EETimes - TSMC's R&D boss addresses 40-nm yields, high-k, litho
Interesting interview especially with 28 nm transition. Time will tell us as usual ...
Thanks to DVHardware
Printable View
EETimes - TSMC's R&D boss addresses 40-nm yields, high-k, litho
Interesting interview especially with 28 nm transition. Time will tell us as usual ...
Thanks to DVHardware
very interesting, cheers! :toast:
only 80.000 12" 40nm wafers per month right now?
thats not a lot, is it?
That's quite a lot, actually, considering the 40nm issues we all know about from ATI and nVidia's products...Quote:
Originally Posted by saaya
It's not 80,000 a month it's 80,000 a quarter, they are hoping to get to 160,000 a quarter soon.Quote:
Originally Posted by saaya
Isnt that per quarter ?Quote:
Originally Posted by saaya
80,000 per month would have been a lot, but for a quarter... not so much
Quote:
Originally Posted by saaya
80k a month would be much. 80k a quarter is ok. GFs production capacity in Dresden is also only 50-60k Wafer per Month i think. And at the Moment only one Fab is producing there the AMD CPUs.
Quote:
Originally Posted by Iconyu
i don't think "by the end of this year" is really that soon... how will they be able to keep up with demand for both amd and nvidia gpu's? how has amd been able to ship "over 1 million" dx11 cards if production is so low?Quote:
we are able to do about 80,000 wafers per quarter at the moment. These are twelve-inch wafers. And this will be doubled by the end of this year
EDIT: i'm guessing the 1.9 million euro "clamp" is actually a crane or hoist?
What is the problem with shipping over 1million chips?
1million chips is even with middle yields with cypress just 10000 Wafer. That's 2500 Wafer per month in 2009 for the production time of Cypress.
probably design issues with fermi i take it?Quote:
Originally Posted by 570091D
They've been complaining about not getting enough 40nm GT2xx chips as well so the problem isn't just Fermi, it seems to be overall 40nm capacity. I wonder how those are yielding at this point.Quote:
Originally Posted by 570091D
oh my bad, 80.000 per quarter... then thats 20.000 per month?
nvidia can get a max of 100 fermis per wafer, right?
so with 100% yield they could get 200.000 fermis per month
then why do they complain about tsmcs capacity? :confused:
even with 10% yield they could still get 20.000 fermis per month...
and lets say they can only use 30% of tsmcs capcity (which is rather pessimistic considering nvidia is historically tsmcs biggest customer AND only get 10% yields... then thats still 7000 fermis per month...
so a 1 month preparation should be enough to launch fermi with good availability... at least from tsmcs standpoint... right?
7000 chips good availability? dont think so.
A quarter is 3 months.Quote:
Originally Posted by saaya
That capacity is for all customers and all chips, not just GF100. And yes, 7000 GF100's a month would be a complete joke.Quote:
then why do they complain about tsmcs capacity? :confused:
Those 20.000 wafers are for all their 40 nm customers, we don't know what's nvidia's slice of that.Quote:
Originally Posted by saaya
the gtx280 launched with less than that, much less...Quote:
Originally Posted by Mechanical Man
doh :lol:Quote:
Originally Posted by trinibwoy
yeah, so thats 25.000 wafers per month then...
how so? ati sold 1 million rv870 and rv840 chips in about 3 months, thats 333.000 per month... im pretty sure 5870s were only around 30.000-100.000 per month...Quote:
Originally Posted by trinibwoy
25.000 wafers per month
worst case, nvidia gets only 8000 of those
worst case, fermi yields are 10% so they can get only 80.000 fermis per month if they use all their capacity for fermi
best case, nvidia gets 15.000 of the wafers
best case, fermi yields are 30% (which is very good for a huge chip)
so nvidia can produce 450.000 fermis per month if they use all their capacity for fermi.
i find it hard to believe that nvidia is limited by tsmcs capacity...
