they dug for themselves because no one joined them in that hole.

If nVidia had just designed their GPUs around native 24-bit support and had ditched their own shader programmability language for the more logical DX standardization then there wouldn't even be a question of their superiority in FPS.

That aside, I still prefer ATi because they have better image quality and a fanless 9800 Pro from Sapphire. (And hopefully now a fanless 9800 VIVO will join that collection before Yule.)

3) I don't believe in the price/performance/stability as a system of measurement. I believe in a person's <b>NEEDS</b> as a system of measurement for that person's hardware. Sometimes their need is price. Sometimes their need is performance. Sometimes their need is something else entirely. Hardly ever however is everyone's balance between price and performance magically the same. Therefore it's just as useless of a standard of measurement as anything else. <i>Individual needs</i> are what matter.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 8, 2003 8:25:41 PM

thanx for realy understanding, and replying in a verry fair way.
So yeah EDEN im from LEBANON,
walla ma fi el kam wahad byefham bel computers bi hal balad, 2ou chi bi 2ayir, cos hata ma fi 2ela intel, 2ou geforce fx 5200, metel el zbele, gher chi ma fi (THIS WAS IN LEBANESE).
October 8, 2003 9:22:25 PM

wow... and that is supposed to be a language?... I'm amazed... (it's not like it's bad or anything, just curious...)

:evil:  <font color=red><b>M</b></font color=red>ephistopheles
October 8, 2003 9:46:03 PM

Very well said slvr_phoenix *round of applause*. :smile:

<b>P4 2.4C @ 3.0GHz 1.525V Stock HSF * Abit IS7 BIOS v1.3 * Corsair XMS 2x256MB PC3200 2-3-3-6 * GeForce4 Ti4200 AGP8X 128MB * SB Audigy 2 ZS * Logitech Z-680 THX 5.1 * Seagate Barracuda 80GB SATA</b>
October 8, 2003 9:58:25 PM

They don't use an English alphabet (I don't think) so it looks quite strange spelled out in our alphabet.

<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>
October 8, 2003 10:48:14 PM

hehe, yeah we use arabic alphabet in our language, but its a lot easier to spell it in an english alphabet, cos the arabic ones is too hard and realy sucks toooooo much.
October 9, 2003 12:49:49 AM

Do you think this time AMD has this advantage with the whole 64 BIT marketing... just like MMX, like you said it's just cooler!

The whole need for 64 BIT right now is true, we don’t need it… but what more important to AMD right now is the perception of need…

Joe six pack walks into a shop, part of the reason he’s upgrading is because his neighbour has. So keeping up with the Jones, Joe as a choice… buy that same old P4 to match or slightly beat it or go with a new 64 BIT Athlon 64 system. The future…

He might hear things like 64 is twice as fast as 32, It could handle more RAM, you’ll need it for HL2 and Doom, don’t forget Longhorn.

In the end this is where AMD now stands a chance to grow, just like MMX it’s just cooler to have the latest in high speed. For bragging rights mostly and never truly for actual usage.

This is why people over clock 3 GHZ processors…

This is Why ATI right now appears better then NVIDIA

This is why Intel paper launches Emergency Edition.

Speed/New Tech is king… hail AMD…
October 9, 2003 1:41:17 AM

Dude, I agree with your point, I always did think that way. People keep thinking I am bashing AMD64. I think it's a totally excellent way if properly advertised in PC Shops or Electronics Stores like Best Buy, even if it isn't useful for these people.

My point has and always will be, which Bbaeyens disagrees and skips most of the time:
I don't think people should tout the AMD64 capabilities for their 64-bit usage NOW given that almost no home user will need 4GB ANY SOON, but rather the extended register amount that actually gives a performance boost. This is what AMD64 is all about now. But in the marketting aspect, yes AMD should tout 64-bit all the way. But in our segment, we're smarter than that now aren't we?
AMD64 is also on Opteron hence why I specify logically I am making these comments for the home user segment, because in servers, the Opteron is likely flexing its 64-bit muscles.

Bbaeyens then comes to create an argument I never brought to discussion, and then claims we'll practically need 2GB next year. To me next year we'll double the RAM as always, to 1GB high-end systems. Yet he uses his personal experience stigma from work to lay it out here. I am targetting home users and home users only. That can include gamers, enthusiasts and basic users. Consider the majority of sites still use 512MB, and are slowly migrating to 1GB for next year, and my point still stands. I've been on 512MB for over a year now BTW. I feel I'll need 1GB in Springtime, and it'll be more than enough. Games don't grow in size by a multiplier ratio, they grow depending on how they are programmed and how optimized they are. You can easily make sure the game is not memory hungry if you know how to program it right. Carmack states 512MB is recommended for Doom III. I do not expect that to go up any soon, and considering DOOM III is practically THE gaming benchmark to come, as in the real kind of pusher, 1GB should be the norm next year.
Maybe you will agree, maybe not.
Quote:
This is Why ATI right now appears better then NVIDIA

Surely you are not depending the allegations against nVidia here are you?! And supporting them right now?

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 9, 2003 1:44:43 AM

The funny thing is we also use numbers to signify a certain pronounciation from the mouth!

The 2 is short stop, like when you say "Ouch", the way you start the O. If I wrote Di2o, you don't read Dee-oh, but rather Dee-O.

When I went to Lebanon 2 years ago, I was pretty surprised by the smart Arabic to English alphabet translation method they developped. Nice way to make the pronounciation work.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 9, 2003 1:45:34 AM

Heheh, there aren't many AMDs huh?

