 Okay, so hello everybody. Thank you all for joining us today in another episode of the IEA's Fusion Breakthroughs webinar. For those who missed the previous episodes, I'll put the links on the chat and in the comments below depending on whether you are watching us live or on YouTube. Before we begin our session today, let me warmly welcome our two speakers who have kindly devoted their time to present to us today about breakthroughs and fusion at their companies and hopefully answer any questions you may have. Our speakers today are Dr. Artem Smirnov, the Chief Technological Officer at TAE Technologies in the USA, and Dr. Michele Leverge, Founder and Chief Scientific Officer at General Fusion in Canada. We again thank them today for their availability. My name is Adam Daniel-Wovich working at the International Atomic Energy Agency and I will be moderating today's webinar. Here at the IEA we work to foster collaboration and coordination on fusion R&D and move forward in developing the peaceful use of fusion and energy. We've been doing this work since 1958. This October we will be organizing the 29th Fusion Energy Conference in London. Registrations are still open for this, for those interested. And in today's session we'll be featuring an overview of recent results from TAE Technologies. This fusion approach relies on field-reversed configuration plasmas. The company was founded in 1998 with over 1,100 patents granted, 1.2 billion dollars in private capital raised and five generations of devices built along with two more in development. The talk will include an overview of recent experimental results from the current experimental machine CTW aka Norman along with TAE's plans for the next phase devices. We will then hear from General Fusion who are developing magnetized target fusion. As far as we're talking about is created inside a rotating liquid metal cavity after which pistons driven by high-pressure gas can press the fluid and the plasmas diffusion conditions. The company claims no expensive technologies are used leading to a more cost-effective solution. Results so far and plans forwards will be discussed. We're about to hear about these topics. The format will be a sequence of two talks, 30 minutes each, followed by a 30 minute Q&A session at the end. Please type your comments in the chat box and to whom it's addressed to and we'll go through the question at the end in 30 minute Q&A session. Without further ado, please welcome Dr. Artem Smirnov, the Chief Technological Officer from TAE Technologies. Thank you, Adam, for the introduction and thank you for the opportunity to speak in front of this esteemed group of people. Appreciate it. How's my audio? Still good? Yes, everything sounds good. Let me share my screen. We should be in the presenter mode, right? You guys see the landing page. Excellent. So first of all, a huge thank you to all of you in the entire plasma physics community for playing a great foundation for our research at TAE. We're definitely standing on the shoulders of giants who came before us in the field and without all of your contribution, our progress would not have been possible. So huge thank you to everyone and also internally to our investors and the team, of course, I'm here representing the team of now over 500. Adam was kind enough to mention a few of his beats of trivia. So that makes my job easier. So as many of you know, TAE was founded 25 years ago. We are a spin-off of QC Irvine, first of California here in Irvine. With the mission to commercialize practical fusion power and deliver fusion derived technologies to solve existential human challenges. We've built a couple of generations of national lab scale devices, which I'll describe later. We made quite a bit of progress with our approach to fusion and validating that in the lab. We set up a lot of truthful collaborations with many national labs in universities, both domestically here in the U.S. and internationally. And for those of you who are interested, all of our research publications could be found on our website on the research library at TAE.com where we have a couple hundred of peer reviewed journal articles as well as several hundred research posters, conference proceedings. So all of it is there. So please check it out. Well, the reason why we're doing what we're doing is because, as all of you know, the global energy demand is rising. The electric power forecast predicts that the worldwide consumption will double over the next couple of decades. And that's a problem because we need a stable and carbon-free source of baseball power. And fusion is an excellent candidate for that. We at TAE strongly believe that fusion, just like the space race, can spawn a lot of innovations in adjacent fields. And from our pursuits of clean fusion, we spun out a company called TAE Life Sciences, which has taken our particle accelerator technology into the cancer radiotherapy space quite successfully. And if we have a little bit of time at the end, I can talk about that briefly. And more recently, we spun out our power management technology into the electric vehicle and residential and commercial energy storage markets. And this is just a couple of examples of how fusion can create tremendous value for investors and society more broadly. Right now, before we even conclude this quest to clean fusion power, which still is sort of at least a decade away. So I'm going to talk today about TAE Vision and Roadmap. I'll give you an overview of the key program achievements. We'll talk about next steps. And then if we don't run out of time at the end, I can say a few words about the spin-offs. So as you know, TAE is in pursuit of the a-neutronic fusion, so-called a-neutronic. And the philosophy for that has always been beginning with the end of mind. We strongly believe that the practical solution for fusion would come as a confluence of three sets of factors. The scientific category where, of course, we need a solution that provides the appropriate plasma performance characteristics, but also we need a solution which where engineering is tractable and doesn't present immense challenges. And the a-neutronic path certainly is attractive from that perspective. And also we need fusion ultimately to be cost-competitive and just commercially viable. And that requires relatively compact size, small environmental impact. And it is the overlap of these three sets of drivers which dictates what would be a practical approach. Now, our ultimate goal is Rottenborn-11 fusion. And this reaction itself is a-neutronic, but of course, as you know, there are some secondary reactions here in the mix which come into play so that some neutrons will be produced. Now, there's been some debate recently as to what can be legitimately called a-neutronic and what not. Well, in PD-11 case, it's less than 1% of output power comes in the form of neutrons. And in the US, interestingly, some states actually adopted this definition that less than 1% is called a-neutronic. So we're just following that definition. But of course, the drastic reduction of neutron production simplifies the engineering and just alleviates many of the severe problems that the DTP app is facing. Hydrogen boron is a very benign and really available fuel, produces very little radioactive waste. The good news is that the cross-section for this reaction is actually larger than was previously believed to be. On the top right here, there are some measurements that we did about 10 years ago. There was a full research dedicated to validating the cross-sections. They are larger. If you do accurate kinetic modeling, which we published a couple years ago, you will realize that the preconception that PD-11 cannot ignite is actually false. It can ignite, although the margin, as you can see on the bottom right, is very weak. And that's not practically, that's not even the point, it will always have to work as an amplifier. But there are nonetheless some very important new research findings that validate the overall approach as published by us, as well as there's some excellent research coming out recently from Princeton, where they also explore various kinetic effects and more subtle effects to offer half-waste, more practical PD-11 fuel. Of course, this approach requires high temperatures and superior confinement, which remains to be validated, and we through our series of experiments trying to get there. So our approach is based on an FRC. As you know, this is a high plasma beta configuration by design. It's compact and offers a high power density. It has large indigenous population of large orbit particles, and we enhanced that by tangential neutral beam injection in the direction of this red arrow here, just injecting neutral beams to drive toroidal, in the tokamak notation, toroidal current here, whereas muthal in our nomenclature. And this fast time population, it further increases the stability and reduces the transport in FRC, as I'll first update later. And this configuration offers radically simplified reactor design and maintenance due to the simple geometry, and also the linear shape offers the unrestricted diverter that facilitates power, cash, grid removal. So there are some very tangible practical advantages that could be leveraged here. And on our quest towards making PB11 practical, we have built, well, depending on how you count, four generations, I