Do you think they were all manufactured in those three months?Quote:
Originally Posted by saaya
They won't. They have other much higher volume products on 40nm to worry about.Quote:
if they use all their capacity for fermi
If TSMC is not able to fulfill orders for both ATi and nVidia, they will have to produce chips for the company they have a more solid contract with. My bet would be on nVidia since they have far more market share and are a much larger source of income for TSMC.Quote:
Originally Posted by 570091D
TSMC has been producting .40nm chips for ATi months before consumers were able to purchase them... and much of those 1 million chips were manufactured before December while yields were very low. This is why I've always maintained the opinion that ATi jumped the gun just to get the crown for a while but forgot to remember that without profits, they wont be able to pay off all the R&D they're spending on very fast product cycles.Quote:
how has amd been able to ship "over 1 million" dx11 cards if production is so low?
nVidia has almost 3x the market share of ATi. A 'low volume' card for nVidia is probably quite similar in units produced as a 'high volume' card from ATi.Quote:
Originally Posted by trinibwoy
Plus I have a feeling TSMC makes more profit from larger accounts like nVidia.
What makes you think Nvidia is trying to produce as many GF100's as it can? I think the limited availability of GF100 won't be because of TSMC's capacity issues. It's because of the yield issues, and even then it's not because there are not enough chips. It's because yields are bad, Nvidia cannot make a profit, so they will try to minimize their loss by producing a very little amount of chips.Quote:
Originally Posted by saaya
My guess is that they are going to sell those chips at a good price and then make the customers wait for the yields to go up and the chip to be actually profitable for them to mass produce, at which stage they will.
Just a guess, sure, but it makes sense
most of them... yes... they reported that they broke the 2million mark in january, that was only a few weeks after they broke the 1 million mark...Quote:
Originally Posted by trinibwoy
thats probably true...Quote:
Originally Posted by trinibwoy
i still find it hard to believe that they are capacity limited...
3x the market share? where did you get that number from? :confused:Quote:
Originally Posted by Andrew LB
yes, thats what i think as well... i was just saying, hypothetically, lets say nvidia is really capacity limited and they try to produce as much dx11 but they cant, because tsmc just cant get enough wafers processed...Quote:
Originally Posted by annihilat0r
to me, that sounds like they blaming tsmc... once again...
and they probably told their investors and analysts that of course fermi is finished working perfectly fine, and yields are fine, they COULD have launched it at any time but simply decided its better to wait until tsmc has better yields and solved their "capacity issues"
and most analysts and investors will probably even believe that... and tsmc being a typical asian suckup happily bends over and apologizes and claims its all their fault... while ati is shipping hundreds of thousands of 40nm parts every month for over a quarter...
This is VERY useful interview, that can clarify some things!
So yield problems with Fermi are not related to TSMC's process, but to the design of the Fermi?!Quote:
You all heard about TSMC is challenge during the early part of last year. I report to you we are glad all this problems was behind us. We resolve this yield problem in the second half of last year. So we're glad the yield issue was over, and we are building the capacity very aggressively to fulfill the very high demand from our customers.
So there'll be no Fermi shrink before H1 2011 at best!?Quote:
The first high-k metal gate we call 28 HP for the high performance application will be introduce the end of September this year, and followed by three months later December will be the 28 HPL. This is the first high-k metal gate introduction for the low power application.
so the pace we'll see in GPU industry will be:Quote:
Going forward, we plan to introduce 22 nanometer node about two years after we introduce 28 nanometer, so the first introduction we'd like to be in the Q3 of 2012, and this if for the high performance version. And followed by the low power version about the end of Q1 2013.
Nex gen - @ 28nm in H2 2011
NexNex gen @ 22nm in Q4 2013
in 2015 there'll be ONLY 4 semi companies in the plane Earth:Quote:
And lithography side the industry seems like the mainstream trend will be to go to EUV and you may or may not have heard about the cost of EUV. If you buy one EUV tool with a matched track, it will cost like 80 million dollars for just one tool. I was shocked to sign a P.O. only couple weeks ago shortly before my vacation. I signed a 1.9 million euro purchase order for a clamp.
