Weebit Nano

BUY

Date of inclusion: 17 June 2021
Share price on inclusion in Marc & Stuart’s Top Picks: $1.78
52-week range: A$0.20 – A$4.27
Risk Level: Medium

 

Weebit Nano has been on Marc & Stuart’s Top Picks list before, from 10 March 2021 to 25 March 2021 when we made a 28% return on the stock. We are including it again at $1.78 per share in anticipation of first commercial deals, which the company indicated it expects around the middle of 2021.

The next big thing in computer memory storage

In information technology ‘non-volatile’ memory is the memory that stays in place after the power gets cut, e.g. when you turn off your computer. For the last few decades, the technology used for non-volatile memory is called Flash. Weebit Nano (ASX:WBT) aims to replace Flash with its silicon oxide ReRAM memory potentially working 1,000 times faster and using 1,000 less energy, as well as lasting 100 times longer than Flash. Weebit Nano went public on ASX in 2016 in order to fund the development of this technology.

A huge market opportunity

The overall market size for Flash is currently estimated at over US$60bn and the market is growing rapidly because the world’s storage requirements are doubling every two years.

Why Weebit has a chance to hit the big time

Weebit has been able to productise its silicon oxide ReRAM cost-competitively with the standard materials and semiconductor manufacturing equipment used in fabs globally. The company has developed an embedded memory device (where memory is combined with processors and other elements on the same chip). Weebit currently has ongoing discussions with semiconductor companies and expects to close it’s first commercial deal for this so-called memory module by mid-2021.

Additionally, the company expects to have its first standalone memory cell for Storage Class Memory, used in mobile phones, computers and datacenters, ready in the third quarter of 2021. A full array should be available early 2022.

Funded for the next stage of its growth

Weebit Nano raised $15m at $1.70 per share in late 2020 and should be able to raise substantially more capital through the deep-in-the-money options that are outstanding. We believe the company is well-funded through to late 2022.

Why we like Weebit

Flash has been the way electronic device makers do non-volatile memory since the 1980s. However, today’s Flash doesn’t cope well in embedded devices and it can’t go below the all-important ‘28nm’ limit on chips where electrons tunnel through gates they shouldn’t go. The industry is looking for a solution and Weebit has one of the more advanced technologies. The fact that in 2020 TSMC, the world’s largest third-party chip manufacturing company (or foundry), announced it started to offer ReRAM to its customers is a major validation of the technology, in our view.

Favourable valuation

In the current phase of the company’s development, we believe Weebit Nano should be valued around A$750m, which is in line with valuations of past industry M&A deals. That would suggest a share price of around $4.75.

Key risks:

  • Delays in signing commercial deals, i.e. well beyond the middle of 2021, may disappoint the market and could put pressure on the share price.
  • Failure to get the technology validated by more potential customers could limit commercial potential of the technology.
  • Competing technologies may limit commercial uptake of WBT’s SiOx ReRAM.

 

Weebit Nano

 

Update 25 June 2021

 

Reason for update: Weebit Nano announced achievement of crucial technical milestone

Weebit Nano’s Resistive RAM (ReRAM) technology is a new form of non-volatile memory, meaning the values stored in this memory will be retained, even when the power is turned off. Non-volatile memory is widely used in electronic systems, e.g. to store pictures and videos on your mobile phone and laptop as well as storage of all sorts of data in data centers. This type of memory is known as Storage Class Memory, standalone memory or discrete memory.

But non-volatile memory is also used as an embedded solution, i.e. in combination with complete subsystems as the on-board memory component that can be used by that system. Weebit Nano is getting close to signing its first commercial agreement for this embedded application.

The announcement on 25 June 2021, that sent the share price soaring 23%, is related to the standalone memory solution.

Successful integration of selector with ReRAM cell for use in standalone memory

Without wanting to get too technical, a selector is used to address individual cells in a memory array for standalone memory. In embedded memory, where memory cell size is less of an issue, this is done with transistors, which are relatively large. But standalone memory needs to shrink down to at least 28nm, which is the current resolution of Flash memory cells. And over time Weebit Nano will want to scale its ReRAM memory arrays to much smaller resolutions, e.g. down to 20nm and lower. In order to do this, a selector is required.

The importance of this integration can’t be overstated

The company has now successfully integrated a selector with a SiOx ReRAM cell, an industry first! The importance of this can’t be overstated. This integration means the company can now move forward and start to build discrete memory arrays; initially kilobit arrays and then megabit arrays and ultimately gigabit arrays. The latter will be required to address the US$100BN Storage Class Memory market, which today mostly comprises NAND Flash memory, a technology which really can’t scale any further.