would say, of large devices. Right now, we are in this green band with a device that called C2W, we call it Norman now, after the late founder of the company, Professor Norman Rostaker. So Norman has been in operation since 2016. And we actually achieved the scientific goals that we set out for this device back in 2019. But then COVID happened and offered us a bit more time of useful operation here. So we pushed Norman now, so well beyond what we ever dreamed to accomplish with it. We're very happy with that. We achieved up to about 6 kV of total temperature in this plasma with now full active feedback control, achieving very robust macro stability. And this really validates our approach, well, at this level of performance, but in the fully collisional confinement regime, collision less, sorry, I may have misspoke, collision less confinement, of course. Now, the next step we're aiming to take is this device called Caternicus. It is in the final design now. We started the construction of the facility. We intend to start operating in 2025. And Caternicus is envisioned as a device that will operate only on hydrogen. We're not going to do DT, but we're aiming for the DT equivalent level of performance with a plasma temperature of about 50 maybe up to 20 kV and appropriate confinement to ultimately be able to demonstrate Q for one performance or close to that. And then, and the reason why we're choosing to do it this way, well, I can talk about that later, but we see this essentially, the market tells us that this would be a great licensing, technology licensing opportunity for us. TAE itself, we don't intend to do duty and treating, but the others may want to take this approach and complete the DT device with it. We will march on beyond Caternicus to the device that we call DaVinci, which will be the first integrated hydrogen boron-durning demo plan. We've hopefully net energy out to the, you know, 50 megawatts, so it's relatively more the scale device. We expect that this could be operational in the early 2030s. Now, we accelerate innovation at TAE by building these experimental platforms that offer fast learning cycles with machine upgrades and rapid prototyping. We heavily leverage the community with strategic partnerships and resources and talent that we pull together with others, including, as I mentioned, the University of National Labs, but also industrial players and commercial like Google and AAPRI, many others. We heavily lean on machine learning and AI in our operational optimization with feedback control of the platform. And we just in general take advantage of the forcing functions and the environment that the privately funded space imposes on us, which is we have to run fast and make fast decisions and maintain a very high pace. Now, let's talk technology and science. So we started from the early, well, devices not quite able to, but transitioning from the tabletop to the bigger laboratory scale. And then from C1, C2, and with a few upgrades, we continuously push the performance of the FRC from, you know, in the early days, as you remember, FRCs didn't last very long, they would crash and burn flames in just a couple of hundred microseconds at best. This plot here just gives a simple metric of the configuration radius sustained over time. And this is millisecond. So as we started injecting beams and improving the fast particle confinement, sustainment became better. And we were able to flat line our FRC to about five milliseconds back about eight years ago in the device we called C2U. And then we upgraded it to what we now call Norman. And it's shown here. I'll try to explain briefly how this whole experiment works. Well, it's a linear device, as you can tell, there's a human figure here for scale here, here and here. I don't know if you guys, can you guys see my cursor a little on the screen? Yeah, thanks. Yeah, I wasn't sure. Don't frequently use Webex. The cross section of the machine, the diagram up on top shows the contour lines. These are the few lines of the magnetic field. And the color map is, I believe that's, that must be the temperature, actually. But you can see the FRC sitting in the middle. And then there's the guiding magnetic field with a couple of mirrors on each side. And then we have the formation sections. So the game, the shot starts in the formation section where we do fast data pinch reversal. We produce two CTs compact steroids. And we slam them together by shooting them peristaltically with magnetic field with electromagnets. They collide and merge here in the middle and produce an FRC, which then takes this beam injection, tangentially. Beams are aimed aiming off axis. They spin up the FRC as milkly. And then the escaping plasma goes through two sets of diverters. We call these the inner diverters, not towards the midplane, closer to the midplane. And then the outer diverters, they both have sets of bias electrodes, as we call them. These are concentric rings with high voltages applied to them that talk to those magnetic field lines and provide the ECROS-B plasma rotation, which conducts all the way to the midplane. And we can create a shared flow right outside of the separatrix and that helps to stabilize global modes and reduce the transport of patrols and data later. So these are the main building blocks of the device. Now, we explored FRC sustainment in fully collisionless regime in Mormon. Sorry, this is a long, long release. I'll focus our discussion on the bottom half of it, but let me just vocalize on what has been done so far. So we demonstrated that the fast time confinement in this plasma is close to classical. We can now reliably and robustly control the stability and transport via end biasing and fast time injection. There's actually a synergistic interplay there. We've achieved the total temperature over 6 kV with electron temperature approaching 1 kV. We can now sustain this plasmas for just over 40 milliseconds, which is limited only by the energy storage we have on site. We store the energy capacitively in all of our power supplies, which by the way, you can, you know, those of you who have two screens open. In my background, that's a quote of the facility behind me there and all those blue cabinets are the power supplies with the bottom half just being the energy storage. So we have a lot of stored energy on site. Recently, we demonstrated how the historical S star over E limit, I'll describe that later, which basically is the density limit in FRCs, which is set presumably by this internal tilt mode. We removed that fast times completely break through that ceiling and that opens up a much larger operation operating space. We demonstrated the real time active feedback magnetic control of the plasma shape and position, which again helps us with the design of the next device. Now using the all these built-in whistles, we can demonstrate the millisecond scale ramp up and hitting in this plasma. And we believe that a favorable confinement scaling is emerging, which now is extended to the collisionless regime. And that offers maybe quite a favorable scaling going forward. So we will focus and I'll show some data supporting well, about half of these claims here for the sake of time, we can possibly cover all of it. So I'll, in the remaining time, I'll just try to go very quickly to some data here. So first of all, how do we know all of these? Well, Norman is a beautiful, very well diagnosed device. So just some, it's a mess, I know, but this is just to give the impression, you know, how much diagnostics we have there, we have 74 diagnostic systems, operational and more on the way some of them, by the way, you're developed through the infused grant program. And I don't know if any of our collaborators are online here, but a huge thank you for our partners. We've benefited immensely from the community in building out our diagnostic capabilities. So, well, typical shot would look something like this now. Well, this only goes up to about 30 milliseconds, but we can actually do more than that up to 40. We're actually preparing a couple of high profile publications now. So some visuals are under embargo, I cannot share them, but they will be coming out, hopefully soon in our peer reviewed papers. So stay tuned. So this is the plasma day magnetism, the, the, the separatric side size, the radio size. This is how we ran the externally applied magnetic field, total plasma temperature derived from the, from the pressure balance. And the neutron signal showing how fast times accumulate towards later fees of the discharge. And we can see that fast times really accumulate. The plasma is heating, pressure is rising. And now we develop the active external field and shape control that allows us to build up this plasma and manipulate both the size of the, the actual size, the radial extent of these bees. Now with fast time accumulation, we then can run our equilibrium reconstruction codes, which have been heavily benchmarked and really validate that the plasma entity that we create is fast time dominated. So this is a, a radial profile of the pressure and the breakdown between the thermal plasma, the fast time and the externally applied magnetic field. And, and you can see the fast times, uh, dominating here. Um, and we don't see, uh, evidence of any significant fast time driven activities to, to, to the extent that it would