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 9, 2003 2:58:18 AM

Dude,

--Please stop that… d00d thing!

I agree with your point, I always did think that way. People keep thinking I am bashing AMD64. I think it's a totally excellent way if properly advertised in PC Shops or Electronics Stores like Best Buy, even if it isn't useful for these people.

--Sometimes it appears if you’re on the Intel side you conveniently leave out the positives of what AMD has done (64 BIT on the desktop) This is great strategy in arguments because you can always turns around and say “I never said that! Look up my post and prove it” – I’m not saying you do this Eden, but I’ve seen a few who you back up do it, and of course this can lead opposition thinking your biased. But what I dislike the most is when you're forced to say a positive about AMD because of being call up on, that you must include negatives while admitting to the positive…(getting a headache yet?) I wish people weren’t so defensive.

My point has and always will be, which Bbaeyens disagrees and skips most of the time:
I don't think people should tout the AMD64 capabilities for their 64-bit usage NOW given that almost no home user will need 4GB ANY SOON, but rather the extended register amount that actually gives a performance boost. This is what AMD64 is all about now. But in the marketting aspect, yes AMD should tout 64-bit all the way. But in our segment, we're smarter than that now aren't we?
AMD64 is also on Opteron hence why I specify logically I am making these comments for the home user segment, because in servers, the Opteron is likely flexing its 64-bit muscles.

--I understand, but what is the harm for AMD to push those capabilities? How does keeping that part of their capabilities quiet help them? And again practical usage for 64 BIT for Joe six packs is relative. This is in the same line as “we will never need more than 640k” its true that day when he said it, but makes him look stupid today when you look back, doesn’t it?

There are so many things we don’t need, doesn’t necessarily stop people from buying.

Bbaeyens then comes to create an argument I never brought to discussion, and then claims we'll practically need 2GB next year. To me next year we'll double the RAM as always, to 1GB high-end systems. Yet he uses his personal experience stigma from work to lay it out here.

--It’s true, in most probabilities we want need that much, but a market for that will begin. AMD will be there first, perhaps. I don’t think he was talking about physical RAM anyways… something about the benefits of 64 BIT and virtually RAM and how some games will stress that limit very shortly. I don’t know about this one though.

I am targetting home users and home users only. That can include gamers, enthusiasts and basic users. Consider the majority of sites still use 512MB, and are slowly migrating to 1GB for next year, and my point still stands. I've been on 512MB for over a year now BTW. I feel I'll need 1GB in Springtime, and it'll be more than enough.

--You might feel it’s enough but it doesn't necessarily mean it will be enough for some one else.

Games don't grow in size by a multiplier ratio, they grow depending on how they are programmed and how optimized they are. You can easily make sure the game is not memory hungry if you know how to program it right. Carmack states 512MB is recommended for Doom III. I do not expect that to go up any soon, and considering DOOM III is practically THE gaming benchmark to come, as in the real kind of pusher, 1GB should be the norm next year.

--Games are not the only applications that stresses RAM, they (correct me if I’m wrong) stress the GPU.

--Microsoft, as any one else should, knows how it’s coming down the pipe. Their support of 64 BIT should be your first clue. They are also anticipating a new wave of digital home entertainment network centres to take over our living rooms and from what I read on that, it seems like we can use all the power we can get.

Maybe you will agree, maybe not.

--Maybe not. But my respect for you has increased ten fold. Reminds me of when I started visiting this place, you are very reasonable and can see both sides! Now get the rest of your clan in order.

Surely you are not depending the allegations against nVidia here are you?! And supporting them right now?

--What?

<P ID="edit"><FONT SIZE=-1><EM>Edited by pied_piper2004 on 10/08/03 11:10 PM.</EM></FONT></P>
October 9, 2003 3:44:42 AM

If it's close or better at performing then a "regular" ***desktop chip*** in similar task of a (let’s say it again) ***desktop chip***, then who the f*ck cares?

Why keep stressing this workstation class horse [-peep-].

One design, multi purpose.

The price premium (*chip only*) is in the same realm as the emergency.

Please don't say, "Well new motherboard, new RAM, blah, blah, blah," The freaks that buy this [-peep-] in the first place are insane anyways. MONEY TO BURN if they need the absolute latest and greatest / On another note, both P4 EE and FX are for perceived performance. Who here has the better benchies will be king.

-pied

<P ID="edit"><FONT SIZE=-1><EM>Edited by pied_piper2004 on 10/08/03 11:47 PM.</EM></FONT></P>
October 9, 2003 7:10:58 AM

<<<AMD's renaming (but not even slightly repackaging) of an Opteron to an A64FX does not fool me in any way into believing that it is anything other than an Opteron. It requires a Socket940 mobo. It requires registered RAM. So even if the CPU itself didn't have the hyperthreading channels for multi-CPU boxes (which it does) it still takes a workstation/server platform to run that CPU.

And since it is the A64FX that falsely wins AMD the desktop performance 'crown', I personally refuse to give AMD that win just because their marketing team has managed to play the world for fools.
>>>>

funny u mention that ... but dont mention the XEON/P4 EMERGENCY EDITION
October 9, 2003 1:01:49 PM

Wow, I'm bout to agree with poopy... *dies*

You are correct though, EE and FX are both just repackaged workstation chips.

What I'd really like to see the review sites start including a dual opteron and a dual xeon in each review just to give you an idea of the performance gains and where.

