GLOBALFOUNDRIES To Acquire IBM's Microelectronics Business

[Cristian_25H]

IBM and GLOBALFOUNDRIES today announced that they have signed a Definitive Agreement under which GLOBALFOUNDRIES plans to acquire IBM's global commercial semiconductor technology business, including intellectual property, world-class technologists and technologies related to IBM Microelectronics, subject to completion of applicable regulatory reviews. GLOBALFOUNDRIES will also become IBM's exclusive server processor semiconductor technology provider for 22 nanometer (nm), 14 nm and 10 nm semiconductors for the next 10 years.

The Agreement, once closed, enables IBM to further focus on fundamental semiconductor research and the development of future cloud, mobile, big data analytics, and secure transaction-optimized systems. IBM continues its previously announced $3 billion investment over five years for semiconductor technology research to lead in the next generation of computing. GLOBALFOUNDRIES will have primary access to the research that results from this investment through joint collaboration at the Colleges of Nanoscale Science and Engineering (CNSE), SUNY Polytechnic Institute, in Albany, N.Y.



As part of this Agreement, GLOBALFOUNDRIES will gain substantial intellectual property including thousands of patents, making GLOBALFOUNDRIES the holder of one of the largest semiconductor patent portfolios in the world. GLOBALFOUNDRIES also will benefit from an influx of one of the best technical teams in the semiconductor industry, which will solidify its path to advanced process geometries at 10 nm and below. Additionally, the acquisition opens up business opportunities in industry-leading radio frequency (RF) and specialty technologies and ASIC design capabilities.

"This acquisition solidifies GLOBALFOUNDRIES' leadership position in semiconductor technology development and manufacturing," said Dr. Sanjay Jha, CEO, GLOBALFOUNDRIES. "We can now offer our customers a broader range of differentiated leading-edge 3D transistor and RF technologies, and we will also improve our design ecosystem to accelerate time-to-revenue for our customers. This acquisition further strengthens advanced manufacturing in the United States, and builds on established relationships in New York and Vermont."

"The Agreement expands our longstanding collaboration, which began when GLOBALFOUNDRIES was created in 2009, and reflects our confidence in GLOBALFOUNDRIES' capability," said IBM Senior Vice President & Director of Research Dr. John E. Kelly III. "This acquisition enables IBM to focus on fundamental semiconductor and material science research, development capabilities and expertise in high-value systems, with GLOBALFOUNDRIES' leadership in advanced technology manufacturing at scale and commitment to delivering future semiconductor technologies. We are grateful for the leadership and investments by the states of New York and Vermont in supporting the semiconductor industry."

This acquisition bolsters semiconductor manufacturing and technology development in the United States. GLOBALFOUNDRIES has robust capital expenditure plans of approximately $10 billion in 2014-2015, with the majority being invested in New York. GLOBALFOUNDRIES has created nearly 3,000 direct jobs in New York and thousands more indirect jobs in the United States since 2009.

GLOBALFOUNDRIES will acquire and operate existing IBM semiconductor manufacturing operations and facilities in East Fishkill, New York and Essex Junction, Vermont, adding capacity to serve its customers and thousands of jobs to GLOBALFOUNDRIES' workforce. GLOBALFOUNDRIES plans to provide employment opportunities for substantially all IBM employees at the two facilities who are part of the transferred businesses, except for a team of semiconductor server group employees who will remain with IBM. After the close of this transaction, GLOBALFOUNDRIES will be the largest semiconductor technology manufacturing employer in the Northeast.

GLOBALFOUNDRIES will also acquire IBM's commercial microelectronics business, which includes ASIC and specialty foundry, manufacturing and related operations and sales. GLOBALFOUNDRIES plans to invest to grow these businesses.

IBM will reflect a pre-tax charge of $4.7 billion in its financial results for the third quarter of 2014, which includes an asset impairment, estimated costs to sell the IBM microelectronics business, and cash consideration to GLOBALFOUNDRIES. Cash consideration of $1.5 billion is expected to be paid to GLOBALFOUNDRIES by IBM over the next three years. The cash consideration will be adjusted by the amount of working capital which is estimated to be $200 million.