This clamp is a custom made clamp only for EUV. We have to mount a special clamp on the ceiling and this clamp will be used to lift the EUV tool when we install the machine, and when we do the maintenance. This tool is so heavy, no other tool can lift it up, and this custom-made clamp costs us 1.9 million euros, just to buy a clamp. It's really shocking.
Intel
GloFo
TSMC
Samsung
"Traditionally" TSMC befriends with ATi more because ATi is more willing to test on new process. But now AMD is tied to GF for future manufacturing.Quote:
Originally Posted by Andrew LB
that actually isn't true.Quote:
Originally Posted by qcmadness
in many occasions AMD executives confirmed that they're staying in business with TSMC, just as they've done it so far.
But the agreement between AMD and ATIC states that after 32nm / 28nm transition, AMD will go for GF and eventually hires GF to manufacture ALL AMD CPUs and GPUs.Quote:
Originally Posted by Nedjo
http://ccbn.10kwizard.com/xml/downlo...915755&sXBRL=1
Page 7
Quote:
Wafer Supply Agreement. The Wafer Supply Agreement governs the terms by which we purchase products manufactured by GF. Pursuant to the Wafer Supply Agreement, we purchase, subject to limited exceptions, all of our microprocessor unit (MPU) product requirements from GF. If we acquire a third-party business that manufactures MPU products, we will have up to two years to transition the manufacture of such MPU products to GF. In addition, once GF establishes a 32nm-qualified process, we will purchase from GF, where competitive, specified percentages of our graphics processor unit (GPU) requirements at all process nodes, which percentages will increase linearly over a five-year period. At our request, GF will also provide sort services to us on a product-by-product basis.
well qcmadness it looks like you've slapped me twice in a row! :D hats down to you ;)
wow, so ati HAS to use GF? that might actually cause them trouble if theres a problem with GF?
Actually, when you read it carefully. You see AMD has to ramp their GPU production in 5 years. Eg. Now they are starting with Lliano (integrated CPU+GPU) and during rest five years they do need to ramp up the production when viable. So if GF cant supply good process for AMD GPUs, AMD does not need to move their GPUs to GF. Also, there is no mention on that exerpt about what end percentage should even be. There is space for discrete graphics staying on TSMC. Remember, most high volume products are integrated GPUs..
Digitimes - GPU shortage unlikely to be solved before May
:shakes:Quote:
The current shortage of graphics chips, which is being caused largely by low yields of the 40nm process at Taiwan Semiconductor Manufacturing Company (TSMC), is unlikely to be completely solved before May, according to sources from graphics card makers.
Although TSMC recently said the defect density of its 40nm technology has already dropped from 0.3-0.4 per square inch to 0.1-0.3, the sources pointed out that the improvement in overall yield still needs more time before catching up with market demand.
To maintain gross margins from the increased cost created by the GPU shortages and rising DRAM quotes, graphics card makers including Asustek Computer, Micro-Star International (MSI), Gigabyte Technology and Tul are planning to increase prices of their products by 5-10%, according to market sources.
Quote:
Originally Posted by Andrew LB
this makes sense....
so ati producing chips and getting them to market and having a good hold on the market for now is bad to recoup the R&D cost ???? sure......
http://www.epicallyfunnypictures.com...orly4-full.jpgQuote:
Originally Posted by Andrew LB
Now let's face the facts:
a)
http://www.dvhardware.net/news/amd_c...dx11_wafer.jpg
those two guys looks to be happy working together!
b)
AMD has no issues with supply of 40nm DX11 chips!
c)
There's one another company that can't get enough 40nm DX11 chips
d)
TSMC boss firmly stated in couple of occasions that they've fixed problems with 40nm, and yields are fine, with continuous ramping of the output wafers!
b+c+d = stories about issues with TSMC's 40nm tech are coming from the side that has broken design for their massive 40nm chip!
