BrainChip Holdings Ltd (BRN) Earnings Call Transcript & Summary

Good morning, good afternoon and good evening to everyone, wherever you're joining us from. Welcome to the Annual General Meeting of BrainChip Holdings Ltd. My name is Manny Hernandez, and I have the pleasure of being your Chair of the company. I understand that a quorum is present, and I'm happy to formally open the meeting. Let me continue. First of all, I wish to thank everyone attending in person and connected into the Lumi online platform for joining us today. The directors and management would like to conduct this AGM in person. We're really disappointed that the continuing -- excuse me, Peter. The directors and management would have preferred to conduct this AGM in person, and we're obviously extremely disappointed that the continuing risks associated with the COVID-19 restrictions in both our jurisdictions, prevented us all from being with you in person for the second time. We're hoping that the next AGM, we will be back to normal. The Lumi platform allows shareholders, proxyholders and guests to attend the meeting virtually. All attendees can watch a live webcast of the meeting. [Operator Instructions] Please note that while you can submit questions from now on, I will not address them until the relevant time in the meeting. Please also note that your questions may be moderated. Or if we receive multiple questions on 1 topic, amalgamated together. [Operator Instructions] Finally, due to time constraints, we may run out of time to answer all your questions. If this happens, we will answer them in due course via e-mail or posting our responses on our website. In the event that I am disconnected from this meeting and cannot immediately rejoin due to technology failures, Mr. Geoff Carrick, who is present in the Perth venue will assume the role of Chair of this meeting. Before proceeding with the business of the meeting, I would like to mention some procedural matters. For those of you attending our Perth venue, you should have registered your attendance as you entered the room today. If any member or proxy holder has not registered their attendance to the door -- at the door with our share registrar Boardroom, would they please do that now. Staff from Boardroom are there to assist you. Visitors are also registered electronically on Boardroom's meeting registration system. If any visitor has not registered their attendance, could they please do that now. Also, could you all switch all mobile phones and any recording devices at this time. Voting today will be conducted by way of a poll on all items of business. In order to provide you with enough time to vote, I will shortly open voting for all resolutions. Those shareholders attending our Perth venue will have received a voting card on registering your attendance which I will ask you to complete as each resolution is proposed. Stacey Spence of Boardroom has been appointed as Returning Officer and will collect all votes at the end of the meeting. For those attending this meeting and voting on the Lumi platform, if you are eligible to vote at this meeting, the new polling icon will appear. Selecting this icon will bring up a list of resolutions and present you voting options. To cast your vote, simply select one of the options. There is no need to hit the submit or enter button as the vote is automatically recorded. You may, however -- you do, however, have the ability to change your vote up until the time that I declare voting closed. Now let me introduce each of the directors present in person or via Lumi Technology today. Peter Van der Made, our Interim CEO, Co-Founder and Executive Director; and Geoff Carrick, Non-Executive Director, are both at the Perth venue. I would also like to introduce Kim Clark, the Company's Secretary of BrainChip Holdings; and Ken Scarince, our Chief Financial Officer. In line with the Corporations Act, our auditors, Ernst & Young, are present and are represented by Jemma Newton, partner with the firm. Ms. Newton is available to take questions on the audit report later in the meeting. I propose to hold the formal part of the meeting first, followed by a presentation from management. Once we have completed these matters, then I will close the meeting. Before the formal part of the meeting, I'd like to say a few words. Let me just start by expressing management and the Board's gratitude for your continued support and patience as we strive to bring Akida to full commercialization. As you know, 2020 was highly impacted by the COVID-19 pandemic. The safety of our employees and the safe interaction with our support service providers and prospective customers were of utmost importance and we're glad to report that we managed reasonably well through this challenging period. 2020, as we noted at the prior AGM, was to demonstrate the feasibility and functionality of the Akida chip, which we successfully completed through a multi-project wafer shuttle with the assistance of Socionext and TSMC, and that happened in the second quarter of 2020. We then evaluated Akida and commenced shipping evaluation boards through our Early Access Program or EAP partners in the fourth quarter. Having proven the functionality of the design, we then focused our resources on the production version of Akida which we successfully handed over to Socionext in the first quarter of 2021 and is now with TSMC for production fabrication. While the Akida chip dominated our focus in 2020, we also continued to promote BrainChip's IP, which landed as our first intellectual property license agreement with a leading Japanese semiconductor company in the fourth quarter of 2020. Though we've made some advances in the fabrication of Akida and in licensing our technology, my point last year about BrainChip being a pioneering startup in the field of artificial intelligence, while being a publicly listed company in the ASX has not radically changed. We are still introducing groundbreaking, highly differentiated AI technology in a relatively nascent market, where even some of our potential customers are themselves still evaluating how to best approach this huge emerging market. As Peter will address more later in his company update, we stand committed to make BrainChip a success. We are expanding our focus to not only harness potential big opportunities with our EAP partners like Ford, Valeo and NASA, but also to create recurring revenue streams. Now let me address certain remuneration matters and recent departures in the company. Though we appreciate certain stockholder sentiment over our remuneration practices, I would like to emphasize that the Remuneration Committee and the Board remains very cognizant of these matters, but also have to maintain competitiveness in the markets in which we operate and from which we recruit from. In 2020, we did not increase the base compensation of our KMPs with the exception of our Chief Development Officer in recognition of his and his team's advancement of the Akida chip. We did not grant incremental equity to our KMPs, except for a customary and competitive refresh grant for our CFO for retention purposes. We did hire our VP of Sales and Marketing in 2020. One of the resolutions for this AGM, as noted in our Notice of Meeting, was a grant of options to one of our Non-Executive Director, Geoff Carrick, who joined us late 2020. The initial director grant of equity was reasonable and aligns our interest with that of our stockholders. Ultimately, we will collectively share the success of BrainChip. And when you realize the return on your investments in the company, then only then so will we. As we previously announced, Mr. DiNardo and BrainChip had a mutual agreement for Mr. DiNardo to step down as our CEO. Though we appreciate Mr. DiNardo's contributions to the company. This also gives BrainChip an opportunity to attract a CEO that is more appropriate for our stage and commercialization phase. This is no easy task given our objective to find a CEO that will take us to the next level. For the first couple of months since Mr. DiNardo's departure, management and the Board focused on tapping our individual networks for potential candidates. We are pursuing certain people on our list and in parallel, have also evaluated search firms that we might engage after exhausting our own network of candidates. Sadly, we lost another director just recently. Ms. Steele, for urgent personal reasons, resigned from the Board. We thank Ms. Steele for her 8 months of service and wish her and her family the best. Peter, our Interim CEO and Executive Director; Geoff Carrick, our Australian-based Non-Executive Director; and I, are all committed to bringing BrainChip our next-stage CEO and recruit directors that will enhance the governance and oversight of our company. Thank you for allowing me to share a few words. Now let's proceed with the formal part of the meeting. I now declare voting open to all items of business. The polling icon will soon appear. Please submit your votes at any time. I will give you a warning before I move to close the voting. We will now proceed with a discussion of the proposed resolutions including questions from shareholders about each resolution. The Notice of Meeting was mailed to all shareholders on 22 April 2021. I propose that the Notice of Meeting be taken as read. With the recent resignation of Ms. Steele as Director of the company, Resolutions 3 and 8a will not be put to shareholders for consideration. We will now proceed with the business of the meeting. Item 1 on the agenda deals with the receipt and consideration of BrainChip Holdings' Financial Report, the Director's Report and the Auditor's Report for the year ended 31 December 2020, which are incorporated in the 2020 Annual Report sent to all shareholders who have requested the report. I would encourage any shareholder who has a question on these reports, our business in general, or any questions of our auditor to raise them now. Since there are no questions, I would now like to move to consider the formal motions of the meeting. Resolution 1, remuneration report. I refer you to Resolution 1 of the Notice of Meeting in respect to the approval of the remuneration report of the company. This is a nonbinding resolution. The company's remuneration report is contained in the 2020 Annual Report. I would highlight that in accordance with the Corporations Act, no votes may be cast on this resolution by or on behalf of a member of the company's key management personnel or their closely related parties. I will refer to these people collectively from now on as prohibited voters. A prohibited voter may, however, vote directed proxies where they do so for another person who is not themself a prohibited voter. As Chairman, I may also vote undirected proxies for a person that is not prohibited voter in accordance with my stated voting intentions to vote all available proxies in favor of this resolution. Before opening this to discussion, I would like to mention that in the interest of corporate governance, the BrainChip Holdings Board has abstained from making any recommendation in relation to this resolution. Is there any discussion on this resolution? Okay. Hearing none, is there -- I propose the resolution, that for the purpose of Section 250(R)2 of the Corporations Act and for all other purposes, approval is given by the stockholders for the adoption of the remuneration report as contained in the company's Annual Report for the year ended 31 December 2020. Please submit your votes or mark your voting card in respect to this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including as Chairman who vote in accordance with my stated intentions in favor of these resolutions, were 118.2 million votes for and 102.2 million votes against, with 11.4 million open and 101.9 million votes abstained. Let's move on to Resolution #2, reelection of Emmanuel Hernandez as Director. I refer you to Resolution 2 of the Notice of Meeting in respect of the reelection of myself as a Director of the company. I will hand the meeting over to Geoff Carrick to chair for the purpose of this resolution. Geoff?