The transaction is subject to the satisfaction of regulatory requirements and customary closing conditions.

View at TechPowerUp Main Site

 

[Frick]

Huh. What does this do for AMD? (that is what affects me the most)

 

[RCoon]

Huh. What does this do for AMD? (that is what affects me the most)

I'm pretty sure GloFo just bought a manufacturing process that was causing a massive financial loss to IBM. Beyond that I don't know the details.

"IBM has agreed to pay the Abu Dhabi-owned firm GlobalFoundries $1.5bn (£930m) to take its unprofitable chip-making business off its hands,"

SOURCE

 

[FordGT90Concept]

So IBM threw Global Foundries a bone by removing dead weight but what makes Global Foundries think IBM's IP is of any use? I think Global Foundries really wanted that $1.5 billion cash.

No reason why regulators would stop it.

 

[Jorge]

This purchase by Global foundries has no immediate impact on AMD though they have worked with IBM in the past. As long as GloFo continues to deliver quality wafers and AMD properly ramps their CPU/APU designs, all is good for AMD consumers. We'll see if Lisa Su can ramp AMD engineering to get AMD back up to speed in all areas not just APUs, embedded, mainstream, etc.

 

[crazyeyesreaper]

THe big thing is this increases GloFo ability to meet demand and strengthens their position against the likes of TSMC. WIth the ability to push out more wafers etc they can meet more demand and take on more customers. The extra money on hand is good as well. The IP might not be worthwhile right now but if other companies use any of those patents GloFo will get a paycheck because of it. This should strengthen their position while making it easier for IBM to mess around with self healing polymers and replacements for silicon based wafers.

 

[Patriot]

Wonder if IBM limits them from using 22nm,14m, and 10m, on other customers products.

 

[rtwjunkie]

Does this mean GloFo picks up their chip manufacturing facilities as well? I know of several in the U.S. Northeast that were chip-producing plants in excess of 5,000 employees each.

 

[Frick]

Does this mean GloFo picks up their chip manufacturing facilities as well? I know of several in the U.S. Northeast that were chip-producing plants in excess of 5,000 employees each.

GLOBALFOUNDRIES will acquire and operate existing IBM semiconductor manufacturing operations and facilities in East Fishkill, New York and Essex Junction, Vermont, adding capacity to serve its customers and thousands of jobs to GLOBALFOUNDRIES' workforce.

Looks like it.

 

[rtwjunkie]

Looks like it.

DOH!, I can't believe I missed that while looking for it specifically! Thanks, Frick!

 

[HalfAHertz]

This could mean that IBM has made a breakthrough in graphene or similar non silicone based transistor manufacturing?

 

[RCoon]

This could mean that IBM has made a breakthrough in graphene or similar non silicone based transistor manufacturing?

That, or they just want to stop leaking a few billion dollars every year.

 

[rtwjunkie]

I can tell you, having worked at one of those two plants Frick mentioned, IBM's money loss wasn't because of silicon manufacturing. I've never seen a company so lavish with their money. Anything that was needed, done. Any refabbing needed? Let's spend double to get it done faster. Anything that needed to be purchased, Done!

 

[RCoon]

I can tell you, having worked at one of those two plants Frick mentioned, IBM's money loss wasn't because of silicon manufacturing. I've never seen a company so lavish with their money. Anything that was needed, done. Any refabbing needed? Let's spend double to get it done faster. Anything that needed to be purchased, Done!

Didn't know you worked in a fab plant. That's pretty interesting!

 

[rtwjunkie]

Didn't know you worked in a fab plant. That's pretty interesting!

Nothing all that glamorous. I was the site security supervisor. I just ran the guard force, still answered to a security director. I did get to pick up on alot of stuff though, had free run of the plant (except clean rooms, although you could watch some of them), engineers, managers. Had a couple friends who actually worked on the line. So, whatever I picked up is just trivia, and good conversation at parties, but has no use in real life!

 

[xenocide]

I can tell you, having worked at one of those two plants Frick mentioned, IBM's money loss wasn't because of silicon manufacturing. I've never seen a company so lavish with their money. Anything that was needed, done. Any refabbing needed? Let's spend double to get it done faster. Anything that needed to be purchased, Done!

