ARM Architecture

Weebit Nano strikes nearer to productisation

Demo chips efficiently full performance testing; scaling embedded ReRAM all the way down to 22nm; expertise switch to SkyWater on-track

April 29, 2022 – Weebit Nano Ltd (ASX: WBT, Weebit or Firm) is happy to supply this exercise report for the quarter ended 31 March 2022 (Q3 FY22), together with the Firm’s Appendix 4C money move report.

Commenting on Q3 FY22, Weebit CEO Coby Hanoch stated:

“Weebit Nano continues to ship in opposition to its commercialisation technique, reaching an essential technical milestone on its productisation roadmap. Practical testing of the primary demo chips with our embedded ReRAM reminiscence module confirmed the technical parameters of our expertise, and we at the moment are progressing to characterisation and qualification with SkyWater.

“Our demo chips, designed as a System on a Chip (SoC), present potential clients with bodily samples of our ReRAM reminiscence expertise inside totally operational chips. Whereas this is a vital step in direction of supply of a industrial product, it additionally allows potential clients to start incorporating Weebit’s reminiscence expertise inside their future product designs.

“To fulfill the wants of superior functions similar to AI, 5G and IoT, we’re persevering with to scale our ReRAM to smaller geometries the place embedded flash expertise is now not possible. We have now already commenced work with CEA-Leti on scaling all the way down to 22nm, a course of expertise solely a handful of superior fabs on this planet can assist. On the similar time, we’re persevering with to enhance the production-level parameters of our baseline ReRAM expertise whereas additionally progressing the event of a discrete reminiscence answer.”

Demo chips full purposeful testing

In early April 2022, Weebit moved nearer to delivering a industrial product with the demonstration (demo) chips integrating its embedded ReRAM module efficiently finishing purposeful testing. Weebit obtained the silicon wafers containing the demo chips from manufacturing in December and along with CEA-Leti, sliced the wafers into chips, packaged and extensively examined them.

Testing included set and reset programming and studying of the whole ReRAM array utilizing good algorithms and varied knowledge manipulations. Weebit and CEA-Leti additionally examined the superior options of the reminiscence module and the operation of the whole subsystem, confirming the demo chips carry out as anticipated.

Clients can use these demo chips to start designing Weebit’s ReRAM into their SoCs. The demo chips, manufactured at Leti, have now moved to the following step– characterisation and qualification.

Cutting down embedded ReRAM to 22nm

Through the quarter, Weebit commenced scaling its embedded ReRAM expertise all the way down to 22nm – a sophisticated course of node for IoT, AI and 5G functions, which just a few of the world’s main fabs assist. The Firm is working with improvement companion CEA-Leti to design a full IP reminiscence module integrating a multi-megabit ReRAM block focusing on a sophisticated 22nm Absolutely Depleted Silicon On Insulator (FD-SOI) course of.

This new 22nm program follows the achievement of production-level parameters at 28nm on 300mm wafers final October. Weebit is leveraging its robust steadiness sheet to speed up improvement plans to scale its expertise to smaller geometries the place current embedded flash expertise is now not viable, enabling future clients to create new merchandise in superior geometries.

Know-how switch to SkyWater’s US fab on observe

Weebit’s switch of its embedded ReRAM expertise to SkyWater’s US manufacturing fab is on observe to tape-out the primary demo ReRAM chips by mid-year. Whereas Weebit’s expertise makes use of trade customary supplies, the semiconductor manufacturing course of is complicated, requiring adaptation to the particular instruments utilized by every fab.

Following tape-out of the demo chips, expertise qualification will start, which entails in depth testing of producing processes to make sure the repeatability of Weebit’s embedded ReRAM at scale. SkyWater will then add Weebit’s ReRAM to its 130nm Course of Design Package, enabling SkyWater clients to license Weebit’s expertise and embed it inside new product designs they develop and mass-produce at SkyWater’s fab.

Demonstrated Weebit’s first crossbar ReRAM arrays

In January, Weebit achieved a key technical milestone with CEA-Leti within the improvement of discrete (stand-alone) NVM chips, demonstrating its first operational Kilobit (Kb) crossbar arrays that mix Weebit’s ReRAM expertise with a selector. Weebit’s crossbar arrays have been developed utilizing 1S1R (one selector one resistor) structure, which in future will allow Weebit’s arrays to be stacked in 3D layers to ship even greater densities.

Weebit’s 1S1R crossbar ReRAM structure has the potential for use as storage class reminiscence, persistent reminiscence and as a NOR flash alternative, and can also be excellent for Machine Studying in-memory computations and neuromorphic computing (Synthetic Intelligence).

Evaluations with potential companions and clients

Technical evaluations and enterprise discussions with potential clients and companions are ongoing. The purposeful demo chips allow Weebit to reveal its embedded ReRAM to potential clients, supporting adoption previous to qualification by enabling potential clients to check and consider the expertise inside their functions and designs.

Abstract of Q3 FY22 money flows

In Q3 FY22, Weebit Nano obtained $8.5 million from the exercising of choices. An additional $300K was raised through the issuing of Placement shares to 3 administrators, following shareholder approval as a part of final November’s capital elevating. R&D bills have been offset by a GST refund in France, totaling $1.9 million, leading to a money influx from R&D of $759K. Funds to associated events over Q3 FY22 have been $752K, which included charges paid to administrators and the CEO’s value of payroll and bonus for the interval.

Trade occasions

Over This fall FY22, Weebit will take part at prestigious trade occasions, together with:

  • Might: Weebit’s Chief Scientist Gabriel Molas will current on the Worldwide Reminiscence Workshop (IMW) 2022 on ReRAM’s excessive temperature reliability. On the similar occasion, Weebit’s CTO Ishai Naveh will take part in a panel session.
  • June: Weebit’s VP of Advertising & Enterprise Growth Eran Briman will current on the New Paradigms for Computing Workshop at CEA-Leti’s flagship Leti Innovation Days occasion on the state of ReRAM.
  • June: Weebit’s VP Know-how Growth Amir Regev will focus on rising supplies and applied sciences for ReRAM at CIMTEC in Italy.

Particulars about these and different invited papers and occasions are listed on Weebit’s web site at

Wanting ahead

Weebit stays targeted on commercialising its embedded ReRAM reminiscence expertise, and is on observe to realize the next milestones by the top of CY22:

  • Know-how switch of its expertise to SkyWater’s US fab and being in superior phases of qualification;
  • Demonstrating the performance and efficiency of its embedded ReRAM module;
  • Signing new licensing agreements with companions and clients; and,
  • Proceed scaling its embedded ReRAM to 22nm.

In parallel, Weebit continues to work with CEA-Leti on the event of an answer for the discrete reminiscence market.

Quarterly money move report

To learn the report click on right here

About Weebit Nano Restricted

Weebit Nano Ltd. is a number one developer of next-generation semiconductor reminiscence expertise. The corporate’s ground-breaking Resistive RAM (ReRAM) addresses the rising want for considerably greater efficiency and decrease energy reminiscence options in a spread of latest digital merchandise similar to Web of Issues (IoT) gadgets, smartphones, robotics, autonomous autos, 5G communications and synthetic intelligence. Weebit’s ReRAM permits semiconductor reminiscence parts to be considerably quicker, cheaper, extra dependable and extra power environment friendly than these utilizing current Flash reminiscence options. As a result of it’s based mostly on fab-friendly supplies, the expertise will be rapidly and simply built-in with current flows and processes, with out the necessity for particular tools or giant investments. See:

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button