Thank you, Manny. Manny Hernandez retires in accordance with Clause 15.4 of the company's constitution. His biography is included within the explanation memorandum accompanying the Notice of Meeting. Is there any discussion on this resolution? If there's no discussion, I will propose the resolution that for the purposes of Clause 15.4 of the constitution and ASX Listing Rule 14.4 and for all other purposes, Emmanuel Hernandez, who retires by rotation and being eligible offers himself for reelection, be reelected as a Director of the company. Please submit your vote or mark your voting card in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including Mr. Hernandez as Chairman to vote in accordance with his stated intentions in favor of this resolution. Manny, you'll have to help me with that because I don't have that information.

Thank you, Geoff. The results are displayed on the screen, but just to share, there were 506.9 million votes in favor and 12.3 million votes against. Thank you, Geoff. Moving on to resolution 4, the election of Geoff Carrick. I refer you to Resolution 4 of the Notice of Meeting in respect to the reelection of Geoff Carrick as a Director of the company. Geoff was appointed as a Director of the company on 23 November 2020 and retires in accordance with Clause 15.6 of the company's constitution. Geoff's biography is included within the explanatory memoranda accompanying the Notice of Meeting. The BrainChip Holdings Board, with Mr. Carrick abstaining, unanimously recommends that you vote in favor of this resolution. Is there any discussion on this resolution? Hearing none, I now propose the resolution that for the purposes of Clause 15.6 of the constitution and ASX Listing Rule 14.4 and for all other purposes, Geoffrey Carrick, having been appointed by the Board as a director to fill a casual vacancy until the next Annual General Meeting, retires and being eligible, be elected as a Director of the company. Please submit your vote or mark your voting cards in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including me as Chairman to vote in accordance with my stated intentions in favor of this resolution, are now displayed on the screen. There were 516.4 million votes in favor and 2.9 million votes against. Resolution 5, ratification of the prior issue of 35 million shares. I refer you to Resolution 5 of the Notice of Meeting in respect of the ratification of the prior issue of 35 million shares on the terms contained within the explanatory memorandum accompanying the Notice of Meeting, including the voting exclusions applicable to this resolution. BrainChip Holdings Board unanimously recommend that you vote in favor of this resolution. Is there any discussion on this resolution?

Yes, Mr. Chairman, there are a number of questions that have been submitted online by a shareholder, MISRA Proprietary Limited. They all relate to the LDA put option agreement. I'll now read them out for you. The shareholders says, "Well done on the operational and development side of the business, but I just have a couple of questions on the LDA put options under the general heading of Transparency and Disclosure. Firstly, what was the put option price advised to LDA in August/September last year when the company exercised the $10 million capital raise, and why was this not disclosed at the time? And will it be disclosed in the future if further exercises are made? There are 4 more questions, but you might wish to respond to that one first.

Thank you, Kim. And our CFO, Mr. Scarince, is also available to elaborate if I fail to explain this enough. The arrangement with LDA with regard to put option price was by agreement subject to a pricing window by which there's a weighted average result of what the discount or the effective price would be. Consequently, you cannot announce what that price would be at the time that we put the notice on LDA. So that's just the pricing mechanism of the LDA agreement. The fact that we have to wait for the pricing window to develop, at which time the proper pricing and appropriate discount for LDA is now -- is then determined. Ken, would you like to add anything to that answer?

No, Manny, you captured it perfectly.

Thank you. Kim, any follow-up on that matter, please?

Certainly. The next question being, after the raising was completed, why was not the new issue price of the new shares advised to the market? It appears from the annual report that it was around $0.41 per share. In future, when a put option is exercised and shares are issued, will you disclose the final issue price quickly after the 30-day VWAP is ascertained? This is an important figure for market participants and shareholders to know.

That's actually a very good point. So why don't we take that as a recommendation or a suggestion from our stockholders. Assuming we're not violating any particular clause or agreement of the contract, I see no reason why we cannot do that. So we'll follow up on that. In the event there's another capital raise subject to a pricing window, it might be a good idea to disclose to the market what the eventual or final pricing was, assuming further that we are not violating any particular conditions on the agreement.

A further question, if the put option is exercised again, will the remaining collateral shares held by LDA also be issued at around $0.41? Or will the pricing process start again?

With respect to future capital put notices, which is what we call it when we advise LDA that we're going to partake or use the facility, it is typically subject to a pricing window. So we cannot tell you at this point what that capital pricing might be. But it will be subject to a pricing window at that point. And subject to the previous discussion, we might disclose to you the final pricing of the next capital raise.

Next question being, do LDA currently hold any significant shares in the company? Or have they exercised their [ put ] options and sold out?

With respect to LDA holding any further options, the answer is no. They have fully executed and exercised their options. As to whether they're holding any further shares of the company, yes, they do. We currently have 8.7 million shares that are parked, if you will, with LDA. In the event that BrainChip decides to do a capital raise, they are already holding 8.75 million of our shares. And we fully intend at some point to do that.

