.09 process AMD vs INTEL

[POPEGOLDX]

WIll AMD struggle with the .09 conversion like INTEL is right now.

There are 2 trains of thought

1. Since INTEL is making CPU TOASTER OVENS... its automatic that AMD Will too. Lets face it , INTEL has the money and engineers while AMD has bubblegum and popsicle sticks.

or

2. AMD will gain knowledge from INTELS initial failure and so will the Machining tool companies of the .09 process as will IBM in helping AMD with .09. The transition will be smooth

My feeling is that .09 hammers wont bake pies like .09 prescotts or dothans , but it will take a lil while for AMD to get .09 perfected similiar to the first t-breds vs the second stepping T-breds

<<< people who just come to this forum to insult me insert insult here.... ( ) ......... >>>>

others who have opinions on this please enlighten me

 

[traviss187]

i posted (I think at least) a desent respnse in a different thread, but to lazy to copy it over here,
I don't think that the process will be supper smooth for amd, yes they will be able to learn some from intel's troubles, but companies arn't going to jump to help amd when the revenue they stand to make from helping them doesn't even begin to compare with what they could lose if they lost the next contract from intel.
also the design and fab changes that they have would have to undergo in transition costs tons of $$$, which amd doesn't have tons of. to convert one fab can cost tens of millions of dollars, if you check finacial news from the last couple of years, you'll see news each time intel upgraded a fab, and the exact amount they spent on it, it's gigantic, not a small undertaking, and seeing the exteme difficult intel is having, I can't imagine amd not having at least as much trouble
but hey I could be wrong, amd might be able to get the process down, BUT even so it'll cost them, and right now they're finacial outlook doesn't look great, eps -4. somthing

 

[Mephistopheles]

Interesting is that it is quite likely that Intel can use AMD's attempt at 64-bit desktops in that way... learn and let AMD spend the initial cash on 64-bit software.

<font color=red><b>M</b></font color=red>ephistopheles

 

[eden]

Then explain to me why their conversion will go well when their 0.13m process took them months before they had it working right, compared to Intel, who also ran into problems?

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=green><b>A sexual experience like never before seen</font color=green></b></A>
Site has now even more sexy members, for your pleasure.

 

[Groveling_Wyrm]

Yet another thought to pass on (or maybe an expansion on the first thought you have) is that is that AMD processors, by their nature, have a higher IPC, thus better "efficiency", thus making them hotter than a comparable Intel, but now that Intel has HT enabled, and upped IPC on P4's the "heat" is on...

Thus, by this analogy, it could be easy to guess that AMD will have serious heat issues, just as bad, if not worse than Intel will.

However, the question that remains is how high can the heat go before they have to start making water cooling solutions commonplace in mainstream systems?

<font color=blue> Ok, so you have to put your "2 cents" in, but its value is only "A penny's worth". Who gets that extra penny? </font color=blue>

 

[eden]

I wholeheartedly doubt Watercooling will take the mainstream.

Joes won't and will never want to maintain the inside of their PC with water rechange, if their cases most of the time are left full of dust for years.

Oh and finally someone else believes like I do, that AMD will run into issues too.
It's almost a given they will. But then, I have fanboys making me out to be anti-AMD with such statements. Jeez.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=green><b>A sexual experience like never before seen</font color=green></b></A>
Site has now even more sexy members, for your pleasure.

 

[Groveling_Wyrm]

I agree. I hope watercooling will not become mainstream, but what other options will be available, considering the current trend in size/heat issues for cpu's? I definitely see a dilemma coming soon for AMD and Intel.

[looking into future] 2010- "Darn it! the heatsink and radiator on my water cooling system is so large that I had to put it outside the window, increase the size of the water pump, and run a 2 1/2 inch water line to the HSF!" [/looking into future]

<font color=blue> Ok, so you have to put your "2 cents" in, but its value is only "A penny's worth". Who gets that extra penny? </font color=blue>

 

[IIB]

Historicly speaking:
AMD had troubles moving to 0.25u, Intel didn't.
Intel had troubles moving to 18u, AMD didn't.
AMD had troubles moving to 0.13u, Intel didn't.
Intel possibly has problem moving to 0.09u, fill in the missing.


