Bystander50 5th Gear February 10, 2012 Share February 10, 2012 Because TSMC's higher yield, they can have higher profit margin. then they can have more resources to invest in the future technology. This becomes a vicious cycle for Chartered. To break the cycle, Chartered made a bold move to work with IBM. IBM has its own spec. You can view IBM as mini TSMC in the foundry business. Through the cooperation with IBM, Chartered can copy IBM process directly, recipe by recipe. The manufacturing process in IBM facility and chartered facility are identical. In this way, Chartered can ramp up their technology deployment and yield faster. However, this efforts was failed, mainly because of lack of customers. The main design houses stayed with TSMC camp and did not switch to IBM side. From what I know, IBM side have only 2 major customer: microsoft (Xbox chip) and AMD. No, IBM and Chrt fabs are not identical. Chrt uses LAM Research tools for most of their front end processing while IBM uses TEL. And for backend processes, it is the reverse. The move to work with IBM was basically taking the shortcut. Chrt is already lagging behind the race by 2 generations. It's like paying $ to catch up, but the sad thing is whatever that was developed over at IBM's East Fishkill site by Chrt engineers, IBM holds the ownership. This joint partnership includes Infineon and Samsung, but i think they may have pulled out. ↡ Advertisement Link to post Share on other sites More sharing options...
Bystander50 5th Gear February 10, 2012 Share February 10, 2012 wafer Fab we are losing to Taiwan, not china. Unless China cleans up with it's IP act, no high tech or advance tech products will be manufactured there. You never know when if you ordered 100 wafers and 150 are produced, with the 50 going into the black market or your technology reversed engineered. But i would say China is catching up on this. My bet is probably in another 5-7 years, GF may completely pull out of sg. Unless there are plans to refurbish some of the older fabs in Woodlands for 28nm and beyond, i find it hard to survive for long. If you do remember, chrt's fab 1 at Science Oark was closed after 20 years. Other than fab7, the rest of the fabs are still producing on 8" wafers. Tech (now Micron) already converted from 8" to 12". Link to post Share on other sites More sharing options...
Wyvern 1st Gear February 10, 2012 Share February 10, 2012 You'd better be more careful of what you post. Plenty of police reports being made now for posts such as this. I used to work in Chartered, to be fair to Chia, he took over Chartered when it was bleeding. He's an accountant by trade, I believe (citation required), and he stripped away what was non-profitable, making painful, but necessary cuts (the killing of Fab1 in Science Park for example) and pointed Chartered towards the future. Chartered's downfall was not because of his management skills, but rather, they didn't have the technological know-how. TSMC and UMC were miles ahead, and CSM was always playing catchup. Chia realised this and solidified CSMs standing in older, but necessary tech (read, not cutting edge, not sexy, but makes money) and used the funds to focus R&D on a narrow range of technologies. IMO, he did the best anyone out there could do. He was always first in the office, on some days, I would come in at 7am...and his black BMW 6 series would be out front, I could never beat him! He was a people person (you could often find him at the food court in Vista Point) and I would like to say that he ensured this was inherited by all his managers, me, as a lowly engineer, would find myself regularly in meetings with VPs and Senior Directors, something which didn't mean much to me then, but now, with a few more years under my belt and a few more companies too, I do appreciate it's importance in feeling committed and accepted by the company. I say, all the best to him (and ultimately, us!) Totally Agree..! I was from Chartered Fab 1 which was close down on 2004. Back then my emp id is still 4 digits. Not sure whats now... Link to post Share on other sites More sharing options...
Machspeed Neutral Newbie February 10, 2012 Share February 10, 2012 Unless China cleans up with it's IP act, no high tech or advance tech products will be manufactured there. You never know when if you ordered 100 wafers and 150 are produced, with the 50 going into the black market or your technology reversed engineered. But i would say China is catching up on this. My bet is probably in another 5-7 years, GF may completely pull out of sg. Unless there are plans to refurbish some of the older fabs in Woodlands for 28nm and beyond, i find it hard to survive for long. If you do remember, chrt's fab 1 at Science Oark was closed after 20 years. Other than fab7, the rest of the fabs are still producing on 8" wafers. Tech (now Micron) already converted from 8" to 12". the 8" lines at GF are their cash cow, probably with the exception of fab3/5. i also know of some fabs running 5", 6" and still staying afloat, so keeping their 8" lines might still be feasible. in anycase, their 12" lines are also not fully loaded, don't really make sense to convert more 8" to 12" then again, i could be wrong. =) Link to post Share on other sites More sharing options...
