Intel unveils 18A node with major performance gains and efficiency boosts

Intel´s 18A process promises breakthroughs in chip speed, power efficiency, and manufacturing simplicity, targeting mass production in 2025 and powering next-gen Artificial Intelligence and cloud workloads.

Intel has pulled back the curtain on its forthcoming 18A semiconductor process node at the VLSI Symposium in Japan, revealing advancements that position it as a significant leap over previous technologies. Scheduled for mass production in the latter half of 2025, Intel 18A integrates Gate-All-Around transistors with a PowerVia backside power delivery system, marking a dramatic shift in metal stack architecture design. By delivering power through the die´s rear, the company has freed up valuable front-side real estate, achieving tighter pitches for critical interconnect layers while relaxing constraints higher up. This reconfiguration not only boosts yield but also streamlines the overall manufacturing process.

Performance metrics from standardized tests using an Arm-based core sub-block showed approximately 15 percent higher throughput at equivalent power compared to the current Intel 3 process. If supply voltage is pushed to 1.1 volts, the node achieves up to a 25 percent clock speed bump with no increase in energy consumption. Alternatively, running at about 0.75 volts, chips fabricated on Intel 18A can realize an 18 percent performance uptick, or nearly a 40 percent reduction in power usage, offering remarkable versatility for everything from datacenter to edge deployments.

Underpinning these advantages are advances in physical design: cell heights have shrunk to 180 nanometers for performance-focused builds and 160 nanometers for density-optimized layouts, both improvements over earlier nodes. The number of front-side metal layers drops from as many as 19 in Intel 3 to a range of 11-16 in Intel 18A, complemented by three new rear metal layers essential for PowerVia. The company has further tightened interconnect pitches from 60 to 32 nanometers across primary layers. The adoption of low-numerical-aperture extreme ultraviolet lithography on key lower layers trims mask counts by 44 percent and cuts back complexity and cost. Intel intends to roll out 18A across a low-power ´Panther Lake´ compute chiplet and the forthcoming Clearwater Forest Xeon 7 series, spanning cost-optimized, balanced, and high-performance 17-, 21-, and 22-layer stack variants to meet the needs of diverse market segments, including future Artificial Intelligence hardware and advanced cloud platforms.

74

Impact Score

Rdma for s3-compatible storage accelerates Artificial Intelligence workloads

Rdma for S3-compatible storage uses remote direct memory access to speed S3-API object storage access for Artificial Intelligence workloads, reducing latency, lowering CPU use and improving throughput. Nvidia and multiple storage vendors are integrating client and server libraries to enable faster, portable data access across on premises and cloud environments.

technologies that could help end animal testing

The uk has set timelines to phase out many forms of animal testing while regulators and researchers explore alternatives. The strategy highlights organs on chips, organoids, digital twins and Artificial Intelligence as tools that could reduce or replace animal use.

Nvidia to sell fully integrated Artificial Intelligence servers

A report picked up on Tom’s Hardware and discussed on Hacker News says Nvidia is preparing to sell fully built rack and tray assemblies that include Vera CPUs, Rubin GPUs and integrated cooling, moving beyond supplying only GPUs and components for Artificial Intelligence workloads.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.