You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
TAIPEI, Taiwan--(BUSINESS WIRE)--United Microelectronics Corporation (NYSE: UMC; TWSE: 2303) (“UMC”), today reported unaudited net sales for the month of December 2025.
Additional information about UMC is available on the web at https://www.umc.com.
This is very sad. TSMC is breaking revenue records while UMC and GF are treading water and it will only get worse. TSMC added 28nm capacity in Japan which is targeted at UMC and GF. Dresden is next which is a direct hit on GF. SMIC has a lock on the China market. We really need Intel Foundry to make a difference here.
This is very sad. TSMC is breaking revenue records while UMC and GF are treading water and it will only get worse. TSMC added 28nm capacity in Japan which is targeted at UMC and GF. Dresden is next which is a direct hit on GF. SMIC has a lock on the China market. We really need Intel Foundry to make a difference here.
Are GF getting no benefits from the Governments attempts to bring manufacturing back to the US?
UMC does appear to be in that middle income trap and unable to move up the ladder.
They have just opened a new 22nm extension FAB here in Singapore , but as you say , that marketplace is getting more crowded by the day
Oddly GF appears to be centralising FAB planning, process and various other FAB functions offsite in Penang Malaysia, which seems odd as they have no FABs in Malaysia.
Oddly GF appears to be centralising FAB planning, process and various other FAB functions offsite in Penang Malaysia, which seems odd as they have no FABs in Malaysia.
GF has a new CEO, he was the COO so expect changes. Before joining GlobalFoundries, Tim Breen was on the senior leadership team at Mubadala Investment Company, the firm’s majority shareholder, where he worked on global projects and investments across multiple sectors.
One of the changes I expect is M&A. They have already purchased MIPS and I expect more to come. Revenue in 2026 is expected to be +6% which is not good. Tim needs to get that back into the double digits, absolutely.
There is also talk of Intel Foundry Acquiring GlobalFoundries. Pat Gelsinger could not get the Tower Semiconductor acquisition through but I expect Lip-Bu Tan could. Tower Semi really dodged a bullet there. Maybe Lip-Bu and Donald Trump discussed it at the last meeting. We can start that rumor right here.
There is also talk of Intel Foundry Acquiring GlobalFoundries. Pat Gelsinger could not get the Tower Semiconductor acquisition through but I expect Lip-Bu Tan could. Tower Semi really dodged a bullet there. Maybe Lip-Bu and Donald Trump discussed it at the last meeting. We can start that rumor right here.
GF should acquire IFS. Not the other way around. (and both options have been discussed extensively and in detail for past 6 years).
Back to statistics. I am suprised the tide isnt lifting all boats more. Is it fair to say the wild boom is only leading edge nodes? And even there, Samsung and IFS have done little to nothing on growth (Intel has grown TSMC, but not IFS). So is the boom just TSMC, AMD, Nvidia, broadcom?
GF should acquire IFS. Not the other way around. (and both options have been discussed extensively and in detail for past 6 years).
Back to statistics. I am suprised the tide isnt lifting all boats more. Is it fair to say the wild boom is only leading edge nodes? And even there, Samsung and IFS have done little to nothing on growth (Intel has grown TSMC, but not IFS). So is the boom just TSMC, AMD, Nvidia, broadcom?
GF is still foreign owned so I do not see that happening. GF is stuck at 14nm and it is not even their 14nm, they licensed it from Samsung. Get the GF sales force to start filling the Intel 7, 3, and 18A fabs. Intel could also use the MIPs business.
TSMC has done a great job of pushing customer to newer nodes. Thanks to Apple TSMC is an expert on low power chips and that is a big driver to new nodes. Mobile and AI datacenter stuff needs to be low power. To me that explains the stampede from TSMC N3 to TSMC N2. It is also a big differentiator between N2 and Intel 18A or Samsung 2nm. We cannot build enough green power so every little bit helps.
75% of TSMC's revenue is from 7nm and below so yes it is mostly leading edge stuff and GF is not participating in that today.
GF is still foreign owned so I do not see that happening. GF is stuck at 14nm and it is not even their 14nm, they licensed it from Samsung. Get the GF sales force to start filling the Intel 7, 3, and 18A fabs. Intel could also use the MIPs business.
TSMC has done a great job of pushing customer to newer nodes. Thanks to Apple TSMC is an expert on low power chips and that is a big driver to new nodes. Mobile and AI datacenter stuff needs to be low power. To me that explains the stampede from TSMC N3 to TSMC N2. It is also a big differentiator between N2 and Intel 18A or Samsung 2nm. We cannot build enough green power so every little bit helps.
75% of TSMC's revenue is from 7nm and below so yes it is mostly leading edge stuff and GF is not participating in that today.
Yes, everyone has optimized their 14nm process to 12nm which is why I said "Intel 7, 3, and 18A fabs". The tape-out would come from the foundry customer of course. I'm talking about new designs.
Intel buying an Israeli company (Tower Semi) for foundry support was not politically correct. The current administration would be much more supportive of Intel buying and American/Abud Abi foundry, my opinion. THe current GF CEO is there to sell the company. Look at his credentials. Again, my opinion.
Lip-Bu is the ultimate deal maker. Can you now think of a better CEO for Intel than Lip-Bu Tan after how he handled Trump and got the other investments from Nvidia and Softbank? Lip-Bu can make the GF acquisition happen, absolutely.
Yes, everyone has optimized their 14nm process to 12nm which is why I said "Intel 7, 3, and 18A fabs". The tape-out would come from the foundry customer of course. I'm talking about new designs.
On this Intel 7 is using custom Intel Tools and is expensive so i don't see any reason to market it for foundry and TSMC N7/N6 is cheaper and not constrained like N5/4/3/2
On this Intel 7 is using custom Intel Tools and is expensive so i don't see any reason to market it for foundry and TSMC N7/N6 is cheaper and not constrained like N5/4/3/2
Not true. Foundry PDKs can be made. TSMC wafers are never cheaper to the customer. TSMC gets a premium for their wafers as they should. Intel has already depreciated the older nodes and yield is optimized so this can be done with a profit. Maybe not at a 60% margin like TSMC but lower margins like GF and UMC.
Not true. Foundry PDKs can be made. TSMC wafers are never cheaper to the customer. TSMC gets a premium for their wafers as they should. Intel has already depreciated the older nodes and yield is optimized so this can be done with a profit. Maybe not at a 60% margin like TSMC but lower margins like GF and UMC.
their process is pretty expensive as well they need to redo some part of their process to make it cost optimized right now Intel 7 Process is tuned for transistor performance power be dammed. Intel themselves admit how costly Intel 7 is
Yes, everyone has optimized their 14nm process to 12nm which is why I said "Intel 7, 3, and 18A fabs". The tape-out would come from the foundry customer of course. I'm talking about new designs.
Intel buying an Israeli company (Tower Semi) for foundry support was not politically correct. The current administration would be much more supportive of Intel buying and American/Abud Abi foundry, my opinion. THe current GF CEO is there to sell the company. Look at his credentials. Again, my opinion.
Lip-Bu is the ultimate deal maker. Can you now think of a better CEO for Intel than Lip-Bu Tan after how he handled Trump and got the other investments from Nvidia and Softbank? Lip-Bu can make the GF acquisition happen, absolutely.
