Sunday, January 22nd 2012

28 nm struggles: TSMC & GlobalFoundries

Making silicon chips is not easy, requiring hugely expensive fabs, with massive clean-room environments and at every process shrink, the complexity and difficulty of making the things goes up significantly. It looks like TSMC and GlobalFoundries are both having serious yield problems with their 28 nm process nodes, according to Mike Bryant, technology analyst at Future Horizons and this is causing a rash of non-working wafers – to the point of having nothing working with some chip designs submitted for production. It seems that the root cause of these problems are to do with the pressures of bringing products to market, rather than an inherent problem with the technology; it just takes time that they haven't got to iron out the kinks and they're getting stuck: "Foundries have come under pressure to release cell libraries too early – which end up with designs that don't work," Bryant said. In an effort to try and be seen to treat every customer equally, TSMC is attempting to launch ten 28 nm designs from seven companies, but it's not working out too well: "At 45-nm, only NVIDIA was affected. At 28-nm any problems for TSMC will be problems for many customers" said Bryant.

GlobalFoundries are also struggling, although perhaps not quite as badly, according to Bryant: "However, there are recent comments of major yield problems with their 28-nm process actually being even worse than at GF [Globalfoundries]". Their 32 nm & 28 nm nodes are struggling, because they are using problematic gate first processing and this is worsened by the fact that they are using processes from two companies, AMD and IBM. Trying to debug two processes at once is causing serious headaches, compounded by a lack of cooperation between bases in Dresden and the US which appears to be caused by bad management.

Note that Bryant's assertions are at odds with what TSMC's CEO and chairman Morris Chang said when he spoke to analysts the previous day about the company's fourth quarter financial result: "Our 28-nm entered volume production last year and contributed 2 percent of 4Q11's wafer revenue. Defect density and new progress is ahead of schedule and is better than 40-45-nm at the corresponding stage of the ramp-up. We expect 28-nm ramp this year to be fast and we expect 28-nm will contribute more than 10 percent of total wafer revenue this year." It will soon become apparent who is right.

Regardkess, the pressure isn't going to let up for these companies, since Intel have been successfully manufacturing at the 32 nm level for a year, making their Sandy Bridge processors. On top of that, in April, Intel will introduce their significantly smaller 22 nm-based process technology in the form of their Ivy Bridge CPUs which have been demonstrated to work very well indeed. These are based on Intel's proprietary Tri-Gate 3D transistor technology too, which gives further performance increases, hence upping competitive pressures significantly.Sources: TG Daily, EE Times
Add your own comment

33 Comments on 28 nm struggles: TSMC & GlobalFoundries

#1
the54thvoid
The real reason for 7970 scarcity, 7950 delay and Kepler's Q2 release???
Posted on Reply
#2
Delta6326
dang I wanted to get a 7950 hope this doesn't mean more delays or a higher price.
Posted on Reply
#3
seronx
"However, there are recent comments of major yield problems with their 28-nm process actually being even worse than at GF [Globalfoundries]," Bryant told the audience.
But no wait if you read source #2 it says the opposite even though GlobalFoundries isn't in volume production the yields are still better

The only issue for GF is that volume production starts after Q2 2012(This has been known for sometime and some how the article makes it more grim than it sounds)
Posted on Reply
#4
Legacy-ZA
Just another reason to raise prices. Pathetic.
Posted on Reply
#5
SteelSix
by: Legacy-ZA
Just another reason to raise prices. Pathetic.
Or to wait for things to shake out. Me, I'm done chasing product launches. It's not been easy, but I'm holding out..
Posted on Reply
#6
Aleksander
Nvidia will again be affected
Anything to do with SWPA?
Posted on Reply
#7
Hokum
Isn't global foundries a spin off of AMD?
Posted on Reply
#8
radrok
I was somehow expecting something like this, isn't TMSC ALSO manufacturing SoC for smartphones/tablets? I am sure about Apple SoCs.
The advantage Intel has with the manufacturing process will probably give them an edge when entering on SoC market with Medfield while all the rest struggles with TMSC 28nm problems.

by: SteelSix
Or to wait for things to shake out. Me, I'm done chasing product launches. It's not been easy, but I'm holding out..
There isn't the need to jump on 28nm graphics right now, assuming you are talking about graphics of course :), consoles are holding the development and until they get a refresh the "older" 5xx/6xxx series are more than enough for playing.
Posted on Reply
#9
Dj-ElectriC
And they told us there absolutely no problems with 28nm manufacturing processes, way to go...
Posted on Reply
#10
erocker
So.. according to TMSC:
TSMC CEO Morris Chang insists the company's 28nm manufacturing process technologies remain on track.
As far as Glo-Fo goes, there is nothing but "rumor" according to TechEye that they are having problems.