affect the plasma performance, which is, but true it is and, and, and almost somewhat, uh, surprising, but, uh, we're, we're studying these now. And, uh, well, this is, this is one example of how the diagnostic suite allows us to understand what's going on inside. Um, we now fully embrace the, the Bayesian and, uh, inference way of assessing the plasma behavior, meaning that we combine different diagnostics and, and reconstruct, uh, the, the plasma state in the Bayesian sense and the probabilistic sense. And that offers a much better accuracy of knowing what's going on inside. For example, uh, well, and that was developed in our partnership with Google. We've been successfully working with them over many years and, um, they brought their expertise and machine learning a, a data algorithms. Uh, but for example, in, in this particular experiment, we reduce the beam power of somewhere in the middle of the discharge just over 17, after 17.5 milliseconds. And you can see how some of the lower frequency modes, which were largely subdued before, they pop up out of non-existence. And then, well, here you would see that there is a bit more action on this. Uh, this is the, uh, interferometry reconstruction of the plasma density. Well, this is the Bayesian reconstruction relying on interferometry and a few other diagnostics. You can see how it starts bubbling up a little bit with M1, 2, 3 later in the discharge. Uh, just demonstrating that fast times actually stabilize, well, help to stabilize some of these global modes. Well, this is another look of what the plasma stability is like. Um, this is, now this is a single diagnostic. This is just the reconstruction from interferometry with the black line showing the infured separatrix radius from magnetics. And the bottom panel, uh, are the magnetic pickup coils, magnetic probes on the walls showing some global modes. First appearing after the collision emerging, but then as soon as the bi-system kicks in, they get subdued. And then the, the plasma seats very, very quiescent mode with just a few gauze, you know, perturbations on the wall, which are benign until we terminate the beams here. The beams are switched off at 15, about 15 milliseconds. Now the biased system keeps going, but alone it's not sufficient. And if you see the violent n equal two just pops out on existence and the plasma disrupts at that point. Now, with all the fast time accumulation, we, uh, we were able to completely undo this star over E limit. Well, many of you would be familiar that in the early days of the FRC research, I think it was Rosamble who predicted that FRCs would be completely unstable to the internal tilt mode, just n equal one, n equal one internal perturbation, just the magnetic moment flipping because it's oriented improperly in the relative to the externally applied field. Well, later on, it was understood that it actually, this instability is, is stabilized by a finite Larmor radius effects. But now with fast time injection, we are basically exploring these FLR effects on steroids. And you can see in the plot on the left here, how the star, we, now a star is just the dimensionless parameters. It's the separatrix radius over the, uh, uh, the, the iron iron gyrate is basically all the skin death. And E is the elongation is the aspect ratio length over diameter. So it used to be, you know, the historically the norm that the star over E would be limited by about three. And this essentially would be interpreted as, as the limit on, on FRC density because the star over E is proportional to the square, square root of density. Now with, with beam injection, you see the more beam current we inject a star over E just grows linearly, which intuitively makes sense. And, uh, we achieve the star over E of instantaneous values of up to eight and nine. And this opens up quite a bit of operating space, which was previously inaccessible because the star over E limit essentially couples density with the geometric parameters of the FRC. You know, if you have, if you have a certain density requirement, then your FRC would need to be just longer than a certain, than a certain threshold value or the other way around. Now in Copernicus, you know, in our next device, we're going to use high energy beams up to 80 kilovolts. And so we need a certain beam, a certain plasma density for efficient capture. Well, uh, that sets a certain value of about mid 10 to the 19 per cubic meter. But now we are completely decoupled with our geometry. We can, we can do anything with, with the radius and, and axial extent. And this just opens up a whole new operating level of operating flexibility. Now, just very briefly, uh, about our plasma feedback control, uh, well, to ramp up the FRC, we need to increase internal pressure for past times, uh, along with increasing the external pressure with, uh, the externally applied magnetic field. But as you start squeezing the FRC, just like a piece of soap, which sleeps out of your hand, you know, top right diagram here shows how to start pressing it. It wants to escape from you axially. So you really need to control with your external magnets, the, the axial position. And that's what we're doing here. You see, you know, the, the, in the middle panel here, the, uh, the centroid would escape from you axially. If you just started pushing on the outside, uh, instead, you, you need to control, uh, right to left, uh, motion. That's what we're doing here with, uh, on a millisecond timescale with external magnetic coils. Uh, and, and, and this offers a way to control the FRC shape here. You see that with coils, uh, at the midplane, which is EQ one and farther away from the midplane, EQ three, you know, by manipulating them in unison and running these feedback loops, we can increase the radial size and, and, and shrink the FRC at the same time or vice versa. So, uh, playing it like an accordion. And, uh, this works beautifully with, with our plasma. So we then, without the active control, you can see that we, we have shots which otherwise look, uh, similar in terms of the radial extent, but we can radically increase the plasma length, for example. Um, and, uh, the Z panel here shows how the centroid starts oscillating, but we're controlling it. And, and that allows to couple the beam power better. And the total plasma energy that's the bottom panel here just keeps growing all the way until we run out of juice. Well, again, this is a 30 millisecond shot, but we have, we have this behavior now well validated all the way up to 40. We keep ramping at that time. So we just kind of run out of juice in the facility and, and we start hitting the limits of our built-in functionality in the machine. So just, I think this, this is my two, two last slides. So first, um, we, we started the transport in this device, uh, uh, quite, quite deeply. And especially with the recent upgrades in our Doppler backscattering system, we see some wonderful evidence of transport barriers. Um, how, for example, here, pardon me, I didn't mean to do that. Um, uh, here in this bottom panel, you can see how, how the turbulence inside the FRC, inside the separatrix is quenched and the, uh, the zone of flow across B shear is formed, uh, just, just around the separatrix. And, uh, that, that offers a quiescent core and somewhat, you know, turbulent open field lines, where the ECROS B sheared flow driven by the bias system can, can suppress the, the turbulence on the open field lines. So this is fully consistent with our, uh, 3D turbulence simulations, which also find, uh, quiescent core and, and, and turbulence developing on this open, open field lines. We published extensively on this. There's a whole series of papers and these glasnosts and nature comes. So I invite you to, those of you who are interested to just explore them. And, and finally, um, uh, an interesting confining scaling is emerging, um, in our experiments on the top panel, you can see the, the particle confinement time. And this is the old data you may have seen before, how, you know, previously there was this, uh, classical data PNH FRC scaling and, uh, our series of devices just took off lyrically from there, uh, indicating that all these knobs like fast ions, bias and mirrors also, um, make this a completely new confinement, uh, scheme. And we see that the, uh, energy confinement time in particular here, featuring the electron energy confinement time in the bottom panel in this collisionless regime shows, uh, a very favorable, uh, scaling with electron temperature. It goes up as almost as T squared. We don't know if this trend will continue all the way, but, um, a few measurements in different devices, modeling and zero DS, they all agree, uh, about this trend. And we, we published on that before and we are preparing publication now too. So this is all very interesting and promising data, which brings us to the next step. And I'll do it, you know, I think I'm running out of time here. So I'll, I'll make this my last slide. And we can save the rest for maybe the Q and a session. This is catharticus. Uh, as I mentioned before, we intend to, uh, get into the DT relevant performance regimes with catharticus, which, uh, means that our ion temperature goes 10 KV plus or more like 15. This will be a post machine too, but with three second long pulses, we will run it also from capacitive energy storage, which is where, uh, our