• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Samsung Considers Foundry Division Spin-Off as Poor 3 nm Yields Deter Customers

AleksandarK

News Editor
Staff member
Joined
Aug 19, 2017
Messages
3,207 (1.11/day)
The grass isn't always greener on the other side, especially as we're running out of sides in the advanced semiconductor manufacturing sector. A recent report by Business Korea highlights Samsung Securities' July publication titled "Geopolitical Paradigm Shift and Industry," which paints a less-than-optimistic picture of Samsung's current state of affairs. The report even evaluates a possible spinoff of Samsung Foundry. The Korean tech giant has faced various business setbacks related to its state-of-the-art 3 nm Gate-All-Around (GAA) FET node. Reports indicate that this node only manages to yield 10-20% of working silicon, making potential customers reluctant to secure partnerships with Samsung. Samsung Securities projects that Samsung Foundry, along with the LSI division, will suffer a 500 billion won (about $385 million) loss this year.

Poor yields and difficulty securing customers have left Samsung facing tough choices, including the possible sale of its massive Foundry unit, which manufactures logic for external customers. It's noteworthy that Samsung is one of only three companies left in the advanced semiconductor manufacturing field, alongside TSMC and Intel. Many companies struggled to deliver results when transitioning to sub-7 nm nodes. Global Foundries dropped out of the race to focus on mature nodes, while Intel faced delays. TSMC has been the only company so far to consistently set and execute its goals, positioning itself as the industry leader. With low yields on the 3 nm GAA FET node, Samsung currently holds 11.5% of the global foundry market share in Q2, while TSMC dominates with 62.3%.



View at TechPowerUp Main Site | Source
 
Oh dear, that's not good. But i guess they are doing ok on their other stuff, phones etc.
 
Oh dear, that's not good. But i guess they are doing ok on their other stuff, phones etc.
Their phone divison is top-notch, and can use any 3rd party SoCs like the ones from Qualcomm.

Memory/storage chips are okay. TVs and home appliances are also selling well. Samsung overall is okay :)
 
Their phone divison is top-notch, and can use any 3rd party SoCs like the ones from Qualcomm.

Memory/storage chips are okay. TVs and home appliances are also selling well. Samsung overall is okay :)

Good job eh. Why sell the foundrys though, seems a bit of a fail rther than try and sort it out.
 
Good job eh. Why sell the foundrys though, seems a bit of a fail rther than try and sort it out.
Ask GloFo the same thing, ask Intel why 10 nm was late. Not everyone can make it in the semiconductor world ;)
 
A 10-20% yield... I would say they are making dumb decisions.
 
Good job eh. Why sell the foundrys though, seems a bit of a fail rther than try and sort it out.
Because sorting it out is EXPENSIVE? And at the same time shareholders strongly dislike the economic pain this puts the rest of the company in. Also see - Intel. And AMD ca. 2009-2012.
 
Because sorting it out is EXPENSIVE? And at the same time shareholders strongly dislike the economic pain this puts the rest of the company in. Also see - Intel. And AMD ca. 2009-2012.
Except they have to sort it out if they have any kind of orders lined up. It's not optional to not deliver to customers in a situation like this.

As for spinning things off, keep in mind that this is a Korean company we're talking about, so it will end up being a subsidiary owned by Samsung somehow anyhow, or a Samsung shell company, since Koreans do not want outside investment into their chaebols. Korean owned businesses tend to stay Korean owned.

On top of that, Samsung still need access to these fabs to produce all the chips they make, including DRAM and NAND, so Samsung is going to want to be in control.
 
Good job eh. Why sell the foundrys though, seems a bit of a fail rther than try and sort it out.
Once you're too big, things like this start to happen.
 
"Global Foundries dropped out of the race to focus on mature nodes, while Intel faced delays. TSMC has been the only company so far to consistently set and execute its goals, positioning itself as the industry leader. With low yields on the 3 nm GAA FET node, Samsung currently holds 11.5% of the global foundry market share in Q2, while TSMC dominates with 62.3%."
1727895762903.png
 
Their phone divison is top-notch, and can use any 3rd party SoCs like the ones from Qualcomm.

Memory/storage chips are okay. TVs and home appliances are also selling well. Samsung overall is okay :)

Samsung home appliances are GOD-AWFUL. It’s a miracle that they are still selling (let’s be real, the Samsung logo probably carries hard for them in the home appliance division).
 