Shadus
October 9, 2003 3:42:16 PM

Can you prove the Xeon's core is the same as the P4's, as in the server component circuits are intact and that the bridges on the chip are the same?

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 9, 2003 4:47:46 PM

Quote:
One design, multi purpose.

The price premium (*chip only*) is in the same realm as the emergency.

If this was a valid argument then why isn't the Opteron 1xx mentioned anywhere? Because the Opteron is for a workstation/server, <i>that's</i> why.

Further, if we're just talking about CPU prices then why aren't the cheap low-end Itaniums involved? Why not? Because we're <i>not</i> talking about <i>just</i> prices, but <i>also</i> about CPU market.

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
October 9, 2003 5:13:08 PM

Quote:
funny u mention that ... but dont mention the XEON/P4 EMERGENCY EDITIOZ

1) You left <i>and</i> were banned (according to you). Why do even feel a desire to stay?

2) It's named the Extreme Edition. You only make yourself sound like an AMDummy to call it an "EMERGENCY EDITIOZ". You don't even see Intelliots calling the AXP an AXPoS or an A64FX an A64FuXer.

3) If you'd actually <i>read</i> my post you'd understand why: "<font color=green>It requires a Socket940 mobo. It requires registered RAM. So even if the CPU itself didn't have the hyperthreading channels for multi-CPU boxes (which it does) it <i>still</i> takes a workstation/server <i>platform</i> to run that CPU.</font color=green>"

Does the P4EE <i>require</i> registered RAM? Is it impossible to run the P4EE on a desktop socket? Does the P4EE even run in a multi-CPU box? Just what <i>exactly</i> makes the P4EE any less of a desktop CPU than any other P4?

For that matter what makes <i>any</i> Xeon different from a P4? Answer that and why the P4EE <i>is</i> a worstation/server CPU at the same time. Go ahead and try. Now try the same thing with A64FX and Opteron 1xx...

<pre><A HREF="http://ars.userfriendly.org/cartoons/?id=20030905" target="_new"><font color=black>People don't understand how hard being a dark god can be. - Hastur</font color=black></A></pre><p>
Anonymous
a b à CPUs
October 9, 2003 5:20:11 PM

>My point has and always will be, which Bbaeyens disagrees
>and skips most of the time:
>I don't think people should tout the AMD64 capabilities for
>their 64-bit usage NOW given that almost no home user will
>need 4GB ANY SOON,

Define "NOW". I buy a cpu NOW to serve as well next year, or maybe even the year after when I spend $800 on just the cpu. there are no applications or games requiring this ammount of speed NOW anyway, so a P4EE or AFX purchase is almost by definition forward looking and to avoid upgrading hassle sooner than needed.

>Bbaeyens then comes to create an argument I never brought
>to discussion, and then claims we'll practically need 2GB
>next year. To me next year we'll double the RAM as always,
>to 1GB high-end systems. Yet he uses his personal
>experience stigma from work to lay it out here. I am
>targetting home users and >home users only.

Your good friend Spud just wrote he is inches away from banging against the 2/3 GB brick wall already now. And what you <b>STILL</b> don't seem to understand is that I am not talking physical ram only, like in actual installed memory modules. You can perfectly well run 1 GB footprint applications with only 256 MB of ram; and even without knowing it, I wouldnt be surprised if you had already. You can not however, run > 2GB footprint apps, no matter how much or how little RAM you have, unless you have a 64 bit CPU, OS and application. So whether or not we will see 1, 2 or 4+ GB desktops next year is irrelevant. What is relevant is that soon enough, a 32 bit CPU and OS will not be able to load certain apps, dataloads (like map editors, DVD authoring, etc) no matter how much RAM you throw in it. granted, this will not affect the large majority of users nex year, but that same large majority has no use for >3 GHz hyperthreading cpu's with 4 different SIMD instruction sets either. The large majority of users doesnt spend >$500 on a cpu anyway, so I don't care about large majorities, its not what these cpu's are aimed at just yet. A large majority of potential P4EE/A64FX customers *will* be able to benefit from 64 bit support, if not now, then at least in the next 6 to 12 months. And unlike new instruction sets a la SSEx or hyperthreading, the benefit will be much bigger than just a few percent better performance. The benefit will be able to run the app or load the dataset, instead of not even being able to run it at all. In my book, that is way more important than a few milliseconds or frames per second you can't distinguish between without benchmarking tool or chronograph.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 9, 2003 6:48:29 PM

Name one program that your average Joe Schmo home user uses that has a footprint greater that 512MB, let alone 1 or 2GB. When you can do that, I will concede the point. All my software runs perfectly well with 512MB of RAM with a 768MB pagefile. If I upgrade to 1GB, I can probably do away with the pagefile... hell I could probably do away with it now and not suffer.

<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>
October 9, 2003 6:51:50 PM

He like's trolling around. That's his life. What else can a Troll do?
October 10, 2003 2:41:12 AM

Whew, finally got around to reply to you!
Before starting, one tip though, use QUOTE tags, it hurts the eyes scrolling through unidentified text, finding your answers to my messages! Use [ quote ] without the spaces, and close with [ /quote ].

Quote:
Dude,

--Please stop that… d00d thing!

But d00d! :wink:
D00d is a funny derivate! I first saw Crashman do it. Say it, it's actually funny! It's cool and I don't consider it 1337 speak (for losers), and it reduces the DUDE I use. :wink:
Quote:
--Sometimes it appears if you’re on the Intel side you conveniently leave out the positives of what AMD has done (64 BIT on the desktop) This is great strategy in arguments because you can always turns around and say “I never said that! Look up my post and prove it” – I’m not saying you do this Eden, but I’ve seen a few who you back up do it, and of course this can lead opposition thinking your biased. But what I dislike the most is when you're forced to say a positive about AMD because of being call up on, that you must include negatives while admitting to the positive…(getting a headache yet?) I wish people weren’t so defensive.