That must have been back in the 90's and early 2000's.

Didn't know you worked in a fab plant. That's pretty interesting!

<---- *looks around* No comment.

 

[rtwjunkie]

@xenocide: Yep, in the 90's.

 

[HumanSmoke]

THe big thing is this increases GloFo ability to meet demand and strengthens their position against the likes of TSMC.

Importantly, it gives GloFo access to more than one foundry process. GloFo binned their own finfet process and licensed Samsung's 14nm. Now they have a choice since IBM also have a 14nm process. Two avenues of R&D, two separate process nodes that can possibly offer a better tailored solution to any given ASIC.

WIth the ability to push out more wafers etc they can meet more demand and take on more customers.

Not initially, it's my understanding that both Fishkill and Essex Junction are both due for refurbishment

The extra money on hand is good as well. The IP might not be worthwhile right now but if other companies use any of those patents GloFo will get a paycheck because of it.

The cash will be a drop in the bucket for the retooling both plants will likely have to undergo, but the IP is the highlight - I don't think any silicon manufacturers have IBM's experience in SOI, and with IBM's FD-SOI (Fully Depleted) going primetime and showing considerable promise GloFo has a chance to put its stamp on the fabbing landscape rather than playing second/third fiddle to TSMC, Samsung, and Intel.

This should strengthen their position while making it easier for IBM to mess around with self healing polymers and replacements for silicon based wafers.

Yep. IBM have been more into R&D and service than manufacturing for some time. The whole dissolution of the Common Platform /gate-first alliance just removed most of the remaining enthusiasm.

 

[remixedcat]

IBM is MAD hurr hurr get it!!??

 

[xenocide]

Not initially, it's my understanding that both Fishkill and Essex Junction are both due for refurbishment.

One of them is in much better standing than the other, and could take over some work from GF once qualifications are done--a few weeks or months depending. They also make fundamentally different products at this point, and I'm pretty sure GF doesn't have any equivalent facilities in terms of what they are focused on manufacturing.

 

[HumanSmoke]

One of them is in much better standing than the other, and could take over some work from GF once qualifications are done--a few weeks or months depending.

It was my understanding that the Essex Junction plant was basically being maintained for military semicon production, older processes, and third party leasing, and East Fishkill had one state of the art fab (323A) but a lot of the plant structures were somewhat lacking? I note that GloFo have publicly stated that both plants will remain open (obviously for continued IBM production and to collect tax incentives) but there has been absolutely no mention of investment in refurbishment and retooling. If anything, it looks like GloFo's Fab 8 will come out the big winner.

They also make fundamentally different products at this point, and I'm pretty sure GF doesn't have any equivalent facilities in terms of what they are focused on manufacturing.

Likewise, the fabs are set up for IBM chips, so unless GloFo has succeeded in finding takers for IBM's 22nm SOI process where IBM itself failed, it doesn't look like the fabs will suddenly swell with extra production - which was my original point in answering crazyeyesreaper assumption that buying the fabs adds capacity for GloFo, and by extension, GloFo's customers.

 

[xenocide]

It was my understanding that the Essex Junction plant was basically being maintained for military semicon production, older processes, and third party leasing, and East Fishkill had one state of the art fab (323A) but a lot of the plant structures were somewhat lacking? I note that GloFo have publicly stated that both plants will remain open (obviously for continued IBM production and to collect tax incentives) but there has been absolutely no mention of investment in refurbishment and retooling. If anything, it looks like GloFo's Fab 8 will come out the big winner.

Likewise, the fabs are set up for IBM chips, so unless GloFo has succeeded in finding takers for IBM's 22nm SOI process where IBM itself failed, it doesn't look like the fabs will suddenly swell with extra production - which was my original point in answering crazyeyesreaper assumption that buying the fabs adds capacity for GloFo, and by extension, GloFo's customers.