Once Intel 18A node is online TSMC is going to have problemas.
 
If Samsung were a US company they would be split up for monopoly reasons
 
Samsung making a decent and competitive node challenge (impossible)

I still remember when N3 hadn't really gotten off the ground, but Samsung was loudly trumpeting its GAAFET "successes".

When was the last time Samsung foundries had a winner on their hands? 7 years ago?
 
Is that what Samsung said or is this what some """"analyst"""" made up from the reports.
the same type of """""analyst""""" that said Sony should sell their sensor division

No one sane and not evil wants samsung to quit semiconductors.
a world where the only options are TSMC and SMIC that do the latest nodes is not going to be a nice world for tech.
 
Once Intel 18A node is online TSMC is going to have problemas.

I’m not betting on Intel 18A re-establishing the lead for Team Blue. Intel spinning off their foundries means those foundries will have to meet production targets for other customers in addition to Intel. Delays for poor yields would not surprise me.
 
Once Intel 18A node is online TSMC is going to have problemas.

Intel's history of delivering new nodes over the past decade strongly argues for TSMC keeping their comfy lead in process technology.
 
Their phone divison is top-notch, and can use any 3rd party SoCs like the ones from Qualcomm.

Memory/storage chips are okay. TVs and home appliances are also selling well. Samsung overall is okay :)
There are strikes everywhere in Samsung right now and they are laying off staff across the board. Doesnt look like they are ok.
 
Samsung home appliances are GOD-AWFUL. It’s a miracle that they are still selling (let’s be real, the Samsung logo probably carries hard for them in the home appliance division).
That probably depends on quality control procedures at newish factories these days. I have a few Samsung home appliances here that have lasted anywhere between 8 - 14 yrs already & one of them has been on 24/7 for 8+ yrs already, not missing a beat.

Back to the news though, what about Micron? there still in the game for making their own ICs for RAM as an example of a product line from their portfolio.
 
Last edited:
Once Intel 18A node is online TSMC is going to have problemas.
Not very likely. Intel's main foundry client is their own stuff anyway and building a client base for a foundry takes years if not decades.
In terms of not being the only one with new tech - yes. In terms of real business - not any time soon.
 
Not very likely. Intel's main foundry client is their own stuff anyway and building a client base for a foundry takes years if not decades.
In terms of not being the only one with new tech - yes. In terms of real business - not any time soon.
If they have the lead and have decent yields they would be fully booked very quickly by a small army of ai venture capital that need their custom chips made yesterday.

you are right about big clients like apple and nvidia those guys will take quite some convincing to dip their toes in a new fab.
 
Quote: "Adding to the company's woes, the Exynos 2500 chip, made with the GAA 3-nanometer process, has a low yield, making its inclusion in next year's Galaxy S25 uncertain. The 2-nanometer process has also faced delays, further complicating Samsung's technological roadmap."
출처 : Businesskorea (https://www.businesskorea.co.kr)

Let´s wait and see...
 
If they have the lead and have decent yields they would be fully booked very quickly by a small army of ai venture capital that need their custom chips made yesterday.

you are right about big clients like apple and nvidia those guys will take quite some convincing to dip their toes in a new fab.
unless the department of commerce is on you, and you're not getting any federal contractor business since they favor domestic tech.
U.S. Govt pushes Nvidia and Apple to use Intel's foundries — Department of Commerce Secretary Raimondo makes appeal for US-based chip production | Tom's Hardware (tomshardware.com)

I think we are underestimating how difficult regulatory life will get for TSMC once the 18A fab is even an option.
 
Last edited:
unless the department of commerce is on you, and you're not getting any federal contractor business since they favor domestic tech.
U.S. Govt pushes Nvidia and Apple to use Intel's foundries — Department of Commerce Secretary Raimondo makes appeal for US-based chip production | Tom's Hardware (tomshardware.com)

I think we are underestimating how difficult regulatory life will get for TSMC once the 18A fab is even an option.
Those would probably shut up if they made something at intel.
nvidia just making the small die for gamers
 
If Samsung were a US company they would be split up for monopoly reasons
I mean yeah but you can say that about a lot of large Asian conglomerates (Matsushita/Panasonic and Hitachi both come to mind), honestly. They have their fingers in a lot of pies and they don't get split up. Just a different way they manage their businesses.
 
Back
Top