Because you're now someone I respect and isn't on my bad side, I accept your criticism and partly agree as well. But do trust me, I do appreciate AMD's work. It's just that, there isn't THAT much relatively to Intel in terms of what to be excited about. Of course AMD64 rocks and the K8 is great, but Intel's also giving lots of cool things. I dunno if I can still praise Prescott now. The feature list from Spud is quite delicious for my core-architecture-fetish, but if that doesn't translate to much performance, it's hard to accept the excitement. I guess I am excited about the core more than the performance, deep down.
What'd really be a shocker is Intel adopting AMD64 in it. That is gonna truly hurt AMD's advantage considering Intel's marketing power. :eek: 

Quote:
--I understand, but what is the harm for AMD to push those capabilities? How does keeping that part of their capabilities quiet help them? And again practical usage for 64 BIT for Joe six packs is relative. This is in the same line as “we will never need more than 640k” its true that day when he said it, but makes him look stupid today when you look back, doesn’t it?


Snif, I think I might cry soon! Why oh why, and where oh where did you conclude I said such?
I swear man, I've never criticized the fact they implemented freely such features! I am no way against it, just like SSE3 added on Prescott. Even if it's not used now, if it uses so little die space (like 5% according to AMD), and will later translate to humongous boosts in performance, then why the hell should I oppose future-proofness?
It's simply I do not like the criticism praising 64-bit as if it comes now to save us. Bbaeyens will strike back but I stick to my main point, no home user will use 4GB any soon. If 64-bit solves some address virtual memory limit, then that is far from because it's 64-bit integers, that's almost an indirect fix! Kinda like the AMD64 performance boosts of today, from the 8 extra registers.

So again, I suppose AMD's move, despite the fact that I wish IA64 would be the real wave. Their move in the 64-bit marketing will surely help them. I'd like to see AMD64 in adds on Best Buy, or Futureshop in Canada. But I am simply not with the opinion of hyping 64-bit usage any soon for the reasons I gave. Doesn't however mean it's a bad feature or that AMD shouldn't use it. Heck, the way I see it, AMD is saying this: Ok programmers, we're now gonna let you use 64-bit addressing whenever you need it for the customers! PLUS! While we're at it, since moving to 64-bit means coding differently, we've added 8 extra registers. So while you're at programming for 64-bit, you can try the extra registers and get more performance!
Kinda nice from them to do that. It's something only Itanium has done with its humongous 128 registers.
Quote:
It’s true, in most probabilities we want need that much, but a market for that will begin. AMD will be there first, perhaps. I don’t think he was talking about physical RAM anyways… something about the benefits of 64 BIT and virtually RAM and how some games will stress that limit very shortly. I don’t know about this one though.

I know there will be a market. Opteron is that. Then you got the home office user, as in the guy who works at home for his work in addition to the workplace. Multimedia for example, especially big rendering. But still, in general, in fact in a 95+% majority, 64-bit is pointless right now. It's a damn good thing AMD added those 8 registers for sweetening.
Quote:
You might feel it’s enough but it doesn't necessarily mean it will be enough for some one else.

Well, I've covered it above.
I think you as well won't need much RAM more than 1GB in a year from now.
I'm still not entirely convinced modern Windows' technology hasn't overcome that virtual memory issue, though.
Quote:
Games are not the only applications that stresses RAM, they (correct me if I’m wrong) stress the GPU

Of course, however RAM plays the part as a buffer for the files as well. And some of the companies bloat their games using unneeded RAM. That's what I was refering to.

Quote:
They are also anticipating a new wave of digital home entertainment network centres to take over our living rooms and from what I read on that, it seems like we can use all the power we can get.

It's a valid argument that I am waiting though for real life proof. As in, I want to see how soon that vision gets realized. For all we know, the digital home center is gonna be the icon of 2005 like HDTVs! So we'll append this to the future's list, although I'll agree with you. As it is logical.
Quote:
Surely you are not depending the allegations against nVidia here are you?! And supporting them right now?

--What?

You sounded like you made ATi to be a faker and nVidia to be badly treated in including ATi in your list in a post previously.

But to finish things, I've said it before, I get lost in my reasoning when I talk to people I can't get along with. They get the best of me even when I try not to. When I talk to you now, and you present counter-arguments, I can totally reply back with no aggressivity. I don't mind that. I've had tons of debates against Slvr, whom I respect and even agree often with. Since I've had a different mind, me and him debated with no aggressivity whatsoever. Same thing with others here. So hopefully we'll continue like that too!

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
October 10, 2003 2:50:18 AM

Good point,
Really guys, now put the top of line amd cpu in one computer, put top of the line Intel in the other, have the same components, & operating system.
Put them behind a wall,(hide them) connected to 2 of the same monitors and how in the hell can you tell which monitor is connected to Amd or Intel cpu??
Can anybody tell the difference?
I imagine 99.9 % of the people couldn't tell.
So for me it is always the best bang for the buck, without any hardcore overclocking.
My checkbook gets my Loyalty.
Bob
October 10, 2003 2:57:54 AM

I currently disagree with Slvr on not giving the A64 FX credit, but agree with him it is essentially a poor attempt at turning a class server CPU into a Workstation one with renaming.
It's a desktop contender nonetheless, although I can see through its name!


