It's true....no 4GHz for a VERY long time.

[Vapor]

The INQ has an amazingly tasty bit of info that I can't reason to be true. However, if it IS true...it's probably the most shocking news of the past year or two.

Wanna know what it is? Intel doesn't plan to *ever* hit 4GHz (except, possibly, in the extreme future). That's right, 3.8GHz is all we'll see (if that). How are they gonna compete (or at least try)? Cache improvements, FSB increases, and eventually dual cores. Wanna read it yourself? <A HREF="http://www.theinquirer.net/?article=19089" target="_new">L'INQ</A>

My speculation:

Can I imagine a Scotty with 2MB L2, 1066FSB @ 3.76GHz beating the AMD 4000+? No, probably not even the 3800+. Okay, what about the 4200+ and 4400+(+) that are in close reach (the 1.8GHz 3000+ 90nm A64 has repeatedly hit 2.5+GHz) of the upcoming A64s (2.8GHz shouldn't really be THAT hard...2.6 is going down at 130nm, afterall...and AMDs 90nm has been surprisingly successful)? I don't even think a P4EE at 3.999GHz with 1333FSB and 6MB of combined L2/L3 cache could keep up with an FX-57 (Q1/Q2 '05), but I may be wrong (I think a Prescott-based core would lose, Gallatin/Northwood might win).

Maxtor disgraces the six letters that make Matrox.<P ID="edit"><FONT SIZE=-1><EM>Edited by Vapor on 10/14/04 06:02 PM.</EM></FONT></P>

 

[wolverinero79]

This was a sort of shock (especially considering how close the current processors ARE to 4 Ghz and anything 3.x Ghz is pretty the same in terms of marketing to the general public). But I don't doubt that it's true. Intel has been hinting for awhile about aggressively attacking multi-core and including large cache. Plus, I think Intel is finally realizing that the consumer PC market doesn't necessarily demand faster frequency chips, or even higher game benching. LCOS, WiMAX, RFID, and other important consumer driven enhancements will probably bring in the money much better than "faster faster faster home pcs". Microsoft obviously is interested in the broader range of focusing on the consumer with their Windows XP Media Center platform.

So likewise, Intel is implementing a new consumer driven approach to the desktop and laptop chips (Centrino was a good idea). And by looking at Itanium's frequency, it's not a huge surprise that GHz will suffer during the transition.

I'm just your average habitual smiler =D

 

[Mr5oh]

I agree.

I don't think we need to process stuff faster, we need to be able to process more stuff at once. (And software support)

My Desktop: <A HREF="http://Mr5oh.tripod.com/pc.html" target="_new">http://Mr5oh.tripod.com/pc.html</A>

 

[mopeygoth]

nice point!

"Pentium 4 666" heh!

Abit IS7 - P4E3.0@3.45- i865PE - 1024mb dual ddr400 - Leadtek 6800 128mb@380/850 - 600W dualfan(front/rear) PowerTek Psu - maxtor 4K080H4 & 6Y120P0 - samsung sm-352b

 

[wolverinero79]

The plot thickens


<A HREF="http://yahoo.reuters.com/financeQuoteCompanyNewsArticle.jhtml?duid=mtfh84275_2004-10-14_19-26-53_n14191720_newsml" target="_new">http://yahoo.reuters.com/financeQuoteCompanyNewsArticle.jhtml?duid=mtfh84275_2004-10-14_19-26-53_n14191720_newsml</A>


I'm just your average habitual smiler =D

 

[raretech]

I tend to agree with a lot of what's been said already on this thread. But, I also wonder if this isn't one way Intel could be dealing with its inventory problem? This would be a good way to convince buyers to buyout existing inventory, since Intel has basically said "it ain't getting much faster than what we got right now"? I wouldn't blame them for taking this approach either.

I just can't believe they won't take it 4ghz, the marketing value of that speed just seems like it would be worth it.

<i>Cigarettes - No cholesterol, high in fiber, low in fat, how could they not be good for you?</i>

 

[priyajeet]

Hey make your subject lines better, I posted a similar article thinking nobody else has.