So what's the measure of TSMC's success??
a) they can successfully build 2 billion transistors chip on 40nm
b) they can't successfully build 3 billion chip on 40nm
SO POINT:
what do you think that TSMC will chose?? To have their equipment producing what they CAN produce, or to have them idle 'cos they can't produce 3 Billion chip with sufficient yield in sufficient quantities??
yeah those PhD and MBA 'tards in ATI don't get the simple fact that they can get paycheck only if they make profit! Please be humane and send them a notice, 'cos you're obviously aware of that simple fact!!Quote:
TSMC has been producting .40nm chips for ATi months before consumers were able to purchase them... and much of those 1 million chips were manufactured before December while yields were very low. This is why I've always maintained the opinion that ATi jumped the gun just to get the crown for a while but forgot to remember that without profits, they wont be able to pay off all the R&D they're spending on very fast product cycles.
:rolleyes:
And you have some source to back up this brilliant statement?? :rolleyes:Quote:
Originally Posted by Andrew LB
well if you have feeling who're wee to confront you :rolleyes:Quote:
Plus I have a feeling TSMC makes more profit from larger accounts like nVidia.
k... so they basically HAVE to keep a foot in the door at GF and give them a try, so to speak... hoping that as soon as GF does a good job or TSMC stumbles they are ready to go full throttle @ GF... makes sense... but wouldnt they do this either way? its weird that they mention this in a contract, isnt it?Quote:
Originally Posted by Mechanical Man
its kinda like signing a contract that tells you that you HAVE to look for other job opportunities that might be more fun or pay you a better salary... well duh! who doesnt do that? :D
thats interesting!Quote:
Originally Posted by Olivon
VERY interesting!
0.1-0.3 errors per square inch...
1 square inch = 25.4x25.4mm
fermi =~ 24x24mm
so on average there is 1 error per 3-10 fermi chips?
that doesnt make any sense... at all...
if this would be true then fermi should have yields of 70-90% fully functional 480 chips and 5-25% salvageable broken chips aka 470s, and rv870 should have yields of 80-95% fully functional and 5-15% salvageable parts aka 5850 or 5830 :eh:
amd/ati's agreements with tsmc&glofo is meant to address the "problem" of spintel escalating a node-process war. if you ask me, it is a brilliant short-term solution to keep up with tick-tock, freeing up amd/ati from manufacturing headaches and allowing them to fully concentrate on gpu/cpu/cgpu design. meanwhile glofo will be tip of the spear fighting the manuf war vs. spintel, with deep financial pockets courtesy of abudhabi.
im waiting to see how chipzilla counters this strategy. or will it be more of the same old trick bribing OEM's.
defects are not only thing that have influence on yield...Quote:
Originally Posted by saaya
false.Quote:
Originally Posted by Nedjo
you cant have it both ways. ATi pays tsmc less money per chip and sells their products at a lower price point. it should be obvious that tsmc is making more money off of nvidia. hell, they probably payed off the majority of the cost of 40nm node.:ROTF:Quote:
TSMC boss firmly stated in couple of occasions that they've fixed problems with 40nm, and yields are fine, with continuous ramping of the output wafers!
b+c+d = stories about issues with TSMC's 40nm tech are coming from the side that has broken design for their massive 40nm chip!