--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
Anonymous
a b à CPUs
October 10, 2003 9:25:16 AM

Just one ? okay: UT2k3 map editor. And Epic is on record for saying the map editor of their next UT game will not just benefit from a 64 bit addresspace, 64 bit addressing will be a <b>requirement</b>

= The views stated herein are my personal views, and not necessarily the views of my wife. =
Anonymous
a b à CPUs
October 10, 2003 9:35:37 AM

Oh dear.. how can one argue with someone that claims to suffer from a "core-architecture-fetish" and then makes statements such as this:"If 64-bit solves some address virtual memory limit, then that is far from because it's 64-bit integers, that's almost an indirect fix!"

Ouch..

May I suggest you do some reading, like <A HREF="http://www.arstechnica.com/cpu/03q1/x86-64/x86-64-1.htm..." target="_new"> this Ars technica article</A>

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 11:23:39 AM

Quote:
But the difference is you will be able to buy P4EE's shortly A64-FX's are all bought up.

Funny, FX-51's are in stock at newegg. Seems A64-FX's are <i>not</i> "all bought up," and the P4EE has never been available enough to get bought at all. :tongue:

Both are rebadged server chips. It's just that the FX-51 is overall a slightly better performer and isn't nearly so much a paper launch.

<i>I can love my fellow man...but I'm damned if I'll love yours.</i>
October 10, 2003 12:56:39 PM

talking about the A64 and the p4 (EE or not, its the same for me, just some l3 cache).
what new AMD did???????????????????????
its simple, first of all the well known 64 bit, but thats not all, all of u tell that 64bit is not for now and rarely for future use, and its not worth it to buy an A64 for its 64 bit support.
But for me, if we can just forget 64 bit now, and go to 32 bit, so we see that the Athlon 64 3200+ is outperforming, and i mean it, the P4 3.2GHz C, and with a lower peice, withought using what u called a SERVER chip, and withought registred rams.
And the fx is also faster than the P4EE, and also at lower price.
The new thing AMD did is not only the 64 bit suport, and the 32 bit suport withought emulation.
but the memory controller introduced in the processor itself.
what i mean here, that we don't have the fear of the memory bandwith with some chipsets, the good here that the memory is connected directly to the cpu, that is the NEW thing for me, more than the 64 bit support.
and for the FX cpu, all of u are critisizing that its a server processor, and it needs registred rams, and and...... so its not for desktop.
where is the problem with that???????? yes its a server processor, like P4EE witch for me is a XEON. but where is the problem if u buy registred rams, and a server processor, and use it as a desktop?????????? i mean for me, its a good move, to get the power of a server on desktop.
AND im just curious of something, why INTEL didn't ever increased the L1 cache????????? the L1 cache is the faster, then comes the l2 than comes the l3.
So if they increase it, i think they will have a good boost in performance, cos the AMD always had 64kb cache in the L1, and the INTEL is only 16KB cache in the L1.
Im not telling here that AMD processors are a lot better than intel ones, but they are indead more advanced technicaly. So for me AMD is in the lead now, and for sure.

but for marketing wise, 80% of peaple will not buy either the A64 and the P4 EE and the p4 3.2 c, cos there are too expensive, lots of peaple buy pcs for office use, and thats the good marketing of intel here, they are giving celerons, but AMD stoped the durons, so they should react.


<P ID="edit"><FONT SIZE=-1><EM>Edited by Parhelia on 10/10/03 04:32 PM.</EM></FONT></P>
Anonymous
a b à CPUs
October 10, 2003 1:20:06 PM

>But for me, if we can just forget 64 bit now, and go to 32
>bit, so we see that the Athlon 64 3200+ is outperforming,
>and i mean it, the P4 3.2GHz C, and with a lower peice,
>withought using what u called a SERVER chip, and withought
>registred rams.

I agree, the A64 really is the hot chip, the FX and soon the P4EE are just for bragging rights and winning benchmarks. A64's currenly sell for <$400, thats one hell of a sweet price for such a powerfull cpu.

>and for the FX cpu, all of u are critisizing that its a
>server processor, and it needs registred rams, and and.....
>. so its not for desktop.
>where is the problem with that????????

Problem is that registered ram is more expensive, and hardly available at DDR400 speeds. It is also slower than non registered memory. Its not ideal IMHO, even though prices will come down, and the performance issue is just theoretical, as even with its 'slow' memory, its still the fastest desktop chip money can buy.

Lastly, lets not forget registered ECC memory also has its advantages: better stability. I don't see anyone touting this, while many people critise AMD's "unstable chipsets" and such. In the past, they would use intels perceived higher stability as a reason to justify higher prices, why isnt that same argument applied to ECC memory ? An Athlon FX is probably the fastest and most stable platform you can buy these days. Yes, its also the most expensive at least until the P4EE hits the market.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 5:19:15 PM

Quote:
Just one ? okay: UT2k3 map editor. And Epic is on record for saying the map editor of their next UT game will not just benefit from a 64 bit addresspace, 64 bit addressing will be a <b>requirement.</b>


When will this next UT game be available? Next year? The year after? Three or four years? I'll admit that I haven't actually used the map editor for UT 2K3, but are you telling me that it requires more than 2 or 3 GBs of address space? Or that it even requires 1GB for that matter?