This post is best viewed with common sense enabled

 

[simwiz2]

Is there a reason for that (like new techniques needed every other process change, etc) or is it just a lucky pattern that may or may not continue?

 

[eden]

Well, it's only an issue now, they have not perfected the silicon at all.

AMD's SOI can come in handy as well.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=green><b>A sexual experience like never before seen</font color=green></b></A>
Site has now even more sexy members, for your pleasure.

 

[IIB]

Well, you can buy most of your litography equipment for 2 process ahead (like 0.18 and 0.13) if you move to 0.13 and dont need to buy alot of new stuff the move should be easier. but that wasn't the case in Intel and AMD recent history. for example AMD's FAB30 (its biggest one) was built and equipet with 0.18 and 0.13 in mind, yet still AMD had much trouble moving to 0.13.
0.13 was a major step for intel buying all the equpment from scratch, yet it went very fast and well.

SO - I guess its just a pattren.
basicly 0.09u are bound to have more leakge and waste more power becouse of the physicaly small size of the gates... maybe AMD could counter it better, SOI primary advantage is that it reduces leakge by ~20%.

This post is best viewed with common sense enabled<P ID="edit"><FONT SIZE=-1><EM>Edited by iiB on 08/21/03 03:39 AM.</EM></FONT></P>

 

[eden]

True, I forgot about that SOI feature.
Still, implementing it was a drag for both AMD and IBM, so let's hope that the 0.09m transition and process will gladly accept SOI's materials over them.
If it does, perhaps, just perhaps they will do better than Intel.

Still, Intel is running into major leakage troubles and it was foretold by historical experience, that 0.18m had little cause for concern, 0.13m started having some leakage problems, but now 0.09m has major problems. So, one should not rely on patterns here if they do, since 0.09m is indeed one of a kind in how it's transitioned to.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=green><b>A sexual experience like never before seen</font color=green></b></A>
Site has now even more sexy members, for your pleasure.

 

[RRAMJET]

definaly

If he doesn't die, he'll get help!!!

 

[RRAMJET]

definaly

If he doesn't die, he'll get help!!!

 

[cdpage]

i wasfealing lazy and didn't accully read all of the post here but thoiught that this might add something to the discution

Apple has used SOI before i am not all to sure if they are using it on the curren G5... however as IBM is making the G5 they are working on the .09u for that processor as well... so AMD and apple may have an upper hand here. as IBM has 2 differat approches with 2 differant budgets mid you not as big a budget as intel with the 2 together or anything.. but non the less it helps. no?

i found this.

IBM beginning to test PPC 970 chips built on 0.09 micron process: With the 0.13 micron "G5" aka PowerPC 970 processor now shipping in volume at up to 2GHz (up to 2.2GHz is practical with 0.13; 2.4GHz is possible but yields are expected to be very low at that speed - to say nothing of heat and power dissipation issues), IBM's G5 development team has just begun testing the first few wafers of 0.09 micron 970's. This early in the process it is unusual to get very many working chips, but Apple has apparently received a handful (less than 20) of these processors running at between 1.8 and 2.8GHz.

It will be about six months before these processors are available in sufficient volume to base products upon (second-version PowerMac G5s at up to 2.4GHz, and the first PowerBook G5s at 1.6-2.0GHz), but reports from reliable sources in Cupertino suggest that IBM is right on target with its .09m production schedule and that temperature/heat ratings for the new chips are even a little better than expected.




ASUS P4S8X - P4 2.4B - 2 x 512M DDR333 - ATI 9500 Pro(Sapphire) - WD 80G HD (8M Buffer) - SAMSUNG SV0844D 8G HD - LG 16X DVD - Yamaha F1 CDRW - Iomega Zip 250 int.