So what's the measure of TSMC's success??
a) they can successfully build 2 billion transistors chip on 40nm
b) they can't successfully build 3 billion chip on 40nm
SO POINT:
what do you think that TSMC will chose?? To have their equipment producing what they CAN produce, or to have them idle 'cos they can't produce 3 Billion chip with sufficient yield in sufficient quantities??
so the real problem isnt defect density but what? transistor variance?Quote:
Originally Posted by Nedjo
neither ati nor nvidia pay tsmc per chip, they pay them per wafer afaik, thats what several articles mentioned at least... and even IF ati would pay them per chip, their chips are smaller and use up less wafers and capacity... so with the same amount of wafers tsmc would make roughly the same amount of money with either one...Quote:
Originally Posted by Chumbucket843
its ironic that you said "you cant have it both ways" because you just claimed that tsmc prefers nvidia because they pay them more money per wafer/chip and thats why nvidia gets lower wafer/chip prices... lol wutt? :confused2 :lol:
well it looks so...Quote:
Originally Posted by saaya
they do pay per wafer but since nvidia has bigger die and lower yields it cost more per chip which is what matters to consumers and nvidia's profit margin. i do remember seeing some research that some company did. gt200 was costed $50 per chip and rv770 was $30 per chip. nvidia probably does get a deal but its not like it makes up for the cost.
when you do such claims back them up and please step back from putting out one wrong statement after another :down:Quote:
Originally Posted by Chumbucket843
you can get every single card of the 5kseries right now both in europe and in the US:
all HD5K series are in stock in the EU
Newegg - where is the shortage???
NCIX - no shortage at all
Quote:
Originally Posted by Mechanical Man
Yep, it probably won't even be 7000 chips a month (if we can keep your 10% yield figure which is too low). This is not a lot by all means, because shipping millions of these chips (like ATI does) would take years.Quote:
Originally Posted by trinibwoy
Lovely, probably a year later than Intel...Quote:
Originally Posted by Chiang
i dont know... there is no source mentioning this or even hinting at it... everybody talks about defects... gate variation affects transistor performance and leakage, but it doesnt break chips...Quote:
Originally Posted by Nedjo
i find that hard to believe... if gt200 is only 50$ how come the cheapest card based on gt200 comes in at above 200$?Quote:
Originally Posted by Chumbucket843
even now that 55nm is mature and the 260 is under price pressure it doesnt sell for less than 225$, which makes me guess that the chip cost is above 50$... probably close to 75$...
well transistors capable of 50% of targeted speed might as well be defected from usability, or yield standpoint.Quote:
Originally Posted by saaya
the other parts were expensive to. nvidia caught a break by using gddr3 which was fairly cheap. there are a lot of other expenses besides manufacturing. testing and packaging are very costly. silicon is actually cheap. if tsmc can pay off the initial cost of the fab then their margins are ridiculous. high end is not a lucrative market (arguably hardware altogether) but its necessary for reputation.Quote:
Originally Posted by saaya
hemlock is very hard to find.Quote:
Originally Posted by generics_user
http://www.newegg.com/Product/Produc...n=5970&x=0&y=0
not to mention its $700.
http://www.google.com/products?hl=en...e2=&lnk=prsugg
yeah, me and my wrong statements. i wouldnt mind have 2 of these in a crunching rig.
how come almost all rv870s can clock past 1ghz then?Quote:
Originally Posted by Nedjo
even the lowest rv870 part aka 5830 comes clocked higher than 800mhz...
if this would be such a big problem dont you think they would disable more units for a 470 than just 2? a 460 or 450 with more disabled blocks or lower clocks would still make a looot more sense than throwing away those gpus that have SOME blocks that cant clock that high...
yeah but here its gddr5 vs gddr5 and i dont think testing gf100 will cost more than testing rv870, right? packaging maybe... but i dont think thats enough to make a difference in retail prices... maybe 5$ more for a more complex package?Quote:
Originally Posted by Chumbucket843
http://www.zipzoomfly.com/jsp/Produc...odlist=froogle
wow, 599 for a 5970, thats a good offer for anybody interested in that much performance...
thats 100$ cheaper than newegg...