<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>
October 10, 2003 5:23:43 PM

Quote:
Oh dear.. how can one argue with someone that claims to suffer from a "core-architecture-fetish" and then makes statements such as this:"If 64-bit solves some address virtual memory limit, then that is far from because it's 64-bit integers, that's almost an indirect fix!"

Ouch..

May I suggest you do some reading, like this Ars technica article


Oh dear me, how could I EVER be myself. I guess loving nature and finding out about it, (or simply taking photos endlessly) but not getting informed on every single bit hurts huh?
Please forgive me father, for I have sinned by not taking up to your godly standards.

That was a pathetic try Bbaeyens, but a nice failed one.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
Anonymous
a b à CPUs
October 10, 2003 5:35:18 PM

>When will this next UT game be available? Next year? The
>year after? Three or four years?

Next year if all goes well. Its called UT2004 after all :D 

> I'll admit that I haven't actually used the map editor for
>UT 2K3, but are you telling me that it requires more than 2
>or 3 GBs of address space?

You only asked me for one example exceeding 512 Mb :) 
But yes, depending on the size of your map/mod/project you'll easily go beyond one gig. i've heard other map editors like morrowind are even more memory hungy.

one thing to remember: current games, maps, editors are designed with x86 limitations in mind. level designers have to restrict themselves when creating textures, on polycounts, etc. Artists and level designers would probably love to be able to use several gigs worth of textures instead of reducing them in size and creating blurry small compressed textures. Don't be surprised next year to see some games making use of the extra addressing capabilities and eg run with much larger textures on 64 bit platforms, and with less eye candy on 32 bit systems. I look forward to see the reactions of gamers that just spend $900 on a cpu and $500 on their videcard, only to find the "use ultra high res texture" checkbox greyed out on their system.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 5:47:09 PM

> Or that it even requires 1GB for that matter?

All options maxed Anarchy Online exceeds that. Hell I think most 'modern' mmp games do.

Shadus
Anonymous
a b à CPUs
October 10, 2003 5:54:38 PM

> but not getting informed on every single <b>bit</b>
>hurts huh?

You missed 32 bits of which you don't seem to understand one bit ;) 

Look eden, I don't expect everyone to know everthing, god knows I only know a tiny little bit, but I will not make bold claims on things I know <i>nothing</i> about; if I do not know, I will simply ask and be gratefull to get answers from, or be corrected by people who know a lot more than me like Imgod2U. However, I do get iritated when I get 'corrected' over and over again by someone who doesnt have a clue, but is simply venting an unsubstantiated opinion that better suits his preconceived idea's.

You have been posting at least 300 posts about why 64 bit wouldnt matter, and no one should care when deciding between a A64 and a P4, while you really havent grasped the <b>fundamental</b> basics, and I do mean basics. Next time, I suggest instead of claiming "touting 64 bits is silly at best" you rephrase that into somehting like: "why would 64 bit be usefull in the near future ?".

And not claiming to have a core-architecture-fetish while you don't know the difference between 64 bit registers, 64 bit pointers or 64 bit integers would save yourself from some embaressament as well.

Lastly, I have posted oodles of information on virtual memory, 32 bit pointer limitations and PAE that if you took the time to actually read them, should stop you from saying things like" <b>If</b> 64-bit solves <b>some</b> address virtual memory limit, ". By now you should at least have understood there is no need for the if or some. And by now you should also have realized that 64 bit is not only about supporting more than 4 GB of RAM, yet every post from you I read on that topic, you will still use the 4 GB Ram as an argument to question the necessacity of 64 bit. I you do not want to learn, fine, but don't expect me to be nice after trying to educate you, and don't expect me not to correct you over and over again.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 8:59:07 PM

It's not that, that is the problem, it's that I haven't seen some real links with known issues on the virtual memory. You can state all the problems you want, but that still won't help me figure out if this is a universal issue that even the most recent Windows' can't fix. Heck I am curious to know what Slvr or imgod2u has to say on this. Usually the latter would be picky and find something to say about it that you don't say.

About that bold claim now, I found it personally only a logical issue. I simply don't grasp why a 32-bit register, and addressing, which entitles you to 4GB of addressing and 4 billions of integers (or was it milions?) to work with, cannot get around some virtual memory issue, and that a processor with extended register size suddenly can. You may have stated why, but I missed it, as I did read everything you wrote. Didn't help because, as I said, I need some firm sources to back up that no current Windows OS can get around this issue, and if you have an explanation why 32-bit x86 CPUs cannot fight that virtual mem limit, I simply cannot accept it.

I often ask for links. Now I don't know if the Ars article covers it. I used to read quite some things from them, but I never got around to read that one, and I certainly can't find much time lately. Ah if only PDF versions were possibly to be ordered in paper form. Sitting by the fire reading about registers, how great.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
Anonymous
a b à CPUs
October 10, 2003 9:47:40 PM

So the point is you don't know, but chose not to believe me even though no one else is contradicting me .. hmm.. oh well, okay, here is link for ya:
<A HREF="http://www.microsoft.com/whdc/hwdev/platform/server/pae..." target="_new"> MS on PAE </A>

>About that bold claim now, I found it personally only a
>logical issue. I simply don't grasp why a 32-bit register,
>and addressing, which entitles you to 4GB of addressing and
>4 billions of integers (or was it milions?) to work with,
>cannot get around some virtual memory issue, and that a
>processor with extended register size suddenly can

Hu ? Its quite simple.. 32 bit cpu's use 32 bit pointers (stored in 32 bit ALU's), which only allows for 4 GB worth of adressing (do the math 2²32). Extend the ALU's to 64 bit, and you can suddenly store large enough pointers to differentiate between 18 million terrabytes. An OS can not support something that the chip doesnt support.

