Samsung and SK Hynix each reportedly committing ₩10-20 trillion to ASML EUV tools; ASML warns on 2027-2028 supply
Samsung Taylor Fab 2 and AMD HBM
Latest: AMD Reportedly Secures Most of Samsung 2nm Taylor Fab 2 for HBM
HBM base dies increasingly require advanced foundry nodes, making capacity bookings years in advance a meaningful indicator of which fabs will serve AI memory demand. ASML's disclosed inability to fill 2027 and 2028 EUV demand adds a supply constraint that could affect production timelines across the memory industry.
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.
The full picture
Samsung's second Taylor, Texas fab has reportedly pre-booked its full planned 2nm output for HBM base dies, with AMD taking most of the capacity, before construction has even begun. Equipment installation is expected around 2029 with mass production planned for 2030. Samsung has not officially confirmed which products the fab will manufacture. Separately, Nvidia is reportedly asking Samsung, SK Hynix, and Micron to fabricate its custom NVHBM base dies on TSMC's 3nm process, an arrangement that would route memory-related revenue to TSMC without requiring it to enter the DRAM market. TSMC denied a related report that Nvidia had pushed it toward 3D DRAM, stating it will remain a pure logic foundry. In parallel, Samsung and SK Hynix are each reportedly committing between ₩10 trillion and ₩20 trillion to secure ASML EUV lithography tools amid supply constraints, with ASML having formally told customers it cannot produce enough EUV tools to meet 2027 and 2028 demand.
How it developed
Tom's Hardware reports Nvidia asking memory makers to use TSMC 3nm for NVHBM base dies; TSMC denies 3D DRAM push
Tweaktown and igorslab corroborated and extended earlier reports that Samsung's planned Taylor Fab 2, a 250,000-square-meter plant using a 2nm-class process, has secured orders covering its entire planned output for HBM base dies before construction begins, with AMD accounting for most of the volume.
Equipment installation is expected in 2029 and mass production in 2030. Nvidia is also named as a potential buyer of base dies from the same facility.
Sources
Related
- Related storyNvidia Rubin memory cuts and HBM shortage
- Nvidia's NVHBM base die at TSMC
- Nvidia AI server prices and memory shortage
- AI Compute Sold Out Past 2028; Supply Chain Strain Spreads to Adjacent Industries
- Goldman Sachs, PwC Project Massive AI Infrastructure Spending Wave
BackgroundCompute: the chips, clouds and data centers behind AI
Get the daily email
When stories like this move, I write about them in an email each morning.
One free email each morning, around 8 AM US Eastern.
Check your email.
I have sent you a confirmation link. One click and you are in.
Not in your inbox? Look in spam or promotions. If it's there, mark it "Not spam" so future editions land in your inbox.