And the final question on this resolution, can you explain in layman's terms what this complicated accounting of derivative and option scheme means to the company? Some $15 million loss has been recorded, which is hard to reconcile against a successful capital raising. Does it represent the loss of opportunity of issuing shares and options at low prices compared to the current and past higher levels? What is the ongoing implication of all of this? Do we need to have the price pull to regain the losses? This logic seems a bit odd and conflicting. Do we need to worry about this as we go forward? And will it cloud and affect your reported results?

Thank you for the question, pardon me for chuckling a little bit because it is actually a very complicated matter indeed. I guess my short answer, particularly to the last point of the question, is no. You shouldn't worry about it. It happens to be a very technical accounting mandate in terms of how these transactions are accounted for. Importantly, we cannot categorize them as a noncash impact, meaning we don't necessarily write a $15 million check to record that transaction as reported in our books. It's an adjustment in value. It's noncash, but we have to do it the way the pronouncements as far as the accounting goes. So I think the best message I want to leave you for whoever asked these questions, you shouldn't worry about it. Any further follow-ups on that, Kim?

No, that's the final set of questions online, unless there's any from the floor, Manny.

Thank you so much for that, by the way. LDA was a very good transaction for us, and there continues to be an active line of capital that we could draw from if it's needed. If we consider it to have been a successful agreement so far, they've also been a very good company as far as supporting our vision. So I appreciate the question. And hopefully, my answers gave a little bit of clarity to the rest of the stockholders as to those matters. So having further no discussion, I now propose the resolution that for the purpose of Listing Rule 7.4 and for all other purposes, approval is given for the ratification of the prior issue of 35 million shares fully paid ordinary shares which are issued pursuant to the put option agreement between the company, LDA Capital Limited and LDA Capital, LLC as announced to ASX on 13 August 2020 and on the terms and conditions set forth in the explanatory memorandum. Please submit your vote and mark your voting card in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including me as Chairman to vote in accordance with my stated intentions in favor of this resolution, are now displayed on your screen. Just to recap, there were 428.4 million shares in favor and 84.7 million shares against. Moving on to Resolution 6, which is also related to the LDA agreement. This is for the ratification of the prior issue of 75 million options. I refer you to Resolution 5 -- or 6 of the Notice of Meeting in respect of the ratification of the prior issue of 75 million options on terms contained within the explanatory memorandum accompanying the Notice of Meeting, including the voting exclusions applicable to this resolution. The BrainChip Holdings Board unanimously recommend that you vote in favor of this resolution. Is there any discussion on this resolution? I think we covered the option question previously, so I'll just move on. I now propose the resolution that for the purpose of Listing Rule 7.4 and for all other purposes, approval is given to the ratification of the prior issue of 75 million options over fully paid ordinary shares, which were issued pursuant to the put option agreement between the company, LDA Capital Limited and LDA Capital LLC as announced to ASX on 13 August 2020 and on the terms and conditions set out in the explanatory memorandum. Please submit your vote and mark your voting cards in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and that the proxy's discretion, including me as Chairman, to vote in accordance with my stated intentions in favor of this resolution, are now displayed on the screen. 427.9 million in favor and 85.3 million votes against. Resolution 7, approval of the equity incentive plan. I refer you to Resolution 7 of the Notice of Meeting in respect of the approval of the equity incentive plan on the terms as contained within the explanatory memorandum accompanying the Notice of Meeting, including the voting exclusions applicable to this resolution. The BrainChip Holdings Board, in the interest of good governance, abstains from making a recommendation in respect of this resolution. Is there any discussion on this resolution? Hearing none, I now propose the resolution that for the purpose of Listing Rule 7.2, Exception 13 and for all other purposes, the company's equity incentive plan, as described in the explanatory memorandum, be approved for the issue of securities under the company's equity incentive plan. Please submit your vote and mark your voting card in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including me as Chairman to vote in accordance with my stated intentions in favor of this resolution, are now displayed on your screen where approximately 136.6 million votes in favor and 109 million votes against. Resolution B, approval of issue of option to directors. Resolution 8(b) of the Notice of Meeting relates to the approval of the issue of option to Director Geoffrey Carrick, on the terms and conditions as contained with the explanatory memorandum accompanying the Notice of Meeting in relation to this resolution. As I've mentioned previously, Resolution 8(a) will not be put to the shareholders. BrainChip Holdings Board unanimously recommend that you vote in favor of this resolution. Is there any discussion on this resolution? Hearing none, I now propose Resolution 8(b) that pursuant to Section 208.1.A of the Corporations Act and Listing Rule 10.14, the members of the company approved the granting of 2.5 million options to Geoffrey Carrick, Non-Executive Director, under the company's equity incentive plan and terms outlined in the explanatory memorandum. Please submit your votes or mark your voting card in respect of this resolution. [Voting]

Votes and proxies received for and against this resolution and at the proxy's discretion, including me as Chairman to vote in accordance with the stated intention in favor of these resolutions, are now displayed on your screen. There were approximately 422.2 million votes in favor and 93.8 million votes against. Ladies and gentlemen, that concludes the formal part of our meeting and our discussion of the items of business. I will now ask our CEO, Peter Van der Made, to give an update on the company. Peter?

Yes. Thank you, Manny. We need to switch over the screen to the presentation. And I'm not sure how we're going to be doing that because the way it is set up is that you are still on the screen, Manny. So maybe you can mute yourself?

Already muted, Peter.

Yes, this might be an issue with the Lumi app because the Lumi app does not allow me to change screens to -- here we go. So ticking down to Slide 2, please. And part of that is being obscured by the -- sorry, that's better. Good. I need to sit down, otherwise, I will not be in the display of the camera. So this is our standard disclaimer. BrainChip is entering an exciting new era as we move from an R&D focus to a manufacturing and sales focus. We have an exciting product, a product that I believe is entering the market at the right time with the right features that enables clients to design Artificial Intelligence into products that are powered from a small battery, wearable electronics or remote sensing equipment that does not require an Internet connection to operate. Information is processed right next to the sensor. This is what we call the edge as opposed to the cloud. No Internet connection is required, and performance is not compromised by transmission delays or bandwidth problems and security is enhanced because nothing needs to be transmitted. We're starting here with our standard disclaimer. We use forecast information about market size and applications from third parties. Just like weather forecasts, a lot of effort goes into getting these forecasts right, but they cannot be right all the time, just like the weather forecast is not always right, even though they have millions of dollars of satellites available to them. And Slide 3, please. [Presentation]