Machspeed Neutral Newbie February 10, 2012 Share February 10, 2012 Totally Agree..! I was from Chartered Fab 1 which was close down on 2004. Back then my emp id is still 4 digits. Not sure whats now... it's 5 digits now, some starts at 1 while some at 7?? Link to post Share on other sites More sharing options...
Happily1986 5th Gear February 10, 2012 Share February 10, 2012 Anyway regarding's GloFlo's trouble with the 28nm process, my guess was the photolithography process. Anyone hazard a guess? Link to post Share on other sites More sharing options...
Klinz Neutral Newbie February 10, 2012 Share February 10, 2012 From what I know, IBM side have only 2 major customer: microsoft (Xbox chip) and AMD. Hmm...then Chia business acumen got problem...anyone with no Financial background knows that we shld follow the leader not to re-invent the wheel provided you are one who Think Different...otherwise ..(no need to say any further we already know what happen) As for Creative, another CSM in making except this one worse....No one wants to buy...if only Mr Sim went into manufacture mobile phones 10yrs ago..maybe Google may buy lor... ..At the rate Creative is heading, it will breed to death..... Link to post Share on other sites More sharing options...
Windwaver Turbocharged February 10, 2012 Share February 10, 2012 he say Semicon is a sunset industry in Sg And I agree with him Compared to finance, yeah Link to post Share on other sites More sharing options...
Windwaver Turbocharged February 10, 2012 Share February 10, 2012 Singapore failed in high tech and will continue to fail in all future tech venture for a simple reason. Because singapore dont believe in science and engineering which is its foundation. What happened to all the local talents Link to post Share on other sites More sharing options...
Somewhat1975 6th Gear February 10, 2012 Share February 10, 2012 What happened to all the local talents banker/insurance agent/property agent. Link to post Share on other sites More sharing options...
Joseph22 Turbocharged February 10, 2012 Share February 10, 2012 he say Semicon is a sunset industry in Sg And I agree with him yeah saw it.. but too late reply liao... Link to post Share on other sites More sharing options...
Wind30 Turbocharged February 10, 2012 Author Share February 10, 2012 I don't think that one can blame Chia for the downfall of chartered. Foundry is a difficult business for anyone, except TSMC. Let me explain why. Customers of foundry are made of those fabless design house, such as broadcom, Qualcom etc. These design house designs their circuit based on a set of circuit parameters. Where do they get those parameters from? From the birth of foundry industry till today, they get those parameters from TSMC. After TSMC develops a new process, they measures their parameters and then gives them to design house. sometimes they called it "T-spec" After the design houses design their IC according to TSMC's specs, they will stick with it. When they do business with Chartered (or UMC), they demand chartered /UMC to produce their chip according to TSMC specs. This is not an easy job for engineers at Chartered/UMC. Chip fabrication nowadays can easily consists of 200-300 processes steps. They have to tweak their processes to match more than 100 parameters from different size of transistors.Usually, it takes months and most of the time, only partial of those parameters are matched. This lowers the yield. Thus, from the outsider point of view, Chartered and UMC always lag behind TSMC in term of technology deployment and yield. TSMC does not need to tweak their process to meet other's spec. They are the rule maker. Because TSMC's higher yield, they can have higher profit margin. then they can have more resources to invest in the future technology. This becomes a vicious cycle for Chartered. To break the cycle, Chartered made a bold move to work with IBM. IBM has its own spec. You can view IBM as mini TSMC in the foundry business. Through the cooperation with IBM, Chartered can copy IBM process directly, recipe by recipe. The manufacturing process in IBM facility and chartered facility are identical. In this way, Chartered can ramp up their technology deployment and yield faster. However, this efforts was failed, mainly because of lack of customers. The main design houses stayed with TSMC camp and did not switch to IBM side. From what I know, IBM side have only 2 major customer: microsoft (Xbox chip) and AMD. you write very long, but still I cannot see why we cannot blame Chia for Chartered Downfall. Was he the CEO who led Chartered into the IBM Alliance? Did the IBM Alliance fail? How can the CEO avoid responsibility in the failure of the Strategy he devised? Common, every corporate Failure has it own reasons. Market not good, no customers. Competition too good. Are those valid reasons? Do you expect TSMC to lie down and let you in? When I design IC chips for my customers, if I am the 2nd source, I am expected to follow all the specs of the 1st source. Is that a reason for my chip to fail? Link to post Share on other sites More sharing options...
Wind30 Turbocharged February 10, 2012 Author Share February 10, 2012 So ok lah....I bought quite a few many years back...jus blamed it on bad timing n tech burst... As I said, I design IC for a living... my wife works in Chartered previously. You think I will buy CSM??? Link to post Share on other sites More sharing options...