:tongue: <A HREF="http://www.geocities.com/priyajeet/Fing.jpg" target="_new"><i><font color=red>Very funny, Scotty.</font color=red><font color=blue> Now beam down my clothes.</font color=blue></i></A> :tongue:

 

[Crashman]

I made a new post after seeing both of yours, because I'm going after how this affects the market, not technological problems.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[Crashman]

I proposed, in a new thread directed straightly at processors beyond 4000+, that AMD will likely NOT realease faster processors as planned.

<font color=blue>Only a place as big as the internet could be home to a hero as big as Crashman!</font color=blue>
<font color=red>Only a place as big as the internet could be home to an ego as large as Crashman's!</font color=red>

 

[Mephistopheles]

Obviously, Intel is going to have to ditch netburst completely. Anything else is suicide, plain and simple.

And before you say Intel dead in another subject line, crashman, I think that's not quite true yet. Netburst is definitely dead. Prescott was to scale to 5Ghz, and it didn't even make it to 4Ghz.

Dothan is Intel's only hope. That or another completely new architecture. If I were in their position, I'd flood the desktop market with yonahs - which have already taped out and all - on the LGA775 platform with maximum support for DDR2 (-667, -800) and 1066Mhz FSB. Yonah will most likely hold its own against dual-core offerings from AMD (remember, those will probably NOT clock a lot beyond 2Ghz at all)...

What is true is that the biggest chip-maker has hit a brick wall. How badly they've suffered from it remains anyone's guess; don't take it for granted that "they're dead", though. Personally, I think they could have released i925XE by now if they really wanted to, but they're trying their best to make a platform that will also support their next big thing...whatever that is... which is probably to come in the next 6 (or at the most, 9) months to rescue them from the disaster that is prescott.

 

[ohnoesaz]

I can understand why this seems like a good idea...

But what about people like me who are just gamers? We want speed plain and simple, and no 8 billion skillionbyte L3 cache is going to help us. We need plain speed, not some ability to multitask. Who plays Rome:Total War AND FarCry at the same time? No one. And that won't change when even crazier games come out.

We need speed. Not multitasking. This new approach is gay and I hope AMD doesnt gay it up too.

 

[Mephistopheles]

Multicore isn't only about multitasking. It's about performance, but only for smart programs. And most programs are written in a single-threaded fashion, even if object-oriented.

However, your point is well taken: dual-core isn't going to revolutionize anything. For Intel, though, they could get, say, a fully-feature Dual-core dothan out at 2.13Ghz or 2.26Ghz and still respect the thermal envelope by a long shot - this would be one killer dual-core chip if compared to toledo, AMD's competing solution. Intel must know this by now.

Both AMD and Intel have announced that the Ghz wars are over. We will be in dual-core-land in 2005. And it's about time they reconsider a few things...

<P ID="edit"><FONT SIZE=-1><EM>Edited by Mephistopheles on 10/14/04 10:05 PM.</EM></FONT></P>

 

[raretech]

I guess you've never heard this line before: You are not the target market. I'm not saying this to be rude, this is kind of a notorious line that tech/geeks/nerds hear who have have ever worked for a software company making pathetic ass products on the suggestion of marketing folks, when they complain about those products.

Gamers are just such a small marketshare. I do disagree on one thing, multi-core IS going to help games down the road. And the beautiful thing about it, is assuming the game companies and developers in general make the switch gracefully, these games will scale well as we end up with 3 and 4 cores per processor.

But it will take time. I remember it took quite a while for games to take full advantage of the 386 instruction set. When they finally did though, things like Doom, Wing Commander, and a bunch of other classics that pushed the envelop at the time were the result. Have patience, that's all we can do know. That and hope that AMD capitalizes on this and pushes the envelope anyway. I'd like to see them hit 3ghz and I suspect they can. We'll see.

<i>Cigarettes - No cholesterol, high in fiber, low in fat, how could they not be good for you?</i>

 

[trooper11]

yeah intel was only a few years late figuring it out lol. but better late then never.

i just dont think dual core dothans would be some kind of dual core amd killer, its easy to say that now that we have no idea what amd will really have out when such chips come out. i do see that they potentially will run neck and neck with amd offerings if the proper advancements are made.

but we know that wont happen on desktops for a while, since intel seems bent on keeping p4's going on dual core.