T power management solutions now excel by enabling that. Otherwise we would need to look forward to a different site for the machine. Um, this is in final design. Now we started the facility construction and we intend to start operating in 2025. And so hopefully more, more, uh, exciting data will be coming from Copernicus that will support this trends. So Adam, I think, I think I'm out of time, right? So sharing here. You'll have more opportunities in the Q and a session. This is fine. This is okay. So I guess I will almost made it through the end. I think just have just one more slide about, um, spin-offs, but we can do that with Okay. Well, thank you very much. So now let's welcome Dr. Michelle LaBerge, the founder and chief scientific officer of general fusion based in Canada. I think you're muted. Hello. Do you hear me now? Yes. Now it's good. Hey. All right. Thank you very much. I'm Michelle LaBerge, the founder and chief scientific officer at general fusion. Let's go in presentation mode, make it a bit bigger. Okay. So at general fusion, what we intend to do is monetize target fusion. Now, we didn't invent monetize target fusion that was developed a long time ago. So in the 70s, they were a program called Linus at the Naval Research Lab. And the idea is to put a plasma in the rotating liquid metal and then use compressed gas to push on piston and compress the plasma with the liquid. And there's a couple of very good advantage for this thing. There is a problem that in fusion in general that it'll produce a lot of nutrients for DT anyway, like a TEE try to avoid that with a different fuel. But the normal DT fuel will produce lots of nutrients that destroy the first wall. Big problem. But here the first wall is liquid metal, so it's not going to get destroyed. If the liquid metal is kind of a couple of meters thick, they will be very little flocks of the actual steel. And this is a very, very good solution to this problem. Another problem in DT is to breathe the tritium. This machine, because the plasma is completely surrounded with liquid metal, it catches all the nutrients. And if you use lead lithium, there's multiplication of the neutron by the lead, and you get a breathing ratio of 1.4. So you produce actually quite a bit more tritium than you burn, so you can get the tritium going around. Also, there's the cost. Here, there's no super constricting magnet. There's no neutral beam. There's no RF. It's very simple. It's gas pushing on piston, pushing on liquid. It's simple technology, high pressure gas, valve, piston, the stuff you get in your car today. So we think that this can be produced at relatively low costs. And it's fairly easy to extract the heat, because all the neutron will dump the heat in this liquid metal. So you pump this liquid metal, you throw it into a heat exchanger somewhere, make some steam. So this had a lot of good advantage. In the 70s, the plasma in the middle was not so good. The control of the piston for the symmetry was a bit difficult, but now we have better technology. So at General Fusion, we took this idea and we keep developing it. What we want to do here is we want to spin a drum, who the centrifugal force will put the liquid on the outside. There is a liquid shaft in the center. You open essentially a tap on the top and you drip the liquid in the center. And in this shaft, we pass current. So the current goes down the shaft and up the liquid wall. And it's a one-turn toroidal field coil. If you're a tokamak person, this will produce toroidal field. Then we inject the plasma with electricity injection. We don't produce the plasma the same way with the solenoid that a normal tokamak do that. It'll be a picture a bit later to show a bit how that works. But we essentially, we is a plasma gun on top that shoot the plasma inside. And what you form is a spherical tokamak. And then you fire the piston around it and it pushes the liquid and the liquid implode, increasing the parameter of the plasma. And at peak compression, you achieve enough parameter to make the fusion. It's a big neutron flash. It'll heat the liquid. You pump this liquid out continuously. This will be a cycle. And we run this cycle about one time per second. And this is what we want to do a general fusion in general. Here's a little video to show that. So there's a drum spinning with liquid lithium. In the drum, there's other set of piston. And the cylinder of this piston is full with fluid. So the gas get compressed by the non-rotating outside piston and the rotating piston push the liquid in. And it's a little bit of a gap of gas in between the rotating cylinder and the thing. It'll be picture a little bit later to show that a little bit more detail. So this is a bunch of parameter that the pre compression and the post compression. If you look at the pre compression, it's a couple of times 10 to the 20th density. This is, if you look at the left column, it's a pretty dense plasma, but nothing too exotic. However, if you look at the right column, which is after compression, you will be a pretty scary number there, like a 100 Tesla or something. For the normal people with the DC coil, 100 Tesla is a RSI. But here what we do is we compress the flux. So you put a certain amount of magnetic field of the order of one Tesla in the middle. And when you compress it, because the flux is conserved, the magnetic field during the implosion does not have time to penetrate the liquid metal. The plasma magnetic field goes very high. And with the very high, you can get a very high density. And with the very high density, you can get enough fusion. Beta gets quite high, near 50% beta. Spherical Tokamak can go to those sort of number, but this is due to the spherical Tokamak. You need a good aspect ratio in order to get that. And one thing that's kind of nice is the confinement required to do that. When we put in this model, what confinement we put is about five times worse than what's achieved in a good jet time Tokamak. So we don't have to invent a better confinement. The confinement that Plasma delivered today is acceptable for this scheme. And even we have some margin for that. A lot of other concept require advance in confinement, which everybody knows pretty hard to do. This system, if the confinement scale properly at the very high energy density, we don't need extra confinement. But this is a big if. We're going to push the parameter way higher than usual. So we need to test that to make sure that the confinement remain good at those very high energy density. Power plant will look like that. So you have to rebreathe the tritium. You have to extract the tritium out of the flow of liquid lead metal. And this is flowing continuously. It goes in the heat exchanger, makes some steam, the steam around the turbine, some electricity to replace the capacitor bank that make the Plasma. And what's important is a lot of recirculated energy to run the piston, but this recirculated energy comes straight from steam. We take the steam off the boiler and push the piston with steam. So that way we don't have to convert the energy into electricity, capacitor bank, you know, energy processing, RF eating, and those steps cost money, the equipment costs money and you lose efficiency. Here with the recirculated energy in the scheme is straight from steam, which allows us to have a cheaper recirculated energy in a more efficient one. Like for example here, this is a little bit of something in my way. Okay, so this is the sort of energy balance of the system. So according to little simulation, we find a fusion yield about 100 megajoule. But in order to put that, we have to put 500 megajoule of kinetic energy in the piston, which produce 100 megajoule of fusion yield. The breathing, the tritium breathing is exothermal, you get another 100 megajoule. Also the plasma, you need 60 megajoule of capacitor bank to make the plasma, but when it goes in, the plasma will disintegrate, eventually that's heat that will be done. We want to recover the energy of the piston. We want to compress this thing and you push against the magnetic field and the magnetic field push back out and you recover this energy. We reckon from calculation that we recover about 400 megajoule. So all this thermal energy, like the balance between 500 and 400 actually got in heat because of viscosity and electrically loss, like a resistance loss. So overall, all this thing goes in heat in the pot. You get about 700 megajoule of heat in the pot. You need 250 megajoule to reclaim your 100 megajoule thermal. This is the Carnot efficiency, about 40%. Spin that turbine, another 40% efficiency loss. Some of the electricity goes to turn on the pumps and the lights. Some of the electricity goes to recharge and you end up with about 100 megajoule per thing. We typically try to sell two balls running one back end. We want to run that at about one hertz. So each ball will produce about 100 megajoule. At one hertz is 100 megawatt. You have two on them. That's 200 megawatt electric. So this is the overall energy balance of this game. So this is all very nice. This is all computer simulation and PowerPoint presentation. But what have we done? So we have compressed the plasma with explosives. Here on the left, you can see how we formed the plasma. There's a flux from server on top. There's some flux installed from a coil, magnetic field. And you fire