Wind30 Turbocharged February 10, 2012 Author Share February 10, 2012 i kinda disagree. i believe the Chief Technology Officer cannot be a non techie, that's true. But CEO can be a non techie. this is becuase the success of tech biz does not lie merely in how good a product is engineered. if u study Marketing, u will know it is about satisfying the demands of the consumers, not about engineering a the best product u can, u can engineer all u want but the consumers still do not like it. e..e gCreative mp3 players vs Ipod. so the problem with tech CEO is that if he is not aware of aspects like marketing, he will be very myopic in his biz decisions in basing on just engineering. Then there is the finance aspect of business like being able to raise capital etc. err.. we are talking about Wafer Fab industry. I think CEO of a wafer fab must be a techie. Everything literally rest on getting up your process nodes. Wafer fab is not selling errr... Ipads. Marvell is not going to produce their chips in Chartered because of a good marketing pitch. Ultimately, it is the performance of the transistors, your process variation, and cost. I believe both technical and management skills is a must. Frankly, any candidates you are looking at for the position of CEO will have YEARS of management experience so that is a foregone conclusion. Process knowledge is very impt. What did Temasek do? they bring in an accountant with zero Wafer Fab knowledge. Ran Chartered Semi to oblivion. What did Temasek learn??? they appoint the very guy as head Strategy of Temasek.... Link to post Share on other sites More sharing options...
Somewhat1975 6th Gear February 10, 2012 Share February 10, 2012 you write very long, but still I cannot see why we cannot blame Chia for Chartered Downfall. Was he the CEO who led Chartered into the IBM Alliance? Did the IBM Alliance fail? How can the CEO avoid responsibility in the failure of the Strategy he devised? Common, every corporate Failure has it own reasons. Market not good, no customers. Competition too good. Are those valid reasons? Do you expect TSMC to lie down and let you in? When I design IC chips for my customers, if I am the 2nd source, I am expected to follow all the specs of the 1st source. Is that a reason for my chip to fail? what i was trying to say is that foundry is a damn tough business for everyone, except TSMC. Foundry business model was invented by TSMC and controlled by TSMC. TSMC makes the rule and plays the game. Even you have the money and capable people， it will be damn hard to challenge TSMC. UMC, chartered and even SMIC tried to challenge TSMC and failed. It was a dead end for chartered to follow TSMC model. Chia tried to break out and failed. but at least it is better than doing nothing and strolled into the dead. Chartered had been struggled long before he became CEO. You can't blame Chartered downfall solely on his capability. Link to post Share on other sites More sharing options...
Relacklabrudder 1st Gear February 10, 2012 Share February 10, 2012 Well if ur in NAND Flash biz, IMFlash is doing it all On their own apart from tsmc. I tink Koreans are also on their own with Hynix and Samsung Link to post Share on other sites More sharing options...
Somewhat1975 6th Gear February 10, 2012 Share February 10, 2012 Well if ur in NAND Flash biz, IMFlash is doing it all On their own apart from tsmc. I tink Koreans are also on their own with Hynix and Samsung IMFlash= flash memory, samsung Hynix=DRAM, they are not foundry. Link to post Share on other sites More sharing options...
Relacklabrudder 1st Gear February 10, 2012 Share February 10, 2012 (edited) IMFlash= flash memory, samsung Hynix=DRAM, they are not foundry. Errr I already mention imflash is in the NAND flash biz. Flash memory got 2 types NOR and NAND. Imflash manufacture NAND flash for micron and intel, they are not ever gg to tsmc. Same with Koreans , die die will not go tsmc I am saying they have their own wafer plants with their own technology development, their own recipes. They are not relying on contract manufacturing to foundries like TSMC Edited February 10, 2012 by Relacklabrudder ↡ Advertisement Link to post Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In NowRelated Discussions
Related Discussions
Any Hokkien Song Lovers here?
Any Hokkien Song Lovers here?
Any nice chinese song to recommend?
Any nice chinese song to recommend?
Anyone into HK cantonese song (80s~end 90)
Anyone into HK cantonese song (80s~end 90)
KNS car park management system at Jalan Kuras
KNS car park management system at Jalan Kuras
Extend retirement age? Maybe it's time to think about that from another perspective
Extend retirement age? Maybe it's time to think about that from another perspective
Josephine Teo: GIC, Temasek won't take more risks
Josephine Teo: GIC, Temasek won't take more risks
Temasek Holdings ...... Piggy Bank broken?
Temasek Holdings ...... Piggy Bank broken?
Singapore's CapitaLand adds D.C., Dhaka apartments with Oakwood buy
Singapore's CapitaLand adds D.C., Dhaka apartments with Oakwood buy