We’ve been big fans of Weebit Nano ever since it listed on the ASX. The recent announcement confirms the company is on the right track and we’re happy holders of WBT in Marc & Stuart’s Top Picks.

 

Update 4 October 2021

 

Reason for update: WBT successfully scales its technology down to 28 nanometers

On Friday 1 October, Weebit Nano announced that it had successfully scaled down (not up) its ReRAM technology to a resolution of 28 nanometres (nm). Together with French development partner CEA-Leti it manufactured so-called megabit arrays at 28nm on 300 millimetre (mm) wafers.

Let’s break all that down.

28nm resolution puts ReRAM at a major advantage versus Flash Memory

Resolution is the term for the linewidth of semiconductors’ circuitry. Remember, smaller is better here, because it means the chip will be faster and use less energy. Flash memory is currently manufactured at 40nm and can’t really scale down below that level, because the individual memory cells with start to display, what is known as, cross talk, i.e. the electrical charge will leak from one cell to its neighbouring cells, which means the value of the cells (1 or 0) will slowly disappear.

Weebit being able to manufacture fully functional Non-volatile memory (NVM) at 28nm is a very significant achievement, which we believe will turn a lot of heads in the semiconductor industry.

Megabit arrays are the last stop before gigabit arrays

As we explain this point, please keep in mind that WBT is initially targeting embedded memory applications, i.e. where a relatively small-sized memory module is part of the larger chipset (it is embedded). And for that purpose, we believe megabit arrays are fine.

However, the largest chunk of the Non-volatile memory market is Flash memory, used in mobile phones, computers, iPads etc. in large sizes, e.g. 256GB. We believe a longer-term play for WBT will be this so-called Storage Class Memory (SCM) segment of the market. Specifically, SCM sits between DRAM and Flash. When it comes to speed, ReRAM is close to DRAM (which is really fast). But ReRAM is non-volatile, like Flash. So, ReRAM is the best of both worlds.

The fact that WBT has now achieved megabit arrays at 28nm tells us that gigabit arrays aren’t too far off (theoretically). In turn, that would mean that the company can start to address the SCM market, which we expect will partially replace Flash memory in due course. Granted, this application is further out than what the company is working on with SkyWater right now, but it is potentially a lot bigger opportunity too.

300mm wafer sizes are today’s norm

Lastly, all of the above has been achieved on semiconductor wafers that measure 30 centimetres, or 300mm in size. When it comes to wafer size, bigger is better as you can cram more chips on a 300mm wafer than on a 200mm wafer. That leads to a 30% cost reduction per chip when you account for the more expensive semiconductor manufacturing tools needed for 300mm wafers.

Additionally, WBT’s manufacturing process doesn’t require new or additional manufacturing machines. So, the ReRAM design can be integrated into a customer’s existing chip design and move straight to the factory floor (after the proper testing and qualification obviously).

The market still doesn’t seem to get it

Looking at the share price response to Friday’s announcement and the company’s fully diluted market cap of $470m, we believe the market still doesn’t understand WBT’s potential. If it did, the valuation would be a lot higher, in our view.

Which is why WBT is one of Marc & Stuart’s Top Picks. Over time, we expect customers will come on board in addition to SkyWater Technology. And WBT will expend beyond embedded memory, into SCM and potentially into neuromorphic processing, which is another beast altogether.

For now, we’re happy and patient holders of the stock.

 

Update 22 December 2021

 

Reason for update: WBT announced receipt of the first wafers with ReRAM embedded into full subsystems on demonstration chips

Another big milestone achieved 

This morning Weebit issued a press release to tell the market about the first wafers it received from development partner Leti that integrate the company’s ReRAM technology into entire subsystems. WBT’s embedded ReRAM module can sit inside the chip of a potential customer to form a so-called system on a chip that contains all required elements, like SRAM, I/O interfaces, a microcontroller etc.

The individual chips will now be cut out of the wafers to be packaged and connected. They will be used by SkyWater, WBT’s first commercial customer, for testing and characterisation. But the test chips can also be used by prospective customers, i.e. to test and see if the technology can be used in their applications. In our view, this is another big step in the commercialisation of the company’s ReRAM technology.

Today’s announcement is a major milestone for WBT that, for some reason, wasn’t allowed to be announced on the ASX Platform. Luckily, you have Stocks Down Under to alert you to these things 😉

 

Disclosure: Stocks Down Under staff and/or director(s) own Weebit Nano shares.

 

Read the most recent article on WBT here