Anand's article about Evergreen development is quite reviling on this subject. Basically ATI has calculated all of those risks, and made TSMC 40nm-friendly design (thanks to RV740 for that).Quote:
Originally Posted by saaya
NVIDIA didn't do that breaking to many rules and pushing TSMC's manufacturing capabilities to the limit.
it's hard to be precise on that matter w/o insight in all the factors that constitute final expense.Quote:
yeah but here its gddr5 vs gddr5 and i don't think testing gf100 will cost more than testing rv870, right? packaging maybe... but i don't think thats enough to make a difference in retail prices... maybe 5$ more for a more complex package?
good price indeed.Quote:
http://www.zipzoomfly.com/jsp/Produc...odlist=froogle
wow, 599 for a 5970, thats a good offer for anybody interested in that much performance...
thats 100$ cheaper than newegg...
:up:
only 75$ more than a gtx295 and only 200$ more than a 5870 (399$)... definitely very nice for whoever needs that much power :)
i read the anandtech article... but what anandtech didnt mention is that nvidia DID tape out other 40nm chips before gt300...
most of them were cancelled because they didnt work at all i think, but all the 10.1 parts were taped out before gf100... so nvidia did have SOME idea of 40nm before taping out gf100...
and especially considering the obvious clockspeed issues nvidia ran into with their 10.1 40nm parts, id be very surprised if they didnt take transistor variations into account when designing fermi...
It makes sense I think. If you look at HD 5830, it has TDP almost as high as HD 5870. This implicates the chips needs higher voltage even they are smaller (the parts which are cutted of are not supplied with power). So probably these are the chips with not so good transistors? :shrug:Quote:
Originally Posted by saaya
As for the price of GT200 based cards - look at manufacturing track of every goods you can think about, not only electronics. You'll discover the manufacturer of first raw material has only few % of the final price. Rest is how the thing travels and everybody who it pases takes his 10% part...
E.g. tea or coffee is good example: the farmer gets few bucks, but it could cost up to hundreds dollars on the stock market on the other hamisphere. So basically if the GT200 costs 50 USD, the 200 bucks final price is actually very low; that's the reason many NVIDIA AIB's were so suffering last year: not only low prices, but low sales too.
ADD// The VGA shortage will be really bad, at least it looks like it really will be. If you look at the market, it looks that GT200, G92 and RV7x0 just are not manufactured anymore (GT200 for sure, but others too I think). Everything is 40nm now and even if yealds were 100%, 80 000 wafers/quarter is not enough for whole RV8x0, GT(S)2x0 and Fermi at this demand.
I haven't read the thread, I'm kinda tired, bored and still gotta do lots of things :(
So please excuse me if somebody else came up with the same thought here before I ( just ) did :D
TSMC's boss saying "We're doing great and we fixed this blah blah" without any real proof, etc isn't as valuable as Jen-Hsun H. now saying "We have a great product and everything is going fine" ? :D
ofcourse it's not a same!Quote:
Originally Posted by BenchZowner
on TSMC's defense we have bunch of 40nm Evergreen based cards from the second wave (MSI, GA, ASUS, Sapphire, PowerColor...) - check out TechPowerup... so TSMC is printing out Evergreens like there's no tomorrow... also it looks like there's plethora of NV DX10.1 chips out there... but there's no Fermi yet! ;)
I saw your nickname from the search page, and I knew what you'd post :DQuote:
Originally Posted by Nedjo
Guess what... MSI lightning = still remain to be seen, and if the info regarding the amount of cards available for the launch is true, it's going to be a bigger joke than Fermi's delay :down:
Regardless of the AMD vs nVIDIA, you do realize that there's a huge difference in producing Evergreen chips ( IRL we should only count the big ones, fully fledged XT ) and Fermi.
I think TSMC "fixed" their 40nm tech process more times than Nvidia announced that Fermi is "Coming soon (tm)!" :rofl:Quote:
Originally Posted by BenchZowner
That would be the issues with vias degrading (which they got around by using double vias, I think), and other things that were difficult with TSMC's 40nm process, right?Quote:
Originally Posted by Nedjo
great blog with lots of pictures about this topic:
http://danielnenni.com/2010/02/28/ts...ndries-vs-ibm/
interesting but no real news... thx though :toast: