Imec advances ferroelectric memory for AI workloads

Imec is testing ferroelectric capacitors and vertically stacked FeFETs as possible routes to denser, lower-voltage memory for data-intensive systems. The work targets scaling limits in DRAM and SRAM as AI workloads increase memory demands.

Imec presented two ferroelectric memory advances at the 2026 IEEE / JSAP symposium on VLSI Technology & Circuits, positioning the technology as a response to growing pressure on memory systems from AI workloads. The research targets ferroelectric capacitors and ferroelectric field-effect transistors as candidates for lower-voltage operation and denser integration as DRAM and SRAM become harder to scale.

The capacitor work demonstrates low-voltage (~1.3 V) operation through ferroelectric layer scaling while maintaining high remnant polarization (>40 μC/cm²) and endurance (≥10¹³ cycles), requirements for DRAM-like memory. A separate FeFET demonstration uses vertically stacked IGZO-based devices and shows the first functional five-word-line vertical stack of FeFET memory cells, with a dual-gate configuration and back-gate designed to improve erase efficiency.

Imec said the two approaches share material stacks and integration strategies, allowing insights from capacitor scaling and interfacial engineering to inform FeFET optimization, while 3D stacking work could support future ferroelectric capacitor arrays. Further work will address FeFET endurance and erase performance, capacitor voltage scaling and reliability, system-level evaluation, and fully integrated 3D memory architectures.

52

Impact Score

SpaceX agrees $60bn deal to buy Cursor

The all-stock acquisition would bring Anysphere’s AI coding platform into SpaceX as Elon Musk’s companies deepen their enterprise AI push. Cursor counts Stripe, Adobe and Nvidia among reported corporate users.

ASML warns Terafab hinges on EUV supply

Christophe Fouquet has held direct talks with Elon Musk about Terafab, but ASML’s bigger concern is whether it can meet another wave of lithography demand. The proposed Texas fab could intensify competition for the machines behind advanced chip production.

Epic opens Unreal Engine to competing AI models

An experimental MCP plugin in Unreal Engine 5.8 gives LLMs a path into real project data and engine systems. The move sets up a workflow contest among model providers rather than a single default winner.

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.