AMD 90nm vs. .13 micron?

[arnold873]

i am interested in buying a AMD processor
but i noticed they talk about process and they bring up
90nm vs. .13 micron whats the difference?
i`m assuming the smaller the chip surface the quicker the
current can flow? Hence the quicker the processor?

are you going to bark all day little doggie?
or are you going to bite?

 

[Ned_Flanders]

90nm chips have a 45% smaller core size, require less power, run cooler, and overclock better.

Get the 90nm varient if at all possible.

______________
Who's the man with the master plan?

 

[Mephistopheles]

No, the fact that they're based on 90nm processes doesn't mean they're faster, it basically means they probably run a lot cooler.

However, these processors are probably faster than equivalently-clocked 130nm-process chips because there have been some changes. I'm not too sure about this, but I think 90nm chips (well, at least some of them) also support SSE3 extensions. And they're marginally faster because of slight enhancements to architecture.

So there you have it. But the single most striking difference and advantage is much reduced power consumption, which allows for much quieter systems.

 

[Mephistopheles]

Yup, that's it....

....I probably talked too much in my previous post.

You said it all.

 

[Ned_Flanders]

lol, you're right about the SSE3 Extensions - but they are only in the latest Vencie core - and who knows when we will see those on the shelves!

______________
Who's the man with the master plan?

 

[arnold873]

do they actually run faster though?


wow i`m confused
should i spend theextra for 90nm
or go for a faster clock?
a 90nm 3200 costs about the same as a .13 3500


are you going to bark all day little doggie?
or are you going to bite?<P ID="edit"><FONT SIZE=-1><EM>Edited by arnold873 on 03/24/05 03:29 PM.</EM></FONT></P>

 

[Ned_Flanders]

They will run at the same speed at the same rating. The 90nm isn't faster -<i> but you could make it easily....</i>

I'd get the 3200 if your going to o/c. You should be able to get to 2.5 Ghz

______________
Who's the man with the master plan?

 

[arnold873]

i`m not really too crazy about OCing stuff
not that i`m not adventurous but with expensive pc stuff
i`m not adventerous. 😉
thats not to say i wouldn`t let the MB manage the OCing
and if i understand it correctly the asus sli-deluxe mb will automatically do this up to a point.
would it be a good idea to get the fastest 90nm i could afford and let the mb OC it? or will i not gain too much from this?

are you going to bark all day little doggie?
or are you going to bite?

 

[endyen]

From what I recall, an A64-3500 is 1% faster on 90 nanotech. A 3200 would be much slower, so @ the same price, take the 3500.
The current 90 nano Amd chips do not have SSE3 (not that it makes any difference). Most of the current A64s can handle the auto OC that the mobos offer.

 

[mozzartusm]

Arnold, exactly how many questions do you need to ask before you get off your butt and build a system?

<b>Ned Flanders said that im a BAD ASS</b> :lol:
Intel 550(3.4)@4.2
ASUS P5AD2-E-Prem
Ballistix PC5300@DDR2 780
ATI Radeon X800XL
TT 680W PSU

 

[endyen]

All of them.

 

[Ned_Flanders]

Twice.

______________
Who's the man with the master plan?

 

[arnold873]

well just when i think i narrow the cpu down to and 64 3500+
then the issue of 90nm and .13micron
comes up
i guess i`m buying the case and psu prob next week or two
then i`m narrowing MB either sli or single pci-e
nforce4 sata
i think the asus sli deluxe
so by the time i get that board cpu prices will have fallen more. when i first started looking fx-55 was 1060 now its around 860. And so on. i will tell you though my last pc took like 1 day to order parts with research and it was very good. i think i might be taking my time because i`m afraid to buy right now. i feel something totally crazy is about to come out.

are you going to bark all day little doggie?
or are you going to bite?

 

[Clob]

Yeah... Dualcore and R520! R520 isnt untill June I believe

"If youre paddling upstream in a canoe and a
wheel falls off, how many pancakes fit in a doghouse? None! Icecream doesn't have bones!!!"

"Battling Gimps and Dimbulbs HERE at THGC"

 

[mrmonsoon]

Ned,

I would like to know-assuming core design being the same, why a smaller core 90 nm VS 130 nm runs cooler and requires less power from a technicial stand point.

For that matter, I would like to know why the smaller core would be a better overclocker-unless this assumes it runs cooler and requires less power.

Not to pick on you Ned, but if you make statements like that-please back them up with why for my and others understanding

Thank You.

 

[Action_Man]

The 90nm versions make use of strained silicon.

Some people are like slinkies....
Not really good for anything but you cant help smile when you see one tumble down the stairs.

 

[endyen]

As you reduce size of "switches" it becomes easier to change thier state. It takes less potential energy to bridge the gap.
Think of it in terms of a large old knife switch, compaired to the current decora switches in houses. It's a lot easier to physically change the switch.
Smaller uses less energy. Energy = heat. Smaller= less heat.
The less heat, the lower the resistance, so the less power is required.
In other words, a size reduction has an exponential impact on thermal output.
Hope that helps.
me

 

[Ned_Flanders]

The 90nm AMD chips don't use strained silicon.



______________
Who's the man with the master plan?

 

[Ned_Flanders]

Sorry, sometimes I don't have the time.

And I assumed since your doing all this research you might have read an artical from Tomshardware / Anandtect etc that explains it <i>all</i>

______________
Who's the man with the master plan?

 

[mrmonsoon]

Endyen,

Thank you for your clear explaination of the issue. I find myself learning new things everyday, but often feel the need to know why something is different/better....

By being able to explain a concept in simple, clear, plain language, you demonstrate your advanced knowledge of the subject.

Thank you for taking the time to educate me.

 

[Action_Man]

Thats right only the FX-55 uses that now.

Some people are like slinkies....
Not really good for anything but you cant help smile when you see one tumble down the stairs.

 

[RichPLS]

But if the siicon is already strained, would not overclocking it produce more stain on an already stressed product?
And the Prescott got hotter with going to 90nm?!?

<pre><font color=red>°¤o,¸¸¸,o¤°`°¤o \\// o¤°`°¤o,¸¸¸,o¤°
And the sign says "You got to have a membership card to get inside" Huh
So I got me a pen and paper And I made up my own little sign</pre><p></font color=red>

 

[Action_Man]

Yep, so that's why you don't overclock with strained silicon or it'll break in half, nah. All of the silicon isn't strained it's layed across the source and drain of the transistor.

<A HREF="http://www.research.ibm.com/resources/press/strainedsilicon/strain_silicon_02_ico.jpg" target="_new">Picture</A>
<A HREF="http://www.research.ibm.com/resources/press/strainedsilicon/strainedsilicon_right_ico.jpg" target="_new">Another picture</A>

Some people are like slinkies....
Not really good for anything but you cant help smile when you see one tumble down the stairs.

 

[P4Man]

>But if the siicon is already strained, would not overclocking
> it produce more stain on an already stressed product?

ROFL ! Ahem.. no.

>And the Prescott got hotter with going to 90nm?!?

It never "went" to 90nm, it was born on 90nm. And Northwood never went to 90nm, so your only point of comparison for 130 vs 90nm is Banias vs Dothan

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

 

[ChipDeath]

I would've like to see s 90nm Northwood, just for comparison.

That's a great quote from arnie though :lol:

---
Winnie 3200+ @ ~2.5Ghz, ~1.41V
1Gb @ 209Mhz, 2T, 3-5-5-10
Asus 6800GT 128Mb