You may have seen this video already. It mentions neuromorphic event-based processing. That's quite a mouthful. What does it mean? Neuromorphic means that the chip is processing information in a similar way that the human brain is processing information. Our brains do not run computer programs to function. The brain learns everything it learns when it comes across images, sounds and other sensors. The brain learns instantly. We do not need 1,000 images of a dog to recognize that it's a dog. A child will see a dog once and is told that it's a dog and then in the future recognizes every dog. The circuits inside Akida have a similar function to the brain cells in the human brain. These circuits are processing incoming information in parallel all at the same time. There is no program code to make them do this. They are designed to imitate the processing method of the cells in the brain including the learning function. Neuromorphic literally means like the brain. Event-based means that only processing -- that it only process the things that are important. The brain does this as well. Only when something changes is it necessary to process something. For instance, if you have a camera pointing at a room that is empty, then the image only needs to be processed once to determine that it is empty. Then no further processing is needed until somebody walks into that room. When somebody walks in the room, then you have -- then the Akida can decide if the person is allowed to be there or not. So only processing takes place when something changes. This saves a lot of power. Slide 4, we look at technology trends. Artificial Intelligence is a fast-evolving market. It's a very young market. Some of the applications we have come across, so far, are remote controllers, financial forecasting, cameras, food safety, food quality, taste control, intelligent refrigerators, disease detection in breath and also disease detection in blood samples. We expect there will be many, many, many more. Artificial intelligence is a technology trend that we will see used in many everyday products. Manufacturers who neglect to incorporate AI into their products in the future will fall behind and expect to lose market share to other people who have incorporated Artificial Intelligence in their products. I'll just give you a few examples of how beneficial AI can improve the quality of our lives. One example is the refrigerator that will be equipped with Akida that could smell if anything is going off in the fridge and tell you what it's likely to be. This would avoid cases of food poisoning. According to Google, each year, 48 million people, 1 in every 7 in the United States, get sick from eating spoiled food. Of those, 128,000 are hospitalized and 3,000 die every year. I did a search on food poisoning statistics. The display on the fridge could give you an overview of what food is in the fridge, how long it has been there, how long it will stay fresh. And doorbells could recognize a person at the door and alert the occupants. Factory robots, in the second image here, autonomous machines. Factory robots are programmed, which is a very slow and tedious process. In the future, they will be equipped with Akida and [Technical Difficulty] benefits to the continuous and incremental learning in Akida, something that only Akida can offer.

Peter, we can't hear you. [Audio Gap] Your car will be able to recognize faces, making for a more personalized experience, for instance, by setting the radio to your favorite station. The same Akida chip can be used to see if the driver is paying attention. 20% of all fatal road accidents involve driver fatigue. Akida can also monitor bearing vibrations and engine noise for safety and preventive maintenance. In medicine, Akida can smell and classify a variety of diseases using the right sensors. We announced our cooperation with NaNose who built a room temperature sensor for the compounds that are present in exhaled breath. Using test data sets that were collected in clinical settings, the Akida chip actually detected COVID-19 with 93% accuracy, much better fast -- and faster than a PCR test. We have also signed up Biotome here in Perth who are working on a diagnostic blood test that will incorporate Akida as well. So over the next 5 years, we expect to see a proliferation of Akida into all kinds of beneficial AI applications that will positively improve all aspects of our lives. Let's move on to the next slide. BrainChip offers several avenues to enable and support the explosion of AI products that we expect to see over the next 5 years. MetaTF, which is our new name for our Akida development engine, is a software tool that works with a Google TensorFlow environment. MetaTF is the development tool that is common to all BrainChip products. We have chosen to use TensorFlow because most data scientists are already very familiar with it. There is no new programming language to learn and data scientists can dive right in. The BrainChip Akida IP is offered as an encrypted box that is dropped into a manufacturer's chip design. It has been configured by BrainChip engineers for the number of nodes that were licensed depending on the available space, the power budget and the performance that is required. For instance, the older classification network can run in just 4 nodes. The Akida IP has been designed to be modular and can be favored to fit the requirements. An IP license is followed by royalties that are payable on each product that is manufactured. The Akida chip at AKD1000 is offered as a low-cost device for development, prototyping and small series manufacturing. We designed this chip to fit a wide range of applications with many interfaces: a microprocessor, a neuromorphic processor array of 80 cores. Once our requirement goes over 1 million units, we will recommend licensing our IP to build a custom chip. And this would be the case, for instance, for consumer goods. Our design partners received training and tools to design custom boards for Akida product clients. BrainChip has designed and will be manufacturing at least 3 different modules. The M.2 science module shown here will be used in laptops. Most laptops have an M.2 slot. The USB 3.0 dongle, right there, will be used in any USB port. We also have a PCIe board that plugs into the desktop computer. Solution providers will use these modules to build products. Peter, are you with us already?

Yes, we are. I'm using my Internet on my phone to connect, so it's working.

All right. We are on this stage that you're seeing under the screen. And I just finished talking about the 3 different kinds of modules, the M2, the USB and the PCIe board. So you could probably pick up here on the [ tachyon ] board. No?

[Technical Difficulty] which is a low-cost computer. It makes a complete embedded system that can be integrated into products straight away. This makes a great educational kit for people to familiarize themselves with Akida as well. MetaTF, together with our modules, are the tools that quickly enable engineers to use Akida in a wide range of home appliances, intelligent washing machines, something that can sense machine loads and temperatures.

Apologies for that, everyone. Looks like the Perth venue has disconnected again. As Peter was saying -- Peter, okay, you're back, Peter.

Sorry, it means that you missed the last couple of sentences?

On the dishwasher -- washing machine.

I'm going to try to see if we're going to switch back to the other end, if it's back up again. Am I online at the moment?

Yes, sir.

Okay. I would like to have the next slide. Or should we continue with this slide and go back about 1 page?

Let's go to Slide 19. Peter? [Technical Difficulty] Oh, boy. All right. They're frozen again in Perth. So let me take us through this until our CEO can join us again. You're now looking at slide that shows the trend that has been observed that is to move intelligence to the edge. What is the edge? Well, basically, it is every piece of electronics that does not need to connect via the Internet to a large data network to work. Most of today's intelligent edge devices, like Amazon Alexa, send everything through the Internet up to the cloud, which means nothing but connecting to a huge computer system in a megawatt burning water-cooled data center. The data center Colossus does all the heavy AI lifting and processing the remote data and sends the answers back to Alexa. Here, a second delay does not matter much. But if you are in a fast-moving car, that could be very important. It could be deadly. The next evolution of this process was to set up regional data centers closer to the user of products for distributed processing. This solved some of the problems. It is faster because transmission lines are shorter. The regional data center is still a large behemoth computer that burns many hundreds of kilowatts and needs forced cooling. The security threat is not diminished. Bandwidth and latency problems are still significant. So maybe now you're traveling 15 meters before your car gets a response, but that's still deadly. Edge of the cloud platform brings processing even closer to the application device. These are smaller computing devices that boost some local processing and communicate with the cloud. Delays are still significant because these computing nodes are not as fast as the large data center computers. And security is still a problem because anything that runs programs can be hacked. With over 30 billion IoT devices and 10.3 billion computers connected to the Internet over the next 4 years, there will be insufficient bandwidth. Metadata, the output of Akida, can solve this problem. The delay to get information back will increase because of this causing a bottleneck. In a self-driving car at 110 kilometers an hour, you're traveling 30 meters per second, much can go wrong in a second. So devices that rely on the cloud do not work at all where there is no Internet or the Internet is down, like it's now on Perth. So connected devices such as cameras, medical diagnostic applications can also be hacked, causing serious security threat. So Akida represents the advanced far edge, sensors and application devices that are far removed from any of the data centers and don't need to have an Internet connection at all to work. All processing is done on Akida and at the point of acquisition. The Akida neural fabric consists of simulated brain cells that do not run a program so the neural network cannot be hacked by inserting malicious code. Akida is fast and delays are minimized. Security is improved, processor efficiency is improved, Internet bandwidth is improved or eliminated and Akida can continue to learn after leaving the factory. Next slide, please. In parallel to developing Akida and MetaTF development environment, we have endeavored to develop an expanding range of application examples. We are not a solution provider, but a chip manufacturer. We have examples of Akida in real-time learning, order classification that can be used for food safety and disease detection, bacteria in blood detection, 3D point cloud object recognition, air quality classification, cybersecurity and access control security, anomalies and fault detections. This covers all main areas that you see here. Expand to many different applications, resulting in expanding opportunities for the Akida neuromorphic processing chip. In health care, for example, We can sniff and detect disease or monitor blood samples, but also see using robotic surgery systems, X-Ray and ultrasound image interpretations, patient statistics, patient monitoring systems, almost like an artificial nurse. Next slide. Figure 1 gives you a quick comparison of Akida to the biological brain and the standard deep learning methods. Important thing to notice here are the power consumption in microwatts. We might be one more slide...