a big capacitor between the inner electrode, the outer electrode. And the force from the magnetic push and stretch the polyadol field coil, the polyadol field, like at the beginning here. And this gets pushed in the pot. It reconnects and it forms the closed flux surfaces. And that's now a spherical tokamak. Now on the right is what happened when we compressed it. Here we go. So we put high explosive around the liner. We set up the high explosive. You can see the time going on the top there. It's about 150 microsecond implosion. And you can see that as the liner come in, it compressed the plasma. And we have observed increase in plasma parameters during the implosion, which I'll show the data in a second. And then we compressed, we compressed. And we stopped essentially when the liner touched the shaft because then it sprays some metal in the plasma and the game is over at that time. So it looks like that in the field. On the left, those are the capacitor and then everything to drive the plasma. In the middle picture, you can see the flux from server. We shoot the plasma inside this flux from server. We put a big explosive. That's done in a field somewhere. You can see the nice tree over there that's in BC. So the first time when we started doing that, in the bottom right corner, those little circular target there, this is the polyadol field as a function of toradol position. We have like 16 sensor at 16 toradol direction and the different circle are different time at later and later time. So you start with a nice symmetric plasma, but as you compress it, it goes unstable and you lose it. So that's a typical problem with plasma tends to go unstable. We learn over a bunch of explosion, how to improve that. And on the right is some of our latest explosion that we did. Do you upload that and stay nice and symmetric? There's no macroscopic instability during the explosion, which was a great success. Here's the magnetic field as a function of time. All those explosions were different shot and they didn't really start at the exact same magnetic field. So we normalized to one to be able to compare that. So you can see that the first explosion, they were actually no gain in magnetic field. There is a gap here in time because we actually changed the liner geometry, but you can see that PCS 16, for example, the magnetic field went up by a factor of hate, which is pretty close to what it was supposed to do about nine considering the compression ratio before the liner hit the shaft. So we managed to compress the plasma in a stable way and the magnetic field increased. And here's the neutron count. This is neutron per second as a function of time. In sort of orange there is for non-compression shot. So when we form the plasma, there's some neutrons, but we don't hit energy to this thing. We make the plasma and it decays. So if we don't compress it, you can see that the heat goes down and the neutron rate goes down. However, if we compress it, it goes to grace thing. You can see that during the compression there's an increase in the neutron yield. If this is thermal, which you can argue that it may or may not be thermal, that would correspond to an increased temperature from 200 eV to 800 eV during the compression. That's pretty good success. Now we also work on the compression system, all those piston rattling thing. So here what you see is a servo control system. You need a very good symmetry in this game. And if you push with a bunch of piston and they're not perfectly symmetric, you will not be a good implosion. Implosion is tricky on symmetry. When it's that symmetric, it grows during the implosion. So the sort of orange dot is if the servo is off. And this is the time the piston arrived at the end of travel for different shot. And when you turn the servo on, it gives the blue dots. So you can see that this is scattered and when the time of arrival of piston goes from 217 microseconds to 7 microseconds, which according to calculation should be enough to achieve the symmetry we want. We build a little sphere like that with 14 of those pistons compressing some liquid lithium. So this was mostly engineering to know how to deal with liquid metal. Like with lots of freeze up pipe get frozen and you have to learn how to operate with this liquid metal, not that easy. And then we fire those piston, but we didn't put plasma in the middle of that thing. That was too small for that. This is a ring that we have with a bunch of driver around it. And it shows a bit this idea of having the piston in the rotor. The upper part is the rotor. And you can see the piston around the rotor those rotate. And inside those piston is liquid. So when the gas push on those piston it implode the liquid. And you can see a stator at the bottom with all the hole where the air come from and there's a bit of those valve and stuff on the backside. So the stator is fixed, the rotor rotates, there's a little bit gap in between. When you put some air in the gap it push on the rotor piston that push the liquid in. That's the system. Here's some result at peak compression. So the little red circle on the left picture is a tracking of the inner surface at peak compression. This was about 8 to 1 compression ratio I believe on this shot. On the upper right is we develop, we unroll that numerically to make it linear if you can see. And then this is delta R over R at peak compression. And you can see that the maximum deviation from around is 4.5 percent. And we reckon that this is enough error that we can do that. This didn't show up on the first shot by the way. When you start shooting those things you get the complete mess, very, very asymmetric. And then you try to understand the computer, look at all that and adjusting and eventually we achieve sufficient symmetry. The symmetry is a big deal. That's a difficult thing to do. Here's a multi-layer one. So the other one is a single layer. This is what we call it, the CWC, the cylinder cold water compressor. You can see on top of the first picture there that's the rotor. This spin, this piston in that spin with the rotor, those cylinder are full with liquid, that's water actually on the machine. Outside is all the gas valve and everything to produce the gas. There's a gap in between the stator and the rotor and the implode. But here we have many layers. So we can fire the top and bottom layer first to try to make a cup geometry, to try to trap the plasma in the cup and compress it. Because the cylindrical thing makes a cylinder and if you would just say cylindrical, the all compression, the elongation of the plasma will become too long and the plasma goes unstable. So what you need is the top and bottom to close faster than the middle in such a way that you make a cup geometry to catch the plasma. So here we project laser on the surface of the thread and we look with a video camera at those lines and we can reconstruct the shape of the implosion with that. And we compare that to the model and we achieve nice symmetric implosion with the proper cupping shape that we want and it fits with the CFD model. So we're quite confident now that we can control the surface of the liquid to the shape and symmetry that we want. Plasma injector, so we need to put a big plasma in there in the middle of this thing. So we've built this plasma injector here, which is a meter radius, like two meters a meter, it's kind of a sphere, about 10 megajoule of pulse power in there. We achieve about 300 eV in this thing, we achieve about 10 millisecond of energy confinement time. Now on the left here the graph that you see is the parallel field as a function of time. It's important to understand that in those things we put a big puff of energy to make the plasma, but we don't add energy. There's no eating, there's no nothing, there's omic eating that comes but there's no external eating or source of power. So you can see the decay of the parallel field and during the middle there between 5 and 10 millisecond is very flat. You can see we're not losing too much energy and we can calculate from the decay of energy and the temperature and the thermal energy and everything that we get about 10 millisecond and we want to compress in about 5 millisecond, that's the prediction that we want to compress in the machine. So we now know that we have enough energy confinement at the beginning anyway, that if we compress it it should go to the to the performance we want, but we do not know that as compression proceed if you can maintain this confinement. This is the temperature of the plasma from terms of scattering a different position on the left is the magnetic axis on the right is the edge. So you can see we have like 300 and a bit in the center and it's fairly flat and this is believed to be because we evaporate lithium on the wall. So we have a gathering system at the wall. So there's very little recycling and this has been shown in other Tokamak and we see that too that the temperature profile is relatively flat across the machine which is good, good thing to do. This is our energy confinement time compared to ITER 97, this is the L mode. So we kind of achieve performance similar to the L mode of ITER. So basically the way we form our Tokamak is a little exotic but the Tokamak we get is not special, it doesn't have any special, it's not even H mode, it's a fairly low and we still think that this is enough