Also, the cpu doesnt really make a difference between physical RAM and virtual memory (paged to disk). Each memory location simply gets a pointer, that pointer is limited to 32 bits on a 32 bit cpu. So actual RAM + virtual memory < 4 GB.

What PAE allows is actually comparable to paging memory EMS/XMS style. PAE enables several apps to see 4 GB adress space, and the OS "shuffles the memory" so to speak that each app/process can have its own 32 bit adress space. This is very tricky to program, its slow, and it still doenst cure the 2/3 GB / process limit. It can be usefull though when running several huge server apps on a single machine. Won't cure your UT editor problem though.

But don't take my word for, see the MS link.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 10:10:25 PM

All right, that's sound proof.

Now my question would be, why hasn't Microsoft so far simply just made their OS to 3GB physical and 1GB virtual in the first place?
Is WinXP any different, and why must they need a 64-bit processor with a 64-bit OS to finally get around that limit? Or is it just because of the inevitable high amount of memory hungry Windows bloating that will eventually come with Longhorn + the apps installed and loaded all the time?

I'll admit you now have a point, but I don't quite fully grasp it.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
Anonymous
a b à CPUs
October 10, 2003 10:30:29 PM

>All right, that's sound proof.

Geez.. hurray.. at least. Look man, I am no tech God, but I don't lie through my teeth either. About time you started getting the point.

>Now my question would be, why hasn't Microsoft so far
>simply just made their OS to 3GB physical and 1GB virtual
>in the first place?

Oh wait NOOOOOOOOOooooooo.. you still didnt get it !! :( 
Read it again.. read the Ars article I linked earlier, honestly, its a good read. sigh.. oh well, since God has charged me with the divine task to talk some sense in these forums, let's just carry on:

XP doesnt have a "3 GB physical" / "1 GB virtual" limit. It has a 4 GB limit, just like the 32 bit cpu's it is supposed to run on. This 4 GB can be 128 MB physical and 3.8 GB virtual or 4 GB physical, that doesnt matter. Maximum virtual memory size is just 4 GB minus physcial RAM or virtual memory + physical ram <= 4 GB.

This 4 GB address space is divided into 2 areas: one for the OS itself (kernel, drivers, page tables for swap file, ..) and one for the applications to do whatever they want. Historically, this division has been made at 2 GB, since "2 GB is enough for everyone, both OS and apps". there was no real reason that I can think off, 2 gigs just seemed plenty. As apps began to outgrow those 2 GB, a quick and dirty solution was to restrict the ammount of address space for the OS to 1 GB, and free the rest for apps. This works as long as the system doesnt have more than 12 (or 16 ? not sure) GB of memory and the application is aware of this extra GB to use. Its a temporarely fix that has been required 5 years ago.. go figure.

anyway, I am not going to type an entire article here; there are plenty of good reads on the web, I do not claim to be able to explain it any better, so I suggest you do some reading.

>Is WinXP any different,

No, and neither is Linux (32 bit), Solaris, FreeBSD, OpenBSD, NetBSD, AIX or even MacOS AFAIK. OS's are written to take advantage of the features of their host cpu, but they can't magically increase what the CPU can do.

> and why must they need a 64-bit processor with a 64-bit OS
>to finally get around that limit?

Well, 64 bit registers and addressing are just the most important feature that makes a cpu (and OS) 64 bit instead of 32 bit. So what you are really saying, is:"why didnt they release a 64 bit OS and cpu any earlier", and thats a pretty valid question.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 10, 2003 10:59:54 PM

Quote:
Oh wait NOOOOOOOOOooooooo.. you still didnt get it !! :( 
Read it again.. read the Ars article I linked earlier, honestly, its a good read. sigh.. oh well, since God has charged me with the divine task to talk some sense in these forums, let's just carry on:

XP doesnt have a "3 GB physical" / "1 GB virtual" limit. It has a 4 GB limit, just like the 32 bit cpu's it is supposed to run on. This 4 GB can be 128 MB physical and 3.8 GB virtual or 4 GB physical, that doesnt matter. Maximum virtual memory size is just 4 GB minus physcial RAM or virtual memory + physical ram <= 4 GB.

This 4 GB address space is divided into 2 areas: one for the OS itself (kernel, drivers, page tables for swap file, ..) and one for the applications to do whatever they want. Historically, this division has been made at 2 GB, since "2 GB is enough for everyone, both OS and apps". there was no real reason that I can think off, 2 gigs just seemed plenty. As apps began to outgrow those 2 GB, a quick and dirty solution was to restrict the ammount of address space for the OS to 1 GB, and free the rest for apps. This works as long as the system doesnt have more than 12 (or 16 ? not sure) GB of memory and the application is aware of this extra GB to use. Its a temporarely fix that has been required 5 years ago.. go figure.

Ok aside from the fix thing.
Why then, if you have that 3GB extended limit, should there be a problem for any future apps within the next year? I understand some games seem to ask for it in map editors, but like I said many times, consider these at the highest cases for home users. Yes there are other kinds at home, but I'll be more open and add the map editor limits lately encountered. So suppose those could also access the 1GB Windows reserved space assuming Windows hasn't used that much with its apps.
Tell me now, why would you see THEN such a limit to be annoying? That is assuming programmers have gotten the hang of the extra 1GB limit from the physical RAM.
I personally don't feel convinced apps are that limited, but that is only if we consider 3GBs. If it's 2GBs, I'm pretty much convinced assuming there are many applications asking for it. (and please, leave the @work applications stuff)

Quote:
Well, 64 bit registers and addressing are just the most important feature that makes a cpu (and OS) 64 bit instead of 32 bit. So what you are really saying, is:"why didnt they release a 64 bit OS and cpu any earlier", and thats a pretty valid question.