Essex is definitely an older plant, but the technology being made there isn't necessarily. There are certain devices that don't need to be made on the lowest imaginable process node, and that's kind of the market they are fishing within. I remember seeing a yearly earnings report that pointed out TSMC made the most profit in the 120nm+ sector as an indicator that cutting edge processes aren't necessarily the most profitable. Fishkill has a much more advanced fab but I can't comment on what the deal is with that site. As far as I know it's capable of doing some legacy GloFo work (in terms of being able to make parts at the process nodes GloFo is trying to sunset) and has a lot of Engineers on site that GloFo wanted. All fabs built more than 5 years ago require refurbishment to some extent, you can push tools a lot further than they are initially designed to go (trust me on this one) but over a long enough time line you start dealing with core structural problems within the fab.

I don't think it has as much to do with capacity as it does with revenue. GloFo probably wants more customers, more engineers, more money circulating, and just overall more weight. They want to surpass TSMC who is edging them out in manufacturing capabilities and absorbing the IBM Manufacturing divisions gives them a lot more resources--keep in mind a lot of IP also changed hands.

 

[crazyeyesreaper]

Also I never said it was an IMMEDIATE increase in production capabilities. The simple fact is they have more foundries. More foundries = more chips even if older processes etc people still need to pay for that thus income generation. Enough to off set it ? NO. Enough to make it viable that long term they have infrastructure in place to expand without needing to purchase more land build all new locations etc. Retooling is expensive but not nearly as expensive as starting from square one aka buying land / getting permits / designs / sewer / septic / water / electrical lines etc. all this adds up and becomes far more expensive than just retooling. This isnt GloFo making a quick grab. THis is GloFo doing what AMD could not. Moving toward mass production where they can seriously compete with the likes of TSMC.

 

[HumanSmoke]

I remember seeing a yearly earnings report that pointed out TSMC made the most profit in the 120nm+ sector as an indicator that cutting edge processes aren't necessarily the most profitable.

When was this, back in the days when Big Mac (Fab 6) was state of the art? IIRC, TSMC has only a single process larger than 120nm ( 130nm BCD primarily for IC's in display boards, and the follow on 180nm CL/CR018 is still at the development stage for Multi-Purpose Wafer usage), and as has been recounted by TSMC's own quarterly's and reported extensively in the trades, the bulk of TSMC's wafer starts, revenue, and profit are built around 28nm HP/HPM/HPL, 40nm LP/G/LPG, and 20nm SOC.

In 2014, 60% of TSMC’s revenue is expected to be from ≤45nm processing.

AFAIK, the bulk of the remaining revenue and profit actually derives from low cost commodity IC's on the 65nm GP/LP/LPG/ULP processes
TSMC's own capacity bears this out. Fab 15 and 12A/12B both use 28nm and 20nm, while Fab 14A / 14B are 28nm. Admittedly the older process nodes (90nm, 65nm) have already recouped their ROI many times over, but their profit margins are also impacted by the low ASP's associated with commodity chips including a lot of Static RAM products.

All fabs built more than 5 years ago require refurbishment to some extent, you can push tools a lot further than they are initially designed to go (trust me on this one) but over a long enough time line you start dealing with core structural problems within the fab.

Their lifespan like all fabs comes down to the litho tools. They would be originally validated for certain nodes, and could/would be further qualified for smaller processes up to a point. If it were a simple case of revalidating existing tooling in perpetuity we wouldn't be at the present bottleneck where the next generation of process wafer lines are limited by the ability to produce tooling (ASML's current NXE:3300B and next-gen NXE:3350B for example)

I don't think it has as much to do with capacity as it does with revenue. GloFo probably wants more customers, more engineers, more money circulating, and just overall more weight. They want to surpass TSMC who is edging them out in manufacturing capabilities and absorbing the IBM Manufacturing divisions gives them a lot more resources--keep in mind a lot of IP also changed hands.

Agreed. GloFo are looking to the future but I don't think they'll overhaul TSMC in any near/medium future - even maxing out the combined capacity doesn't put them in the same ballpark as TSMC. IBM has been shopping its foundry business for a while (and TSMC are on record as saying they weren't interested). If IBM's IP and foundries represented such a threat it would be almost certain that TSMC would have acquired them just to consolidate its own position even if it leased out production to the current leasers (like Sernam) at break-even/loss (tax incentives included) just to tie up the IP.