to go to good condition if we compress it. Slick, so this is a machine where we turn our explosive machine upside down and we put a pool of liquid lithium at the bottom and what we do is we fire a current in the shaft and the lithium because of magnetic field shoots up and it covered the bottom half of the machine then we fire the plasma in there and we're interested to know if the plasma is happy when there's a liquid metal. Typically we evaporate lithium on the wall but it's a solid lithium and our scheme relies on having a liquid lithium and we say maybe the liquid lithium will kill the plasma which would be bad. So we do this experiment here when the shaft current we see the pool of liquid lithium at the bottom here and when we shoot the current this lithium goes up and we let it go up to about the equator because if it goes above the equator then it goes on the ports and then you can't see what's going on and all your sensor and diagnostic are shot so we don't want to, we'd be nice to cover the whole thing with lithium but that will also cover all the diagnostics so we presently we don't want to do that. So in dark green is what we achieve if we have a evaporated solid lithium on the wall and in pale green is the liquid lithium so this is just a magnetic field as a function of time but you can see that the parameter of the plasma is actually a little better when you when you put liquid lithium and other people have observed that it is the lithium token back at Princeton they get good performance of that. So lots of people are lithium addict they think that lithium is good for the plasma which is good because we want to implode lithium on the machine. A quick diagnostic this is for the next machine we want to build the big compressor and we have an array of diagnostic it's not as big as TAE but we have something that we think will be able to measure what's going on it's a bit difficult in this situation because you cannot put anything on the outside like the liquid wall will block you so all the diagnostic have to be inside and that tick compression you will only have the shaft in the center so for this next machine we want to build a shaft will not be liquid because we have to put sensor in it the shaft will get broken at a high energy shot so we have to put a different shaft with a different set diagnostic and implode it in a power plant we would like a liquid shaft but then it would be very little diagnostic access would be very difficult. For our next machine we will put a solid shaft with diagnostic in it. We also make some simulation and some computer stuff so the first thing we do is we call it the integrated model it includes the valve and the air and the piston and the liquid and there's a relatively simple liquid model in there and this runs very fast and it like it takes you know 30 seconds to run or something so you can optimize how big should the piston be what the pressure should be and things like that and when you have one that you think you have descent then you do some CFD we have open form which we modified a bit to include the magnetic force in it and then that can simulate the exact but that takes longer to run this is probably a day run to run those things we have an a combustion with quick merry on that then we do a lot of mhd we use a computer called vac we modified it to be able to deal with the surface that change most mhd code are for a fixed box when you start compressing the box they travel with that so then the energy confinement time is a calculate with tjaro cjaro from a general atomic and then we have plasma stability with decon this is per instance doing those calculation what you see in a little graph is as a function of compression with different characteristic and what you need it's you need to stay in the blue all the way so where it says you know back trajectory it stays in the blue it means stable but you have to be a bit careful because if you're not at the right place during the implosion you're going to go unstable but we we've found that we can actually find trajectory and initial plasma thing in the shape during the compression that stays stable during the implosion holding to decon and then we do plasma wall interaction especially near the center the plasma gets very hot there's a lot of energy the wall is predicted to evaporate the the transport of this vapor metal into the plasma is very important i would i call it a smoke problem we don't want to smoke out the plasma so the the wall will evaporate and you have to try to calculate what the transport of this smoke going in so far it looks okay but this is something that we're still in progress right now that's we're doing that with oak ridge and finally the the structure of this machine takes quite a beating it's very high pressure gas in there so we do dynamic fea pretty standard and generating thing to make sure that we won't bend anything when we shoot so the for nothing we do that with all bunch of existing code and we kind of loosely couple them on a time step to time step but uh i'll come up plasma our computer guy said that it's a bit Mickey mouse and it would be nice to have a single code that have all those things included in one thing so we want to develop this code called pulsar and it will have all those bits but in a single grid because sometimes you have to change from one grid to the other cost some issues because that big compression we think this is going to become pretty couple like the way the fluid move with the plasma move it's all going to be very couple so this loosely couple different code that we do like for example the cfd for the liquid and the mhd for the for the plasma if we can put that in a single code would be better all right so now what are we going to do next so this is stuff we're doing right now if uh if we get the financing for it that we're trying to do right now we want to build this machine so you see the little man on the bottom right corner it's a pretty big machine uh this is the same idea this is going to be a spinning rotor with piston but we scale that pretty big and uh so here's a simulation of of the shape this is what i was you see you see you fire the top and the bottom liquid in red and makes a cupping geometry that trap the the plasma in the center so that's what we we want to do in this machine this is the iron temperature prediction so it'll go to about 10 kv there's the current this doesn't show the the density but this machine should start at about one time to 20 and end up at about you know one time to the 23 density that's big density at 10 kv so this is actually this should produce about 10 percent of loss on if the confinement happened in the middle the big it the big uncertainty in this machine and in a general fusion approach is what will the confinement be at peak compression we do those little scaling law and everything and it's okay but nobody have measured plasma those very high density there's quite a bit of of risk there and that's what we want to remove with this machine here and this is the the magnetic field you can see we have open field line uh on the outside the symmetry axis at the bottom here and we compress that this way so it's take up like we we have the the outside compressing the thing and keeping it in the middle so this is uh the end of my presentation I think let's see if we can get out of here okay thank you very much thank you very much I guess I will direct the first question to you uh Dr. Michelle LeBerge and the first question would be if you're using lead lithium as a compression medium don't you have to worry about the lead getting into the plasma yes this is a big concern we are trying to calculate and and when we do the experiment we'll have some idea of how much transport of the metal goes in the middle lead of course is terrible lithium is better I'm scared of lead anyway we haven't finished the calculation so we don't know because the vapor get ionized very quickly and then it's trapped by the magnetic field and then it's you know magnetic transport going on if you think you're beating the heat transport which is the the stuff going out you will also beat the stuff coming in but we need to show that if it becomes to be too much of a problem we have this aspect idea of floating a sin liquid of lithium on the surface of the lead lithium we need the lead lithium for the inertia to keep the thing together longer in the middle to achieve something so we need the density of it but we could have the surface being pure lithium to avoid this problem but we don't really want to go there because it had complexity to the rig but uh if it turns out that lead is too much which we don't know yet we would do that okay thank you my next question to you uh our term would be what the term would to be to field reverse configuration efficiencies and circulating power fractions are assumed for the TAE reactor can you guys hear me yes yes sorry I had a little problem here with my audio um recirculating reactor powers well it's it's quite high um I was just checking the numbers here it's uh it's about 30% or so I mean to make these I was following the chat threads here no to make these a viable reactor you you of course would have to uh harvest uh brimstrahlan right you do you do want to harvest x-rays absolutely you cannot reject that um and and with that the there is circulating power would