 

[Mephistopheles]

since intel seems bent on keeping p4's going on dual core.

They said so, but I'm not so sure of that... Let's wait and see. I'll be the first one crying out loud if their first dual-core products are netburst-based.

 

[ohnoesaz]

So theyre now into making cpus with dual cores in one cpu chip?

Because I could totally see them doing it based on 2 cpus so consumers have to buy double. Jerks.

In fact I could see them coming out with a Dual motherboard setup. You need 2 motherboards, each with its own matching cpu, matching ram and matching video card. Just so they can sell more.

No wait, how about 2 cpus in ONE motherboard as part of a DUAL motherboard setup. Thats 4 cpus to buy! The possibilities are endless!

 

[Imp]

Games and entertainment may be a bigger part of the picture than we think. Last year, sales of games alone surpassed the entire revenue of all Hollywood movies combined. And when you think about it sure businesses make up a big part of the total computer market, but home computer ownership is becoming almost universal. And volume is a factor that can really add to the bottom line.

 

[Spitfire_x86]

When are we going to see 10 GHz? Didn't clock speed mean everything till yesterday?

------------
<A HREF="http://geocities.com/spitfire_x86" target="_new">My Website</A>

<A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new">My Rig</A> & <A HREF="http://geocities.com/spitfire_x86/benchmark.html" target="_new">3DMark score</A>

 

[TechMan]

====
I just can't believe they won't take it 4ghz, the marketing value of that speed just seems like it would be worth it.

====

Maybe they really wanted to, been doing that all the time. I mean, always introducing those speed bumps to stamp out non-believers of their superior technology.

Just that, somebody else proved to have a better technology at present. Take it with a grain of salt. Netburst failed.

 

[ohnoesaz]

nothing will ever out weigh clock speed. period.

Instead of coming up with ways of increasing speeds instead of by clock, how about come up with ways to cool higher clocked cpu's better.

A Pentium 6ghz would cool great if its spread out over a core thats 3inches by 3inches. Get the picture?

Clock beats all. Cool it right. bishes.

 

[Mephistopheles]

9 square inches? Wow, that's like ~7cm x 7cm, 49cm^2!

I think they'd have electrical signal integrity problems with that...

 

[Ned_Flanders]

BWAHAHAAHAHAHAHA - You are funny

_______________________
Whos round is it? Down that beer quick, smash my glass down, fall over the table - all rowdy and pissed

 

[wolverinero79]

No one plays Rome:Total War AND FarCry at the same time, that's true (or if they do, I think they're nuts ). But take your average strategy game. How would you like one processor to handle the AI, while the other processor handles user input and other items? Or one processor to handle complex calculations while the other better coordinates the transfer of data from RAM to the CPU core doing the calculations (sort of an on-board Northbridge chip).

Multi-threaded applications are harder to program, but not impossible. And what's kind of discouraging is that if you code a program to be multi-threaded and you run it on a processor that can't handle simultaneous threads, you won't see any performance increase. The hardware always needs to exist before the software. The software can sometimes help drive the hardware requirements, but you still need to be able to design for the hardware improvements (which means actually having a piece of said hardware to test against).

Define "plain speed". It obviously isn't pure Ghz (as Intel has found out). AMD has never been about pure speed and they're rocking Intel for gaming. Plus, AMD is slated to have multi-core desktop chips before Intel anyway.

I'm just your average habitual smiler =D

 

[Zoron]

nothing will ever out weigh clock speed. period.

LOL.

Thanks for the chuckle.

Really, now, if this were true, then how do you account for the fact that a much lower clocked AMD CPU competes with and even outperforms a much higher clocked Intel CPU? Clock speed is NOT king, and AMD has been proving that for the past 3 years.

<font color=red> If you design software that is fool-proof, only a fool will want to use it. </font color=red>

 

[ohnoesaz]

"How would you like one processor to handle the AI, while the other processor handles user input and other items?"

Thats an interesting concept.