Embedded MRAM available on 22nm FDSOI

February 28, 2020 //By Peter Clarke
Embedded MRAM available on 22nm FDSOI
Globalfoundries has announced the availability of production volume manufacturing of embedded MRAM on its 22FDX fully-delepleted silicon-on-insulator process.

The embedded spin-torque transfer (STT) MRAM bit cell has been tuned to withstand high temperatures making it suitable for industrial and automotive applications, Martin Mason, senior director of embedded memory told eeNews Europe. It still provides 10^5 cycles endurance and is effectively intended to be a non-volatile memory replacement for embedded NOR flash code- and data-storage that can scale.

Globalfoundries has one client taped out and several others expected to tape-out in 2020, Mason said. Rival foundry TSMC is also preparing to offer 22nm eMRAM. TSMC had plans to enter so-called "risk production" of both embedded MRAM and embedded ReRAM in chips in 2019 using a 22nm manufacturing process (see TSMC to present 5nm CMOS, 22nm STT-MRAM at IEDM).

Globalfoundries has been making magnetic memory for a while at both the 40nm and 28nm CMOS nodes, but this was mainly as foundry supply agreement to MRAM pioneer Everspin (see Everspin ships 1Gbit STT-MRAM). One or two other companies have been given access including French startup eVaderis, which developed an MRAM-based microcontroller at 40nm (see Startup tapes out MRAM-based MCU).

Today’s announcement demonstrates the scalability of eMRAM at advanced process nodes for Internet of Things (IoT), general-purpose microcontrollers, automotive, edge-AI (Artificial Intelligence), and other low-power applications, Globalfoundries said.
Designed as a replacement for high-volume embedded NOR flash (eFlash), GF’s eMRAM allows designers to extend their existing IoT and microcontroller unit architectures to access the power and density benefits of technology nodes below 28nm.

Next: Robust

Vous êtes certain ?

Si vous désactivez les cookies, vous ne pouvez plus naviguer sur le site.

Vous allez être rediriger vers Google.