28nm Chips Remains in Short Supply at TSMC

Status
Not open for further replies.

phatboe

Distinguished
Sep 2, 2006
239
3
18,685
I thought TSMC scrapped the 32nm node to focus on and accelerate the 28nm node. Now they claim to want to be more conservative? Yeah this may save TSMC a few bucks here and there but it is destroying it's partner's profits. Mean while Intel is gaining market share in almost every sector that they are involved in. It's so sad that the TSMC is holding back progress for so many companies all to save a few bucks.
 
If yields are 50% and a wafer cost $1000.00.
You process 1000 wafers a day that equals half a million($500,000) dollar a day loss in wafer cost alone ,not including labor and processing cost!
Can you blame them now?
 
They arent doing anymore yields/good di scenarios, that ended with 40nm.
Its strictly per wafer
This ramp up is actually much faster than 40nm, as the 4770 was one of its first, and came way before we saw decent yields and numbers on 40nm, so 3rd qtr still looks good comparatively
 

maxinexus

Distinguished
Jan 1, 2007
1,101
1
19,360
You are so generous with 50% yields Ricky! Sometime ago I read article about closer to 20% yields per wafer. That is why we don't see Keplers cards anywhere and if they are in small quantities. That is no good for nvidia.
 
G

Guest

Guest
Intel may have to produce everybody's chips. They are the only company that knows how to do this thing.
 
I have not read any articles stating yields in a while just that they had gotten better.
I thought i was being conservative with the 50% and it gave good round numbers for my example.
Looks like it is worse than i thought.
 
G

Guest

Guest
@samIHam: Well, TSMC is leading Intel in process technology. Last time I checked, 28nm is smaller than 32nm.

Of course, you like most other Intel fanboys will just pretend like Intel's 22nm process is already out and doing well, even though here in the real world it's severely delayed, with no hard ETA yet. Even when it is finally available, it may not be available in much volume.

::Queue up people stating that Intel could do better, but they're deliberately choosing not to because AMD isn't 'competive enough'::
 

dreadlokz

Honorable
Mar 30, 2012
312
0
10,790
short supply? sure, just as diamonds and every other expensive thing in the world... ppl who believe that is just dumb! shortage is how they do money
 

rex86

Distinguished
Aug 23, 2011
70
0
18,630
[citation][nom]twenty-eight[/nom]@samIHam: Well, TSMC is leading Intel in process technology. Last time I checked, 28nm is smaller than 32nm. Of course, you like most other Intel fanboys will just pretend like Intel's 22nm process is already out and doing well, even though here in the real world it's severely delayed, with no hard ETA yet. Even when it is finally available, it may not be available in much volume.::Queue up people stating that Intel could do better, but they're deliberately choosing not to because AMD isn't 'competive enough'::[/citation]

Intel is producing 22nm products as we speak, but they won't release it until they sell out their SB based products, which is quite logical as the PC market has been shrinking for some time now. At least that's what I read on the Internet.
 
G

Guest

Guest
28 nm is not necessarily a better process than 32 nm. You are showing your ignorance twenty-eight. There is more to this than just a die shrink (including the type of semiconductors, material type, gate type, etc.). One cannot blindly state that one 28nm process is greater than one 32nm process. One could argue that Intel's 32nm process is as advanced or even moreso than TSMC's 28nm process. Intel's 22nm process is much, much, more advanced for sure.

Please don't try that ETA BS either. You are just being an ignorant troll. Every vendor out there is getting products now in an attempt to ramp up production and have systems ready by May. Did you think they announce one day, then start taking orders the next? You are as daft as they come apparently.

So, again, TSMC is in no way leading Intel in process technology. No one iota. In fact, they are WAY behind the curve that Intel has set in many ways. Next, you'll be telling us that AMD processors are better because they run at 3.9 GHz turbo and Intel's run at 3.8 turbo. Good grief!
 

Parsian

Distinguished
Apr 28, 2007
774
0
18,980
man AMD needs to become independent again or find a worthy partner. This is bad, especially if they are supplying the next gen console chips. This could be the opportunity for AMD to recover.
 
G

Guest

Guest
Fool: I guess you're going to tout that Intel's process nodes are not only ahead of everybody else by 6 to 12 months, but you are going to suggest that they are actually better at the same geometry?

That's completely laughable, Intel's 45nm didn't outperform GloFo's or TSMC's 40/45nm in any metric. Even the 45nm SOI process had less leakage than 45nm HKMG, which was what HKMG was supposed to address.

Then there was Intel's 32nm shenanigans, they spent an eternity selling budget dual-cores and ultra-low-volume 6-cores before they could finally get yields to the point where they could sell high-volume mainstream quad cores. Meanwhile, GloFo ramped their 32nm much more quickly, with moderate availabilty of the entire line available from day 1.

But, where you lose this argument is Ivy's specs: It's barely more than a die shrunk Sandy Bridge, and they only gained little to no clock-speed, at less than 20% power savings? That doesn't sound like a very good 22nm process to me, are you sure it will be any better than competing 28nm processes? Based on evidence from current 28nm GPUs and known specs of Ivy Bridge, 28nm appears to tie or beat 22nm in every metric, even transistor density.
 

ojas

Distinguished
Feb 25, 2011
2,924
0
20,810
[citation][nom]twenty-eight[/nom]@samIHam: Well, TSMC is leading Intel in process technology. Last time I checked, 28nm is smaller than 32nm. Of course, you like most other Intel fanboys will just pretend like Intel's 22nm process is already out and doing well, even though here in the real world it's severely delayed, with no hard ETA yet. Even when it is finally available, it may not be available in much volume.::Queue up people stating that Intel could do better, but they're deliberately choosing not to because AMD isn't 'competive enough'::[/citation]
Lol@that intel's already building 14nm fabs.

anyway, IB launches on the 23rd of this month, that's a solid release date for you. They're taking orders from other companies too, i'm sure they're not facing that many problems.