You're on the same slide, Manny. We need to advance one further.

Yes. Advance one slide, please? Okay. You want to take it on now, Peter? Slide 21.

I'm not sure if I have enough bandwidth, Manny.

Why don't I -- if you wish, I could carry on and you could just interject, please, anytime you wish.

Yes. Thank you.

So Figure 1, as we're now looking at here, gives a quick comparison of Akida to the biological brain and standard deep learning methods. Important thing to notice here are the power consumption, microwatts. That is millions of watt to a milliwatts. To put that into perspective, an odor classification system could run on a penlight battery for 5 months. Akida is a green technology. It also learns in real time on the chip and it is very efficient and fast compared to the old GPU method. Next slide. Our current examples cover all 5 senses of humans, a visual object detection, such as manufacturing. Accelerometer and tactile sensors, bridge, resonance, fault detection. Auditory meaning speech and sound recognition, keyword detection, engine noise classification and preventative maintenance. Olfactory, detection and classification of odors, detecting disease in breadth, detecting spoiled food, food manufacturing, quality control or air quality and gestation, food and drink, taste control. Next slide. Now Akida AKD1000 has 5 main advantages over the large data center, the Himax computer. Akida does the same job but in a very different way. In the same manner as the brain, it is processing data simultaneously in many individual simulations of brain cells. It uses the same learning methods as the brain, which learns instantly. It is thousands of times more power-efficient. It is a small and lightweight, which is an advantage for flying machines such as drones, aircraft and spacecraft. It generates almost no heat because it uses very little power. We have no on-chip convolution -- oh, sorry, we have on-chip convolution. Convolution is a technique that is used in deep learning networks. This makes it possible for Akida neuromorphic processing to run existing deep learning networks with extreme efficiency. Next slide. Next slide, please. Yes. The future looks bright indeed. With all the advantages that Akida is offering the expanding artificial intelligence industry, we are happy to move from R&D focus into production, manufacturing of modules and sales. We released the design of the Akida chip to our production partners, Socionext in Japan. Akida is not a simple chip. It has many layers which are processed at one layer per day. We expect to have several thousand working chips by August of this year. We received and tested engineering samples back in 2020, many of which have now been delivered to our early access partners. Our sales and marketing team are gearing up to hit the road running. Our priority is to enable design partners and solution providers with the tools that they need to start integrating the Akida technology either as a chip or as a module or as an IP. BrainChip sales team will continue to drive revenue through licensing of the IP, chip sales and module sales as well as expanding our design partner program. As we mentioned last year, we had our first order for our IP license in the fourth quarter. This project is running well, and the client is happy with our level of support. We are also working on a proof of concept with a large European manufacturer of car electronic systems. We have accomplished much with a small team, more than some of the large guys with deep pockets and large teams have. Next slide, please. I just want to walk you through the 7 stages in our sales process. It starts with creating awareness through online blogs, trade shows, direct sales, social media, scientific papers and articles, and it moves on to consideration. Does the product match the customers' requirement? We focus on sensor data classification, wearable electronics, medical diagnostics and consumer electronics, and that includes all the many categories I showed earlier. Then we move on to evaluation. This is where the customer engages and evaluates the Akida technology in their application. Then we move on to customer support. During the evaluation process, BrainChip engineers and scientists provide the client with support to design the best possible solution to integrate the Akida technology. And hopefully, that leads to IP licensing. Once the evaluation process is finalized, the client either purchases an IP license or places an order for an Akida 1000 device. And we go into development and testing. The client then develops their own product and test their products' functionality and market acceptance. Then we move to production and sales. At this point, the client buys quantities of Akida chips. Or in the case of an IP license, they start to pay royalties for each product that they manufacture. The next slide is a video. So if you could play that, please. [Presentation]