be quite high but um we can make our beams very efficient we in fact build reactor prototypes which uh reactor prototype uh level beams uh at one mega volt uh level injected energy which can be 85 to 90 percent efficient so so the overall balance works fine uh but of course assuming assuming the confinement is is in place to support it okay thank you my next question to michelle laberge would be has general fusion demonstrated stable reversible implosion of quasi-spherical linear compression of a target example a trapped gas uh yes we we do shoot the liquid and when it comes in the center it reflect actually if you look carefully those little circles that I showed during my presentation you can see the after the implosion the circle go back up on those machines we don't have the rig to recover the energy we don't have like this the fancy valve and stuff to to recharge so really what they do is they go in and out and in and out and in and out and bounce a couple of time dies out so we don't have the equipment to recover the energy but we certainly see that the liquid rebound now we haven't got plasma or magnetic field in there so what it rebounds on is we can put some gas to rebounds on gas but mostly it rebound on the on the rotation in order for this thing to be RT unstable you need the rotation centrifugal acceleration to be higher than the radial acceleration like this so in all those experiments if you want to go a good experiment you have to spin it fast enough compared to the implosion velocity so it bounce off the we call it bouncing off the rotation that's that's what bonds are now sometimes we put gas in there but if you put too much gas so it because it starts to rebound on gas more than on the rotation then it goes RT unstable then you see all those little finger and stuff going on so uh yes we've seen reflection but we have not recovered the energy because we have not built a gear to do so okay thank you my next question is to autumn as you apply more and more neutral beam injection past particles do you see any kinetic alpha instabilities yes we believe we do see it and thanks thanks for the question Glenn um it's uh we're trying to understand what exactly we're seeing now we're running some um um sim simulations which which is a peak road you know there's many of you may know and it seems that the alphanic alphanic modes and alphanic driven reconnection should be there now we're just trying to formulate experimental hypotheses on what to test for in the experiment but what we see doesn't seem to lead to any performance you know significant performance deterioration so maybe we're not there yet but uh yeah we're exploring stay tuned thank you my next question to michel laberge is um how are you handling the repeated application of very high magnetic fields on the center conductor of your talking back like arrangement yes so right now for the machine that we want to build the pressure will be so high that it will destroy the shaft so we have to put a new one on each explosion not very good for power plant power plants be a liquid when the maximum compression happened the shaft will go unstable and everything we calculate that the gross rate of stability is good enough compared to the implosion time but uh after the shot it was key like you know we expand but the shaft would keep beginning your shaft will disappear it'll get wiped out but you have a tap here finding a vertical jet of liquid metal so all this twisty shaft will fall off the bottom of the pot and a new shaft will be created and we want to do that at once again time rate so basically people from the question they're worried that the magnetic field will mess the shaft well it will but then we'll put a new one be it liquid or be it solid solid is a pain we have to go slow and put a new one but in the power plant we want a liquid one that will get wiped out every shot but it will reconstitute as a as the jet keep pushing down okay thank you my next question would be uh back to our term is what is keeping the electron temperature so low compared to the total temperature well it's kept at whatever level it's finding self consistently i guess a generic answer to to the question if we're bringing power in it finds a balance so the power in is counterbalanced by losses right so um i wouldn't i wouldn't necessarily say that it keeps it that low one kv is a very respectable temperature for mirror device with the power with the heating capacity that we have we inject a couple of megawatts of power which mostly goes into the electron channel and our losses are mostly convective as well they actually pretty well plugged we've done a great job with our expander dieter so we measured that the the energy loss is about six to seven temperatures which is very close to the ideal limit so all of that self consistently it establishes the electron temperature approaching kv so that's my long answer to how the electron power balance work okay back to michel laberge following up from the from your previous answer how would such instabilities in the shaft affect the plasma itself do you worry about that yes not only the shaft but also the outside like the outside liner will develop some asymmetry the shaft will start to bend so so as you compress the whole thing will start to be a little asymmetric we're we're trying to tackle that with simulation right now and i would have to say that we haven't gotten an answer right now because most of our code are cylindrical symmetry so now you run that in 3d to try to figure out what what level of unroundness will affect the transport like how much transport do you lose when the things start to be asymmetric because the shaft start to bend or the outside shaft and we don't have an answer for that exactly we're working on this like typically in the tokamak people they are very careful with their magnetic field they think that they will tend to minus three magnetic field error start to increase but that's an external magnetic field here's the liner so the plasma will always adjust a little bit to the liner because it's reflecting current in the liner so it's not totally clear to us at this point what's the maximum error that we can accept and that won't you know what the effect of a certain amount of error on the transport we're working on that numerically and when we start shooting this machine we will get some answer but at this point we don't know that exactly okay thank you so back to Arton have you been able to raise the magnetic field from 0.1 tesla to higher values in the CTW reactor yeah yeah we used to operate actually below kilo gauss we used to be somewhere 750 kilo gauss and we basically doubled it we can go to about 1.5 now which is b squared four times what it used to be so yeah to that level yeah we can okay my next question is without the particle deposition the effective beta of the plasma target increases what happens to the plasma stability for reactor level nuclear gains that's the michel abhash yes so that's a good question when when we want to produce those yield enough to show the little diagram that I showed at the beginning included in there is alpha heating and the alpha heating is a sizable amount of the increase in thing however when the alpha start eating we over beta like we will go over the maximum beta then the then the plasma can take in experiment in normal tokamak there's two ways that this fail some people try to put some more heating in there and the confinement degrade and the extra heat escape that would be okay with us basically it would make a clip like at 50 beta any other energy would just go away to the wall and that'd be acceptable however sometime when you over beta it goes unstable and you lose the configuration loop that will not be acceptable so at this point it is not clear to us when we over beta will it go unstable and we lose it or will it graciously the confinement will decrease and the extra heat will escape which we would like if the situation and it goes unstable we could considerably put some impurity in it some AIZ impurity that when the temperature increase they radiate more and you control your peak beta with radiation like if the heat goes too high then you radiate more and then it's stabilized so you might want to get rid of the extra heat with radiation by putting impurity or the extra heat will escape by increased transport naturally from the machine but if the case is that when you over beta goes unstable and that would not be good okay thank you my next question is regarding the neutral beam injector how much recirculating power as a fraction do you think you would have in the reactor and what efficiency does this require since traditional efficiency is usually no more than a third and more sort of developments will be required in the nvi for future reactors okay well i guess we touched on that briefly before but we have been researching and developing beams for 15 years now both positive ion based and negative ion based and as i mentioned before we prototype mega-volve level beams with unpaired level currents for our future pv-level reactor so we're very comfortable with the approach we're taking as far as beam efficiency is concerned well to support the practical reactor you of course would like to do better than the gas neutralizer for sure and there's a couple of options there is a plasma neutralizer which