No you misread here.
I am asking, or WAS asking why would 64-bit in itself help remove a 2GB limit. But it sounds now that it was a choice by the big companies, rather than to start from scratch with a 3/1 GB assignment since Windows' files would really not go that high any soon.

My perception is now that if it was a reservation thing, well it's their own fault, but in reality if it were a free-for-all space usage all the while Windows has first hand rights on memory usage, then I would continue stating that 64-bit is not exciting because no user will be using so much memory any soon. Looks like I can't anymore.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol: 
Anonymous
a b à CPUs
October 11, 2003 6:57:46 AM

>Why then, if you have that 3GB extended limit, should there
>be a problem for any future apps within the next year?

Because the /3GB isnt practical; it requires users to run the switch (no one does), it requires apps to be recompiled/written for it (why not rewrite for 64 bit while you are at it ?) and it can give a ton of problem with drivers and apps. If the switch gave you 30GB instead of 3 (and in most cases, that 3 isnt even 3 but closer to 2,5. Among other things, I thought the videoram or AGP aperture needs to be substracted from the full 3 gig), it might be worthwhile persuing, but as it is, no one is bothering except for one or two server apps like SQL server and exchange where it runs in a controlled server environment on a computer handled by an admin.

>My perception is now that if it was a reservation thing,
>well it's their own fault, but in reality if it were a
>free-for-all space usage all the while Windows has first
>hand rights on memory usage, then I would continue stating
>that 64-bit is not exciting because no user will be using
>so much memory any soon. Looks like I can't anymore.

I don't understand what you are saying here, but fact is the OS really needs the 1 gig, and there is no guarantee future OS's like Longhorn will not need more. That only leaves so much for the apps. Also keep in mind, unless you are runnning PAE (which you are not) that all your background apps eat into the remaining 2/3 GB, including firewall, antivirus, file sharing apps, IM, your email client, browser, one or two office apps. Don't be surprised you only have 1 gig left when you launch that game, map editor or authoring package. If you would run PAE, prepare for a really serious performance hit.

Bottom line: 4 gigs is still workeable for most people, in most cases, but now really is the time to prepare for 64 bits, and if you are buying a new platform for the next few years, I wouldnt just ignore this; especially since affordable 64 bit cpu's and OS's are here, expect among others game developpers to make use of the extra capabilities. I wouldnt want to buy into an expensive platform that wouldnt allow me to leverage this. SSEx or hyperthreading can me missed, after all, they "only" speedup certain things, AMD64 may soon be a matter of being able to run it or not, there is no slower workaround.

Also keep in mind every other platform has moved to 64 bit now, Sun (sparc), IBM (power), Alpha, MIPS, Intel (Itanium) even Apple. ISVs writing cross platform applications are almost guaranteed to be writing 64 bit clean code now, which will really help get those 64 bit applications sooner rather than later. If it wherent for intel's market share, I could even see them not bothering with 32 bit x86 ports anymore in the near future.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
October 11, 2003 7:36:10 PM

By that logic that still doesn't prove that the true nature of 64-bit is what helps alleviate the issues. It only shows that you can then extend by millions the virtual+physical addressing, when in reality if the 3GB barrier was conceived right from the get-go for apps, it wouldn't be an issue before 2005 likely. As for my reference to the 1GB, most computers have about 200MB of commit charge for Windows loaded with some startup programs. How can that even near the 1GB usage any soon?
My system is a 2-user one, bloated with open programs and background ones, and often reaches minimums of 400MBs without gaming. But that's because of the programs open, not Windows itself.
Longhorn will use a lot of resources, but since the tradition has always followed a doubling of usage, I don't expect it to utilize any more than 300MBs whenever it's loaded. I still don't see, in an inexistant 3/1 scenario, how the 1GB would soon go down, causing lots of troubles.

My point of all this is to tell you that this still doesn't make the TRUE USE of 64-bits to be removing the problems made my Microsoft or whoever created such norms of 2/2. The true use remains that you can calculate 64-bit integers and extend programs size usage to over 4GB and use more RAM. So while you did show me what the problem is, when speaking in technical on the 64-bit technology itself, this does not in any way in my eyes translate into its true purpose, but, kinda like AMD64's extra 8 registers, an extra side-effect that helps a lot in x86 or other 32-bit platforms doomed with such norms.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>This just in, over 56 no-lifers have their pics up on THGC's Photo Album! </b></font color=blue></A> :lol:  <P ID="edit"><FONT SIZE=-1><EM>Edited by Eden on 10/11/03 03:38 PM.</EM></FONT></P>
Anonymous
a b à CPUs
October 11, 2003 9:23:56 PM

>My that logic that still doesn't prove that the true nature
>of 64-bit is what helps alleviate the issues.

Okay thats it; God knows I tried, but if you don't want to listen, do not understand, do not want to do any reading on the matter to educate yourself, fine..But I'm not ready to keep discussing consequences of things you don't understand the fundamentals off while you refuse to take my word for it, so why don't you just keep spouting those 4 GB claims and get yourself one of those nice 32 bit CPU's next year.

= The views stated herein are my personal views, and not necessarily the views of my wife. =
!