I'm not sure, however, how one would compare the manufacturing process, but if production volume is any metric, i would say Intel's ahead.

[citation][nom]SM4RT3R_TH4N_U[/nom]Fool: I guess you're going to tout that Intel's process nodes are not only ahead of everybody else by 6 to 12 months, but you are going to suggest that they are actually better at the same geometry?That's completely laughable, Intel's 45nm didn't outperform GloFo's or TSMC's 40/45nm in any metric. Even the 45nm SOI process had less leakage than 45nm HKMG, which was what HKMG was supposed to address.Then there was Intel's 32nm shenanigans, they spent an eternity selling budget dual-cores and ultra-low-volume 6-cores before they could finally get yields to the point where they could sell high-volume mainstream quad cores. Meanwhile, GloFo ramped their 32nm much more quickly, with moderate availabilty of the entire line available from day 1.But, where you lose this argument is Ivy's specs: It's barely more than a die shrunk Sandy Bridge, and they only gained little to no clock-speed, at less than 20% power savings? That doesn't sound like a very good 22nm process to me, are you sure it will be any better than competing 28nm processes? Based on evidence from current 28nm GPUs and known specs of Ivy Bridge, 28nm appears to tie or beat 22nm in every metric, even transistor density.[/citation]
You're partly confusing architecture with the manufacturing process. AMD, Nvidia, Intel have different applications and different ways to use that silicon. I'm not sure what you're trying to say here. How is IB a good example of a manufacturing process comparison? if by simply shrinking the process while retaining the architecture, you're saving 15-20% power, how isn't that bad? I mean Haswell=change of arch+22nm, why not use that as a metric against sandy bridge after that releases? I mean they're both on a "tock" cycle, after all.

I don't know what you're whining about. I mean FFS let's just hope TSMC gets their act together so we can buy our graphics cards...if i have to wait past june i'll simply buy a GTX 560, something i don't want to do.

p.s. Nvidia wanted Intel to manufacture their chips for them...i think that says everything, doesn't it? If it weren't embarrassing for AMD to do the same, i'm sure it would too. And get the same answer, "no". :p
 

Microgoliath

Distinguished
Dec 13, 2011
113
0
18,680
@smarter_than_u

10-20% lower power yet still attaining same / higher clocks than sandy which uses more power to get there.

So give ivy the same power that sandy uses and ramp those clocks up.
 

vaughn2k

Distinguished
Aug 6, 2008
769
4
19,065
[citation][nom]samIHam[/nom]Intel may have to produce everybody's chips. They are the only company that knows how to do this thing.[/citation]
Their scientist are being paid very well, they also tend to camp outside of Intel just to get 'things' done. Missed those days...
 

ukee1593

Distinguished
Jun 8, 2009
290
0
18,810
Without a working knowledge of both Intel AND TSMC's manufacturing processes there is no way that anyone on the tom's hardware forums can tell whether Intel's manufacturing process is more advanced than TSMC's or not.

While I personally belive that Intel's 22nm is the better process I cannot definitely state this!! From the articles I have read Intel's 22nm is less efficient with space than TSMC. Then again Intel uses FinFET transistors now.

Please don't try to compare the manufacturing processes
 

jkflipflop98

Distinguished
[citation][nom]ukee1593[/nom]Without a working knowledge of both Intel AND TSMC's manufacturing processes there is no way that anyone on the tom's hardware forums can tell whether Intel's manufacturing process is more advanced than TSMC's or not. While I personally belive that Intel's 22nm is the better process I cannot definitely state this!! From the articles I have read Intel's 22nm is less efficient with space than TSMC. Then again Intel uses FinFET transistors now. Please don't try to compare the manufacturing processes[/citation]

Actually, anyone that knows anything about the semiconductor industry knows that Intel is far and away the leader in process technology. Intel's 22nm Finfets are higher performing, lower power, and more robust than TSMC's 28 nm technology. Heck, no one else has even gotten close to the performance characteristics of Intel's 32nm HKMG transistors, and Finfets are almost 50% faster.
 
G

Guest

Guest
http://www.chipworks.com/en/technical-competitive-analysis/resources/technology-blog/2011/07/more-hkmg-hits-the-market-%E2%80%93-gate-first-and-gate-last/ - Ahem...

http://www.tomshardware.com/reviews/medfield-krait-smartphone-mobile-soc,3117-5.html - more

TSMC is just starting to do the things Intel has been doing for years.

As for GloFoundry, ahem:
http://www.anandtech.com/show/4894/amd-confirms-32nm-yield-issues-at-global-foundries - apparently had a lot of yield issues at 32nm (LMAO).

Intel has the most advanced gate-last high-K process and now adds FinFET to the design and a shrink to 22nm. I have no idea why anyone is claiming that TSMC is ahead of Intel or that GloFo has better yields or any of the other nonsense that has been said on this thread.
 
Status
Not open for further replies.