Applied Materials enables emerging memories for IoT, AI & cloud

Applied Materials enables emerging memories for IoT, AI & cloud

Technology News |
By Rich Pell

These manufacturing solutions, under the Endura brand, are aimed at accelerating the adoption of these memories in the Internet of Things (IoT) and cloud computing. These allow novel materials – the key to these new memories – to be deposited with atomic-level precision.

MRAM for IoT
MRAM is the leading candidate for embedded applications in the Internet of Things. Deployed at the edge these nodes may combine sensors, local processing and communications and non-volatility is key to low power consumption. MRAM is favoured for storing IoT device software and AI algorithms.

MRAM incorporates delicate magnetic materials commonly found in hard disk drives but which need to be incorporated in conventional semiconductor manufacturing flows. MRAM may eventually be used as an alternative to SRAM in level 3 cache memory or even closer to logic and has the advantage that is can be built in the back-end interconnect layers of a chip sitting above logic, thereby enabling smaller die sizes and lower costs.

However, MRAM memories require precise deposition of at least 30 different layers of material. The Endura Clover MRAM PVD platform includes on-board metrology that measures and monitors thickness of the MRAM layers with sub-angstrom sensitivity. Applied claims it is the industry’s first 300mm wafer capable MRAM system for high-volume manufacturing capable of individually depositing up to five different materials per chamber. It has nine separate chambers.

Next: ReRAM, PCRAM in the cloud
As data generation grows exponentially, cloud data centers require order-of-magnitude improvements in the speed and power consumption ReRAM and phase-change RAM (PCRAM) are non-volatile high-density memories that can be used as storage class memory.

Resistive RAM use novel materials – often metal oxides – to create filamentary connections while PCRAM uses chalcogenide materials in memory cells that can be moved from amorphous to crystalline with a change in resistance. As with 3D NAND memories there is scope to arrange memory in multiple stacked planes and cells can be used to store multiple bits.

ReRAM is also a candidate for in-memory computing architectures whereby computing elements are integrated into the memory arrays to help overcome the data movement bottleneck associated with AI computing.

To address this opportunity Applied is introducing the Endura Impulse PVD platform. “We specify the Applied Materials Endura Impulse PVD system with on-board metrology in our ReRAM technology engagements with memory and logic customers because it enables a breakthrough in these critical metrics,” said George Minassian, CEO and co-founder of

Sung Gon Jin head of the advanced technology thin film group at SK Hynix, said: “In addition to providing continued innovations in DRAM and NAND, SK Hynix is pioneering the development of next-generation memories that can help boost performance and reduce power consumption.

Related articles:

Applied, ARM to develop CeRAM for neuromorphic applications

ARM, Applied, seek to replace SRAM with MRAM

Spin Memory: Using MRAM in place of SRAM

PCM, MRAM will lead in non-volatile memory

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles