Close reading of the information from Google AI suggests the relationship between TSMC and Intel is getting very, very close.
The most important "data point" may be the one in red bold down below.
Intel does not currently have a 2 nm chip but is focusing on its 18A (1.8 nm) process node, which is in risk production and expected to be ready for high-volume manufacturing in mid-to-late 2025.
This 18A node uses RibbonFET transistor technology and is intended to be Intel's competitive response to TSMC's 2 nm node. Intel also plans to start mass production of an enhanced 18A-P process in the second half of 2026, though it has also reportedly placed orders with TSMC for some of its Nova Lake processors, which are designed for the 2 nm-class process.
Intel's 18A Node & the 2 nm Label
- Intel's Approach:
Intel has adopted an angstrom-based naming convention for its nodes, with the 18A node representing a process that is comparable to the 2 nm (or 20 angstrom) range.
- Technology:Intel's 18A node features advanced RibbonFET transistor technology and backside power delivery, marking a key step in its manufacturing strategy.
- Timeline:
The 18A node entered risk production and is expected to move to high-volume manufacturing by the second half of 2025
Intel's Relationship with TSMC's 2 nm Process
- Dual-Sourcing:Despite developing its own competitive process, Intel has reportedly placed orders with TSMC, a rival foundry, for its Nova Lake desktop processors.
- Strategic Shift:This dual-sourcing approach signals a strategic effort by Intel to ensure its future processors benefit from the most advanced manufacturing capabilities available on the market.
The 18A-P Process
- Enhanced Process: Intel is also working on an enhanced version, the 18A-P, with mass production targeted for the second half of 2026, potentially further improving its position in the market.
No comments:
Post a Comment