Looks like it is definitely a bunch of FUD.
Posted on Reply
#11
GC_PaNzerFIN
I recently downgraded to GTX 460 from GTX 580 because I plan to upgrade to some even faster 28nm card. Although I wouldn't really NEED one right now (I can play everything I have now just fine, at decreased quality settings tho) I can't wait to get my hands on one. HD 7970 looks pretty tempting, especially the DC2 card, but I would really prefer NVIDIA. GK104 is definately coming first but that really isn't what I am looking for either, I am waiting for the GK100. Too bad it looks like I have to wait till H2 this year... :(
Posted on Reply
#12
seronx
by: Hokum
Isn't global foundries a spin off of AMD?
Of the foundries AMD owned + More

Fab 1 - Dresden, Germany
Fab 8 - Luther Forest Technology Campus, Saratoga County, New York, USA

Are the only two that AMD use to have ownership on

http://www.globalfoundries.com/manufacturing/300mm.aspx
http://www.globalfoundries.com/manufacturing/200mm.aspx

Everything else was Chartered Semiconductor

http://www.globalfoundries.com/manufacturing/fab8_overview.aspx

Take note of this:
Several centers of excellence for semiconductor and nanotechnology R&D are in close proximity to Fab 8. GLOBALFOUNDRIES is a member of the Joint Development Alliance (JDA) centered at East Fishkill, New York with several world leading product companies including IBM, Renesas, STMicroelectronics, Samsung Electronics, and Toshiba. With this alliance, GLOBALFOUNDRIES is driving the global standard for new technologies such as High-K Metal Gate. As a founding member of the JDA starting from 90nm, GLOBALFOUNDRIES has been able to leverage the collaborative R&D model to reduce the cost of process R&D that is projected to reach more than $1B US for 20nm and below. GLOBALFOUNDRIES also collaborates closely with the The College of Nanoscale Science and Engineering (CNSE) of the University at Albany - State University of New York (SUNY), a global education, research, development and technology deployment resource dedicated to preparing the next generation of scientists and researchers in nanotechnology.
If I had anything to speculate I would be saying GlobalFoundries is perfecting the process before going into volume rather than releasing the process in an unoptimized way

My guesses:
28nm-SHP => 28nm FD-SOI UTBOX HKMG eSiGe (Kaveri APU/Steamroller CPU)
28nm-HPP => 28nm Bulk HKMG eSiGe
28nm-HP => 28nm Bulk HKMG eSiGe
28nm-SLP => 28nm Bulk HKMG eSiGe
28nm-LPH => 28nm FD-SOI UTBOX HKMG eSiGe (Wichita/Krishna APUs(The current Bobcat is getting an extension on 40nm while these APUs are releasing next year))

Bold = Q2 2012 and later
Red Bold= Q4 2012 and later
Posted on Reply
#13
badtaylorx
so much for the early summer price war :cry:
Posted on Reply
#14
qubit
Overclocked quantum bit
by: seronx
But no wait if you read source #2 it says the opposite even though GlobalFoundries isn't in volume production the yields are still better

The only issue for GF is that volume production starts after Q2 2012(This has been known for sometime and some how the article makes it more grim than it sounds)
Yes, quite, thankyou. I'd meant to fix that bit before I published the article, so I'm not sure what happened. Fixed now. :toast:

by: erocker
So.. according to TMSC:



As far as Glo-Fo goes, there is nothing but "rumor" according to TechEye that they are having problems.

Looks like it is definitely a bunch of FUD.
Yup, I'd meant to add that bit in too (TSMC's CEO statement). Fixed. :)
Posted on Reply
#15
a_ump
i really am never surprised about this when new process nodes are released. Every single one for the last 3 nodes or so seem to run into hiccups, i just accept it as that's how it'll always work. nothing new imo
Posted on Reply
#16
OneCool
Seems that the 28nm process is a mother... how long has this been giving them issues now 6-8 months?