Thank you for that. Let's move on to the next slide, please. Now this table gives a quick overview of the type of devices and functions enabled by the Akida processor. For every device can be a wearable electronic device, can be configured by the user, can be a battery -- can be battery operated or can be used in remote or battery remote sensing, Akida checks all those boxes. In autonomous machines, Akida enables user configuration, solar power or battery power-operated and remote sensing. In augmented reality, Akida enables artificial intelligence in wearable electronics, user configuration, real-time learning and battery operation. Artificial intelligence in home appliances, refrigerators, washing machines, TVs, microwave ovens and entertainment systems, to name a few, Akida enables user configuration and intelligent monitoring or sensing. It could put on your favorite channel when it sees you sitting on the couch. In all appliances, it increases security and privacy because nothing needs to be uploaded to the Internet. All sensory perception is processed locally on the Akida AKD1000 chip. Next slide. Okay. This slide shows the unfair advantage that we have over competing products. Only the Akida AKD1000 check all boxes. We have micro to milliwatt power consumption, real-time on-chip learning and training, a familiar TensorFlow development environment that the data scientists already knows, the possibility to run in stand-alone mode, and we have on-chip convolution to make it easy to port existing deep learning networks to the Akida AKD1000. IBM have a neuromorphic chip called TrueNorth. First, this is an experimental chip. It is very large, about 2.5 square centimeters and therefore, very expensive to make. It does not learn. It does not have on-chip convolution. And to use it, you must learn a new language called the Corelet Language. Intel has also developed an experimental neuromorphic chip called Loihi. It has a programmable core if engineers want to add their own learning algorithm. It does not have on-chip convolution. And to use it, you need to learn the complex rural engineering platform. Both these chips are low power, but not as low as the Akida 1000. Deep learning accelerators or DLAs are math chips like small versions of a data center. The neural network is executed in the computer, but the additions are accelerated on the DLA chip. They do not support any neural operations, no learning, cannot operate independently from a computer and generally use 2 to 5 watts, or 10x that or 1,000x that more than the Akida depending on the application. Next slide, please. Now this is third-party forecast of the market for the edge devices up to 2025, prepared by Tractica Research. Many of those devices could be using the Akida 1000. You can see that by 2025, the market size reaches nearly $58 billion, distributed over industries, users, mobile users, consumer goods, automotive systems, ADAS, preventive and safety systems and drones. I expect that the air and spacecraft will also be included in this group. The prospects for the BrainChip Akida technology are huge. Many, if not most of these systems will need artificial intelligence processors that is small, light, fast and power efficient. We deliver all those capabilities in Akida 1000. Next slide. The number of Internet of Things devices will triple between today and 2025. All these devices will be competing for internal bandwidth. IoT is a category of machines, and this is different from Internet of people. With on device, near sensing processing enabled by Akida, Internet bandwidth problems can be solved, even when metadata that is processed already by Akida is uploaded to a central system. For instance, a video stream takes a lot more bandwidth than a simple text warning that an unauthorized person has entered the room. We expect that many of today's dumb devices will be replaced during -- over this period by intelligent Akida-based devices. In addition, there will be some 20 billion new devices. This is another great opportunity for the Akida -- BrainChip Akida technology. You can tell we are excited about these prospects. Next slide. Now if we split up the $58 billion market into market segments, we get the chart that's shown here now. All of these categories match up with the expanding opportunity slide that I showed earlier in the presentation. Akida technology is flexible and can be used in a large number of very different application fields. Next slide. Now the IP licensing model that we use in financial -- is financially attractive. After the initial licensing fee in successive years, the client pays royalties on each product that is manufactured, royalties continue over the life of the product. This has a cumulative effect, royalty stacked up over the years to come. Every new license has the potential to add a new layer of income to the stack of royalty income. We are right at the beginning of this evolution with our first license project progressing well. Next slide. Akida 1000 chip is in production, and we expect sales to take off soon. Initially, we expect to sell complete modules, while our clients are developing their own products and boards. Those modules will contain Akida 1000 chip and the MetaTF tools. This is where we expect the first production chips to be absorbed. Sales are expected to increase once clients have finished their development work and are integrating Akida 1000 into their product. Module sales are likely to continue into the future, followed by new products such as Akida 2000, which will be a chip with additional features to execute sequence learning networks known in the industry as LSTM and transformer networks. Important part of Akida 2000 are already working now in the lab here in Perth. Akida 3000 chip is in development and will be aimed at capsule network that are under development at Google and the cortical neural networks of the future. Next slide. I already gave away the clue to this slide. The Akida 1000 is indeed the beginning of an exciting range of advanced neural network products, each aimed at specific market segments of this evolving technology. I'll illustrate the reasons for these products in an example. Today, if a plastic bag blows across the street, a car equipped with AI sees an object and hits the brakes or worse takes evasive action. Our future networks will be able to learn the difference between plastic bag blowing across the street and a rock by their behavior. They will learn from sequence of events and from observing behavior. A rock does not get blown in the wind. A door has a handle to make it open. Objects have not only a shape, but also behavior and location in space. Our brain understands and predicts the behavior of objects and sounds. We aim to build that intelligence into future products so that we can enable our clients not only to build intelligent products, but safe and beneficial products as well. Next slide. The decision to end Mr. DiNardo's employment was mutually agreed upon between the company and Mr. DiNardo, as I mentioned. BrainChip is moving from an R&D preproduction phase into a commercial production phase. It was agreed by both parties that the company needed to recruit a CEO with the appropriate skill set, the drive and vision to lead the business through the upcoming commercialization phase and beyond. Lou joined the company in 2016 when we were still in a research stage. Since then, we have made much progress, culminating with Akida 1000 chip being in production. And we also got upgraded to the OTCQX in the U.S. and also got included in the ASX 300 in Australia. To attract the right people, we must be attractive. Part of that attractiveness is to issue options. BrainChip has to compete with big boys like Qualcomm, Intel, IBM and a dozen others who can offer incentives and salaries that are far beyond what we can offer sometimes. To attract the right quality of people, we grant option as a long-term incentive plan. Each option has a strike price that is equivalent to the share price at the time the employee joined. They can buy those shares at the price when the options vest. It is, therefore, in the interest to increase the value of the company. This is the incentive that is offered by options. The share price must go up for an option to be worth something. In this way, we align our interest -- we align the interest of our employees with those of our shareholders. We have created a trust that holds a pool of shares available for allotment to our employees during -- when they exercise their options. We are growing our sales and marketing team, our engineering team and our business operations to meet the needs of our clients. We must have enough people to support a team of certified solution providers and certified design partners. In parallel, to our effort to make -- to market the Akida technology through sales channels, we continue our social media campaign, publishing in scientific journals and professional publications, trade shows and direct sales. Next slide. With the recent elevation of BrainChip into ASX 300 and the OTCQX upgrade in the United States, we expect to attract more institutional investors. I will be doing a road -- Peter will be doing a road show in the Eastern states after this AGM. We are also improving our Investor Relations communications, and you will notice this going forward. We recently employed a professional Investor Relations Manager, who is based in Perth and speaks Australian rather than techy. His name is now on all our releases, and he is happy to answer your questions. We are evolving, not only as an industry leader in artificial intelligence hardware, but also as a professional company. Now we don't make sensors. We make them intelligent, making AI simple, that is also part of our mission. Thank you. Sorry, I had to [ append that ] for our CEO due to technology issues. It's now back to me, Manny Hernandez, your Chairman. What we would like to do here is answer questions that we have received in advance of the meeting. Peter and I parceled this into questions that were more appropriate for me as Chairman to answer and for him as CEO. So let's see if he is able to answer his sets of questions.

Let me start with mine. We got questions about executive and Board remuneration and incentives. As you noticed on the resolution for the REM report, there were obviously some concerns over what we do as a company. Now the way I would like to answer that for you folks is BrainChip is a relatively small Australian company competing for high-end talent in a fiercely competitive global environment dominated by tech companies with far greater resources than we do. Now one of the ways we must compete is by offering competitive remuneration package to attract the best talent. This package are benchmarked against our competitors in Silicon Valley, where most of the best people in our industry want to work. Now while the Board and the executive remuneration packages look very generous by Australian standards, they are benchmarked to the U.S. market where the best talent that we need resides. If we want to attract the best people, we must pay them what the market calls for. As a matter of urgency, however, the Board will be seeking to align the interest of directors and executives and the stockholders by ensuring that most, if not all, remuneration packages and incentives are directly linked to performance hurdles or performance or key performance indicators or KPIs, as we call it. And this will be based on independent global benchmarking reports as well. Now the second most frequent question we got was with regard to Ms. Christa Steele's resignation. So just to elaborate. Ms. Steele resigned from her position as a non-Executive Director due to a serious illness within her family, and we fully support her decision to do so. Ms. Steele has served in good faith as a Director of the company over the past 8 months and we wish her well and her family for all -- and we wish her and her family all the best for the future. In fact, on behalf of the Board, I'd like to thank Ms. Steele for her service. She will not receive any option or any other compensation upon her resignation. Now a search for a new director is already underway, and we hope to be in a position to make an announcement in due time. The next set of questions we received were in relation to our listing in the U.S. OTC system. So let me say a few words about that. The primary purpose of our OTCQX listing was to open BRN or BrainChip up to the U.S. market, particularly institutional investors and expose our company to a bigger pool of professional and tech investors in the world. U.S. investors want to be able to trade on a platform that is domiciled in the U.S. and denominated in U.S. currency. Now this is not an exchange where new shares are issued and traded. Rather, it's a pool of investors in the U.S. who can trade BrainChip between themselves in U.S. dollars and can buy and sell BrainChip without going directly through the ASX to do it. So we have recently upgraded our listing on the OTC to the highest level, which makes it possible for a wider range of institutional and professional investors in the U.S. to trade and invest in our shares. This next question is actually a frustrating one for you folks. It has to do with why cannot we disclose Early Access partner participants? And why don't you provide more recent updates on those? Now many of the large corporations that are EAP participants insist that their involvement remain confidential as they don't want their competitors to know what technology they are evaluating and which companies they are talking to. We are bound by confidentiality agreement with these companies. We are extremely limited in what we can say, if anything, about who the participants are and how they're evaluating process -- how the evaluation process is progressing. However, we are also bound by our continuous disclosure obligation through the ASX listing rules to provide updates on any matters that may be reasonably deemed to be material to the company's share price. So our approach is to only make announcement on these matters if they are material and not name any of the participants in the EAP program unless they physically say it's okay for us to do so, such as NASA. Now we understand that this is frustrating But ultimately, it is about BrainChip becoming a trusted and reliable business partner to some of the world's largest corporations. So it is in our interest and the interest of our shareholders that we respect this confidentiality agreement and the wishes of our potential customers. Another question was raised on proxy advisers. Now this is our first year in the ASX 300 and our first opportunity to engage with proxy advisers. So we will be reaching out to the leading proxy adviser firms to seek feedback and learn more about their expectations, including in regard to remuneration matters, dilutionary matters and other business matters. Now this is only an example of why we recently hired our Manager of Investor Relations, Tony Dawe. Tony will be our point person for engaging with capital markets in Australia and in the U.S. and ensuring that we understand market expectations of our business as we grow and expand. Peter, are you able to speak?