can be up to about 85 percent efficient demonstrated experimentally or you could be if you are ambitious and could do even want to do even better you could do a photon neutralizer which is a very cool process where you can detach a loosely bound electron to the h-minus accelerated ion with light and by the way there is no there's no opposing process unlike gas and plasma so not you know photons don't see the electrons back so it's very efficient and you could actually get over 90 percent efficiency which again in smaller laboratory setups we demonstrated so we're very comfortable that we can do at least 80-85 percent efficient beams okay at mega-volve level well thank you Atom you also mentioned in your presentation about spin-off technologies and that TAE has been developing so if you'd like to touch upon them and then we might move on to Michelle afterwards if he also has any spin-off technologies that general fusion's developing sure I'd love to oh they're generous let me just bring up my slides back if that's okay no problem there is all right so we are on beyond fusion and spin-off technologies right and I can so our first spin-off was with our accelerator technology which leveraged all that development work that I just commented on in the negative ion beam space in particular and we developed a an electrostatic beam which uses tandem tandem accelerator platform at about well of course this is a very different beam than what we use in fusion this is higher energy about two over two megavolves total energy with just about 10 to 15 milliamps of current so much less current higher energy to produce to develop a neutron source produce neutrons by converting protons beams into neutrons for the cancer treatment modality which is called boron neutron capture therapy it's a unique cancer treatment technology which is a combined therapy where the therapeutic effect is produced by the drug and the the neutron radiation you can think of it as neutrons triggering the the action of the drug which of course just boron 10 capturing neutrons and sending an alpha particle out which kills that cancer cell but not the healthy cell next door due to the vector action of the drug that pumps boron 10 preferentially into the cancer cells so it's a very efficient way of killing cancer very highly selective the the other way to think of this is a it's like a biomedically targeted radiation therapy as opposed to physically targeted radiation therapy like x-rays or protons which will have physical limits on accuracy this now has biomedical limits cellular limits of accuracy we started treating patients late late last year we so far treated 14 patients and with pretty great clinical results so far of course the observation period is not very long yet but the immediate response of those patients is terrific and we're growing the water book for CHE life sciences in US Europe and Asia so that's that and our more recent spin-off it had to do like any invention that grew out of necessity we for those of you who may have visited us or know our our facility here we operate in a very inconspicuous commercial campus in Southern California the energy feed the power feed into our building is very limited just a couple of megawatts whereas the experiment takes many hundreds of megawatts in that short pulse of 40 milliseconds or so so we've had to develop technology to store energy on site capacitively and then very efficiently on a millisecond timescale channel it in and out of our capacitors and batteries and very efficiently managing that power those power flows into all of our electric electromagnetic loads and once that was developed with all the power electronic circuitry and topology and control networks and algorithms then we realized that it has the same technology because it's very modular and very scalable it has applications for example in electric vehicles turns out both on board with the power train and on the charging side and can produce very meaningful improvements to the vehicle range longevity of parts operating characteristics of the vehicle and also this can go into a residential and commercial energy storage applications which is we've we've further developed these now outside of our fusion experiments and spun out key power solutions the name of the company and it's addressing a trillion dollar size market right now with all the EV and storage applications that's that's the site application story for us not so I guess so I guess I'll give the floor to Michelle if he has any similar spin off or comment in that regard okay so a general fusion making spin off of different piece of technology was often discussed and I personally resisted quite strongly doing the spin off because in my opinion it will distract us from doing fusion I just started general fusion because I wanted to you know deliver fusion to humankind so there will be a problem there so in my opinion if we start building spin off like some fraction of the personnel involved with the spin off we'll have to go with a spin off too so I thought it would be a distraction to do so now the investor like it because you know in case your fusion doesn't work you still have something so there's some discussion that perhaps we should have done fusion spin off but so far general fusion have not and I have been resisting I think we should do fusion that's the idea well thank you I guess we have a few more minutes so I'll find some more questions in the chat I guess one more question to Artem would be regarding earlier regarding the s s star parameter and the question is is that based on the beam ions or or some average ion energy the that plot that I that I shared before that's s star for the for the thermal plant the whole point is that you know if you have no fast ions then s star over e as conventionally defined would generally you'd operate over s star over e less than three or just about three well fast ions break the conventional limit and definition so yeah it's the conventionally defined s star okay thank you very much given this let me check I have one more question for general fusion regarding the plasma aspect ratios and simulations and whether that increases during compression or does it remain constant throughout as this would affect the high beta attributes it changes like we try to maintain it as much as possible but due to the maximum copiness that we can put in the fluid if we do too much it starts to produce jet at the equator actually so we cannot cop too much so within the limit of what the fluid can take the aspect ratio increase I can't remember the number but go say from 1.5 to 2 or something like that for the k expected elongation factor uh this stays stable we do check with decon that the plasma stays stable it affects the transport when you put those different at a k into the little equation like the gyro code it changed it but all that is included in our in our simulation so it's it appeared that it should still work yes it affects beta beta change with the elongation all that is is all in the system and it works in the computer yeah remain to be seen if it'll work when we actually shoot but yes the elongation change but it's included in the calculation it should still work okay thank you very much uh one more question just showed up so general fusion has announced that reactor project is being built in the uk now which is is that correct and the follow-up question is is ta systems considering building something outside of the us or is it currently confined geographic to geographically where you are now the the plan is to uk but it's subject of financing we're presently in financing mode trying to finance this big machine so no money show up no no machine no money no machine big money big machine now to autumn well i i i fully second michelle statement big money big machine for money don't forget so that that's that's generally true uh no look we are ta is is is exploring a few a few opportunities outside of the us we are we're having our dialogues open with various parties uh it's it's it's a huge commitment to the company and of course we would need to scale accordingly we are fully committed to the tournament because it is being built here now if we find a and we would be open we are open to the ideas of uh accelerating the path to practical fusion and uh provided that there are committed and the motivated partners and financing can be showed up then we can certainly support parallel activities or staggered activities where maybe we can accelerate the vinci uh while kapturnicus is still operating so yeah so those conversations are taking oh thank you thank you to both of you panelists for contributing so you know contributing your time contributing your time and knowledge to this presentation and and again answering all the questions within the comments regarding your presentation i guess uh we could take away that you know we hope we hope to see you know more results in the future and i'm sure to the audience we encourage them to follow along their respective websites and more information in the future from both of the companies so thank you very much to both speakers and thank you for answering the questions and we hope to see you the audience in future episodes thank you very much thank you very much appreciate the opportunity and thanks for inviting us some questions thank you very much cheers right