You know Intel is sitting back laughing :rolleyes:
Posted on Reply
#17
seronx
by: OneCool
Seems that the 28nm process is a mother... how long has this been giving them issues now 6-8 months?
Most of the issues for GloFo is Gate First while TSMC doesn't have the same excuse
Once 28nm becomes full volume production it will mostly roll out
GF 28nm > TSMC 28nm

by: OneCool

You know Intel is sitting back laughing :rolleyes:
I would actually wait and see Intel has been getting some backlash about lying about tidbits of their 22nm Tri-gate process
Posted on Reply
#18
ice_v
by: OneCool
Seems that the 28nm process is a mother... how long has this been giving them issues now 6-8 months?

You know Intel is sitting back laughing :rolleyes:
lol true

of all the chip players, Intel is one of the few that can actually relax for at least the entire Q1.
Posted on Reply
#19
Protagonist
Damn, but I'm not in a harry to get new graphic card, too bud for TSMC they've been having this issue for a while now.

While Intel seem to be yanking out new process with less strain than TSMC or any other fab.

And there was an article here stating that Intel has 14nm chips already being tested in their labs as of last year, and its not due till 2014 i think, that's how Intel does it, If they are to release new nanometer they start preparing years in advance.
Posted on Reply
#20
dude12564
by: st.bone

And there was an article here stating that Intel has 14nm chips already being tested in their labs as of last year, and its not due till 2014 i think, that's how Intel does it, If they are to release new nanometer they start preparing years in advance.
That should be correct, i think it's Broadwell (Shrink of Haswell) and Airmont (Successor of Silvermont, which is then the successor of Medfield) coming out, if Intel keeps their Tick-Tock going.
Posted on Reply
#21
Protagonist
by: dude12564
That should be correct, i think it's Broadwell (Shrink of Haswell) and Airmont (Successor of Silvermont, which is then the successor of Medfield) coming out, if Intel keeps their Tick-Tock going.
You seem to know the details. I've had this info before, do you by any chance know where i can find Intel's read map, i used to find them on their website but nowadays i cant seem to see them, so if you can find the road map please post it or post a link preferably PDF
Posted on Reply
#22
xenocide
by: st.bone
And there was an article here stating that Intel has 14nm chips already being tested in their labs as of last year, and its not due till 2014 i think, that's how Intel does it, If they are to release new nanometer they start preparing years in advance.
It said that they were testing early forms of it. It's not like they have tons of those sitting around. Intel goes out of their way to find the best engineers they can. GF has been doing a lot of talent hunting in the past year or so for a plant they are openning in NY as well, but Intel is definitely top dog.
Posted on Reply
#23
seronx
Capacity: 1st. Intel 2nd. GlobalFoundries & Samsung
Development: 1st. Intel 2nd. Anyone part of the SOI Consortium(IBM, GlobalFoundries, STMicroelectrics, etc.
Production of the Highest End: 1. GlobalFoundries 2. Intel

GlobalFoundries 32nm-SHP should be the most expensive process yet it is cheaper than Intel's 32nm high-end process

There is no way 315.8mm² of 32nm-SHP actually costs $245- and 32nm from Intel at 200-295mm² costs $285+
Posted on Reply
#24
largon
by: seronx
Capacity: 1st. Intel 2nd. GlobalFoundries & Samsung
And I thought AMD actually had to buy fabbing from a subsidiary. (Sorry for the sarcasm.)
Development: 1st. Intel 2nd. Anyone part of the SOI Consortium(IBM, GlobalFoundries, STMicroelectrics, etc.
I guess by "development" you only meant the lenght of the development cycle, not technological advancement, shrinkage of fab processes, etc.
Production of the Highest End: 1. GlobalFoundries 2. Intel
Yet, AMD's chips are slower and consume considerably more power.
Posted on Reply
#25
xenocide
by: seronx
GlobalFoundries 32nm-SHP should be the most expensive process yet it is cheaper than Intel's 32nm high-end process
Intel's is more expensive because it is a better product with more reliable yields. They first shipped product developed using High-K in 2007, GF just barely did it in 2011. Then there's the whole gate-first vs. gate-last situation. Overall Intel is just doing better than GF, as they should, since they cherry-pick the best engineers in the worldindustry.
Posted on Reply
Add your own comment