I do not have video because the folks have stopped it [ with us ].

So Peter, would you like me to just carry on then?

Yes, please, Manny, of course, we did not have broadband internet.

Yes. Apologies, everyone. We actually practiced this. We rehearsed this, the technology worked. So it's not for lack of diligence that we're experiencing this. It's unfortunate, and I'd like to apologize to everyone. But in the spirit of continuing in with our agenda, let me address some of the questions that were fielded -- that our CEO, Peter would have gladly answered for you guys. One is almost a rehash of Mr. DiNardo's departure. Now Mr. DiNardo was brought into BrainChip to grow the company, and they achieved that. We became an ASX 300 company with a market cap approaching $1 billion on Lou's watch. However, both Lou and the Board agreed that the company required a CEO with a different skill set as we transitioned from R&D phase into production and commercialization phase. Now Lou remains a major stockholder and a strong supporter of the company, and he is now preparing for his well-deserved earned retirement in the U.S., and we wish him all the very best. How are we doing on the search of a new CEO? Now since Lou DiNardo's resignation, the Board has been focused on the search for our new CEO as a top priority. We're pleased to advise that we have a short list of very suitable and impressive candidates for the position of BrainChip's CEO, and we are in the process of vetting them, conducting due diligence and ensuring a strong alignment of personal, professional values, deep industry knowledge, formidable contracts -- contacts and connections and the drive and ambition to unlock and realize BrainChip's incredible potential. In addition, we have a parallel process of vetting professional search firms that might be able to help us if we have exhausted our internal effort through our network. A common interest is a question on volume of Akida chips being manufactured. On April 13, we announced that the commencement of volume manufacturing of Akida 1000 by TSMC has commenced. This represented a major milestone and a transformational moment for our company. We also indicated that the first [ production erected ] right around August of 2021, and we are currently on track to meet that deadline. Now commencing Akida chip production is a major achievement, particularly at the time when there is a global chip shortage and chip production slots are very limited. So we've been able to manage that really well. It is important for our shareholders to know that each and every Akida 1000 chip takes approximately 13 weeks to fabricate and produce, and this requires extensive testing and evaluation. This process usually takes several months to complete. Now it's vitally important that we ensure the highest level of quality control in this initial production runs. So Akida gains a reputation of high-quality, reliable performance. If we rush and cut corners, which we won't do, we will damage our brand reputation in a market that is highly demanding and ultra-competitive. Our long-term success depends on getting this early-stage production right, which is why we're starting small and being very diligent. Our initial production run should produce enough chips for our internal validation as well as to support certain EAP partners that are expecting them. The more important point, though, is once we've identified high-volume chip customers, we will be able to order production runs for them on a just-in-time basis. That is also beneficial to us at BrainChip so we don't have to build inventory in advance and just let it sit on our shelves and burn cash. With proper planning and coordination, we can get those units to our customers within a 3-month lead time, which is customary in our industry. Moving on. So this is one of the more important questions. So how does BrainChip make money? Now our sales and marketing strategy brings our products to market in a way that maximizes chip sales and cash flow, but also meets a wide variety of growing range of customer needs and technical requirements. Now our strategy provides a 3-tiered approach to customer engagement and allows us to meet the needs of multiple customers simultaneously. The first tier is to sell Akida IP through a license fee and residual royalties. This provides us with cash up front and then a recurring income stream over the life of the IP agreement. The second tier is to sell silicon, currently, Akida 1000, to OEMs, system houses and board and module manufacturers. In short, selling throughout the whole supply chain where there is a demand for AI on edge-based devices. Now many electronic goods manufacturers want to incorporate AI in their products, driving their strategic direction is the cool thing to do and gives them a competitive edge. Few know how to accomplish this. There is a shortage of data scientists in the world, which may not -- which may get worse before universities can provide enough people to fill that gap. The shortage of data scientists and the demand for incorporating AI enables BrainChip to leverage our skill set by providing implementation, integration and design services for an additional path to revenue. Now finally, with the drive for AI, there is a need for simplicity and a plug-and-play approach. Our third tier is focused on proliferation of Akida. We are producing a development kit to proliferate to the masses. This development kit will be a low-cost, easy-to-use, plug-and-play kit. We will work out, out of the box and let the users run through the examples of BrainChip has developed for odor, image, vibration, sound, gestation, tactile classification. The goal is to focus on the inexperienced and experienced users, enabling them to get results quickly. Understanding of AI, so they can start designing with Akida 1000 devices and integrating it into their system. This kit can also be used to integrate Akida for prototyping and low-volume environments. All right. Those are our answers to the prepared or questions sent in advance. I will now open the floor for any further questions from the shareholders. Kim Clark, our Secretary, will be moderating and capturing those questions.

Manny, there's a number of questions, many of which center around the volume production and quantification of the chip and whether we have another slot booked with TSMC. I believe you've responded to many of them. But what I may do, with your consent, is just to summarize those to see if there's anything that you wish to add. A question, as I said, being whether there is another slot prebooked with TSMC for another run of wafers. How many chips -- sorry, Manny?

Why don't we do this? I think the volume and the units and be able to order production I've already addressed. But I think Anil's here, our Chief Development Officer, Anil, who has a deep and personal relations with Socionext and TSMC. Maybe an opportunity to answer that question, Anil, in terms of production run slots at TSMC.

Yes. Simple answer is yes, Manny. We can, but we have to first test the first log that we have, and that's progressing well. Once we get there, everything is good, as the customer demands come in, we will always be going in for another slot, depending on the demand.

Thank you, Anil.

Sorry.

Go ahead, Kim.

So just some further questions on that point. What percentage of the chips this year have been presold and the status of any subsequent orders on those?

Let me first answer that and then I'll let Anil follow as well. We have not purposely presold products because we want to make sure, as Anil mentioned, that the production ones coming out of TSMC are perfect, functional, high-quality, reliable products. So we don't want to presell chips until we are all assured as a team and as a company that we have a really good product to offer at the production scale. Anil? Anil, you might be on mute.

Yes. Yes. So there is no need to any presell. I mean it depends on our what we call just-in-time supply chain. So we will manage it as our requirements from our customer goes up. We have a good relationship with our partners, and they have a good relationship with TSMC. Not be an issue with supply of the chips. But no, we have not presold anything for obvious reasons that Manny mentioned.

Thank you, Anil. All right, Kim.

On the back of that, there's 2 questions in relation to the Renesas transaction. One being have we received any royalties from the Renesas IP deal yet? And the second being, does it represent an actual plant manufacturing run of Renesas products?

We have not received royalty yet, but we have already received a significant portion of the upfront agreement, which is the licensing value of our agreement. Royalty, as you all know, comes later on when they've successfully integrated Akida into their product and start shipping products. Anil, would you care to add to that, please?

Yes. The royalty is only shipping when they start shipping their production quantities, which typically takes some time. This is a standard practice in the industry.

Thank you.

The final question on this point is, can we clarify current chip production numbers, which you've done and likely sales in the next few years? Are we talking thousands or millions of units?

I would rather not speculate on that. Let's all wait for this production output that we're all anxiously waiting for in August. And given how good the development team has been in delivering to their specs, I don't expect us to have a problem, but it's best as a company to wait and see. We will gladly announce that news when we finally get there. I think it would be reckless, careless for us to be projecting millions or thousands of chips at this point. I wish I could, but given once we have visibility as to the repeatability and predictably of our pipeline, I will have no problem sharing projections with you folks. But I think for now, that would not be proper for us to do.

Okay. The next question involves Akida and asks what safeguards, if any, will be in place to stop Akida being used in less beneficial applications such as being weaponized?

Good question. Peter?

The -- any technology that you develop can be used or abused. We are designing Akida to be used in beneficial applications. That's the focus of the whole project from the beginning. If anybody is building Akida into a weapon system, we cannot legally stop them from doing so. So if you are looking at the reach of a neural network in a weapon system, it will be no different from any other computer system that will be used in a weapon system in that we have a million neurons in Akida. And it can be used for classification and so can our components be used for classification, [ without altering the business ].

At least the way we could at least influence that is knowing our customers, knowing the applications that they're intending to use the product for and have a little bit of influence on whether we want to do that or not.

There's another question, Manny, which I believe you've covered off, but in case you wish to add anything further, asking whether BrainChip is the only group developing this type of technology? If not, who are our main competitors in this field and how advanced are they? Is there anything else you want to add on that subject?

Maybe for Peter and/or even Anil to add.

I can think of one. I think we already answered that in the 5 advantages that we have over the rest of the field. If you're looking at other neural marker chips, such as Loihi and Intel. Intel and the TrueNorth chip from IBM. They are also very complex chips, but they do not have the on-chip convolution. So this was used in deep learning networks. They also do not have native on-chip learning. So even the power supply, the power consumption of those chips is higher than Akida. So I believe we have really a winning part over other products that are out there. Of course, there's also DLAs, Deep Learning Accelerators, which accelerate the execution of neuron networks on a computer. But you still need a computer around the neural network in that case. So in applications where you have sufficient power available, where you have passive 10 watts available, you produce [ in accelerated ]. But any product that is wearable, battery powered, remote sensing and solar powered, Akida stands out and is the only product that make sense.

Thank you, Peter.

Another somewhat technical question, Manny, is the Akida IP core delivered as a hard or synthesizable core? At what technology nodes? And can you comment on how portable the IP core is to different nodes?

I think we'll have Anil answer that one.

Yes. It is delivered as an RTL core, which is synthesizable, and it is process-agnostic. It can go to any process technology, whether it's 7-nanometer or an old technology like 65-nanometer.

Thank you.

[indiscernible] nanometers if you want to [ address that ]. There's no limitation on which [ chip ] policy we can use.

Go ahead, Kim.

Manny, there are a number of questions also seeking updates in respect of NASA, Ford, the NDAs and the Magik Eye announcements that have previously been released, basically asking for a status update on each of those.

Yes. Why don't I just lean on the answer that we've -- or the comment that we've already made earlier relative to disclosures. It's our obligation, obviously, based on the tenuous disclosure rules, significant material events that could be sensitive to the stock in particular. So we're very mindful not to just be announcing anything and everything. In the first place, there's still confidentiality rules that need to be complied with as far as these partners, it was good enough for them to let us say their name, but it is not always okay to tell the world what they're doing with our technology, for reasons we already mentioned. Number one, they don't want their competitors to know what they're doing; and number two, which companies they're talking to. So at the proper time, ladies and gentlemen, when there is something material and significant development with our EAP partners, we will let you know as that is our obligation as well.

Manny, the next -- there's 2 questions that are somewhat related. One being what is the update on the significant revenue uplift that was promised to materialize from the last quarter of 2020? And the second being, where does the company see itself positioned at the end of 2021?

Yes. I think I'll -- as I mentioned earlier, we are really not in a position to be sharing short or even long-term forecast, particularly in revenue with you folks until we can get a relatively predictable sense of how our pipeline is converting into real sales. Otherwise, it will just be all hope in expectation that it happens. So I really would not like to build expectations out there as to specific numbers or ramp rates. I think what's most important for us as a company is to, in fact, see the production run through, which a lot of our partners are quite excited and waiting for. And once we've gotten that production units out of TSMC, hopefully gives us a better sense of what kind of production units or sales can we get out of that once it's available. In the meantime, We have also continued to push on the licensing of our IP. Now we've obviously had a contract with Renesas. There are several leads that we're also pursuing. But like anything else, those take time and subject to a lot of evaluation by our potential partners. So sorry for the long answer, but I don't think we, as a company, are in a position to give you sales forecast, ramp rates at this time, either for 2021 or 2022. That will happen someday, and I hope that's soon.

Manny, the final point which I can respond to was to ask us to record and upload a video of the presentation we are trying to give because of the content drop out with the Internet issues. So on that, the presentation and the addresses of both the Chairman and the CEO were released to the ASX immediately prior to this meeting. And Manny, that's the last question received from online shareholders.

Thank you, Kim. Thank you, everyone, for sticking with us here through this whole AGM session. And once again, for your support and continued patience with us as we take BrainChip to commercialization. Just a formality, please ensure that you have cast your vote on all the resolutions. For those stockholders in attendance at our Perth venue, please make sure that you place your voting card in the poll box held by the Boardroom representatives. All right. I will now declare voting closed. We will publish the final voting results with the ASX and on our website as well. Thank you all. That now concludes BrainChip Holdings' Annual General Meeting. Thank you all for your attendance and your continued support of the company and call the meeting to a close. Thank you.