Speaker 0:  ai data centers are buying gpus by the warehouse, but all the silicon is useless if the machines cannot agree on what time it is. a montreal startup just raised $55,000,000 to replace centro's quartz timing with microscopic silicon clocks built for the ai era. so strathtera closed and oversubscribed 55,000,000,000 series b, bringing its total funding to us $75,000,000. the company will mass produce its gen two silicon timing products and develop gen three components for ai dentists data centers with initial customer samples targeted for 2028. so canada's building the new ai clock, and the ai infrastructure is becoming a coordination problem, not merely a gpu counting contest. the better timing could make networking, optical links, and distributed computing more stable as clusters grow to tens of thousands of processors. what do you think about this, john?

Speaker 1:  yeah. i love this. i love that this is a canadian company taking a different approach than trying to compete with nvidia because they're not competing with nvidia. they're enabling nvidia. and so, many people assume that every semiconductor startup wants to build gpus. and so stathera, i think, is pursuing a different part of the semiconductor stack, like you said. so its products are mems timing devices, which i think we've covered in the past. we've covered so much. it provides an ultra precise clock signal that synchronize digital electronics. so think of it like an orchestra. without a conductor, every musician would slowly drift apart. and computers have the same problem. every computer has, whether they have a crystal in it before, i don't know if they still do, but, a clock. so every processor, operates according to a clock. so if the clock drifts even by incredibly small amounts, then you get performance degradation. your errors increase, and communication between trips become unreliable. and so modern ai clusters, they contain gpus, cpus, modern asics, memory controllers, optical interconnects, and every one of these components depends on precise timing. so as ai clusters scale to tens of thousands of accelerators, maintaining that synchronization is gonna become increasingly difficult and increasingly valuable. so i guess people are asking if you don't follow us, what is mems timing? so, it's microelectrical mechanical systems. and so mems combines microscopic mechanical structures with semiconductor electronics. so traditional timing devices, they often rely on cool. it was quartz crystal. that's right. yeah. so quartz oscillators are extremely accurate, but they have limitations. they're larger in size. they have a greater sensitivity to shock, actogen complexities, and manufacturing constraints. whereas mems resonators are fabricated using semiconductor manufacturing techniques. and so they have a smaller footprint. they have a better integration. they have higher robustness. they have lower power consumption, and they're easier for mass manufacturing. so i think this is a smart move. what are your thoughts on this? do you, on this company and on the investment and on the technology?

Speaker 0:  i think it's really interesting. you know, there's always a race for better timers as we know. we've talked about the different types of atomic clocks and whatnot. and as we get into these clusters and ai and global networks, it's more and more important. and the gen three platform is where the ai data center story really begins. the next generation is being architected from the ground up for ai workloads, enterprise communications, and hyperscale data center applications. initial customer sampling for this platform is targeted around 2028. so, basically, when you build something like colossus with a 100,000 gpus, you need systems like this that could control who can talk when and, like, in what order things are happening. otherwise, things are gonna get messed up. and, as we build more and more of these data centers, we really need i think this is great. i think, we've got a lot of very interesting lead investors here. bdc capital, celestia capital, mediatek innovation fund, txc corporation, ultratech capital partners, seiko, epsilon. the question is, i didn't see what the competition is like. i like that this is canada. i like what we're doing, but what competition is out there? and are we poised to take this market, or are we, just another player?

Speaker 1:  yeah. what i liked i don't know if you caught this in the article when you're reading it, but, oh, you have it there. his his decision to stay in canada. i love that.

Speaker 0:  yeah. the founder, george zeres, has received direct offers to relocate his company in united states and he's turned them down. this is not a trivial thing. canada's a well documented pattern of commenters a commentator sometimes calling, branch plant problem in tech. promising startups get acquired by american firms or the founders themselves move to the us because that is where the capital is. the customers and the talent concentration exists. so it's not a moral failure. it's a rational economic decision. he describes building a chip company in canada as more painful than doing in the us. he has been direct about the structural challenges securing later stage funding is harder, recruiting executive talent and semiconductor experience is harder, and the proximity to major customers requires deliberate effort rather than being a natural side effect of geography. so, yeah, it's great to see, canadians staying in canada and keeping the tech and whatnot here.

Speaker 1:  and so it just because he's in canada, doesn't mean everything's happening in canada. right? so, they're taking an interesting manufacturing approach. so the mems component is manufactured in canada. there's other electronic components that are sourced internationally. assembly and packaging is performed primarily in asia, and sales are focused on silicon valley, obviously. so i think his approach is a smart approach. if he wants to stay in canada, that's probably one of the best ways to do it, is not have everything happening in canada. do it where it makes the most sense.

Speaker 0:  and i mentioned we didn't have much on the competitors, but we do have one on systime, which is, the current market leader in silicon mems timing. it's a publicly traded california based company. has been over a decade establishing itself as a go to alternative for quantum, oscillators at precision timing. so they're taking them on, head on saying they're better. so we'll have to watch this carefully and see what's happening.

Speaker 1:  yeah. and the other thing that i learned while doing some research is that canada is quietly becoming an ai superconductor hub. so we've got, photonic, which does quantum computing. we've got thales, which does ai processors. we've got tenstorrent, which does ai chips. and we have spark microsystems, which does a ultra low power wireless chips. so i didn't know that, even though we're in tech. jesus. i can't believe i missed all that. that is awesome.

Speaker 0:  there's a lot more chip manufacturers or, maybe not manufacturers because they all use the same foundations that build the chips. but the, the peep like groc, g r o g r o k. k. yeah.

Speaker 1:  q. q, i mean. no. q.

Speaker 0:  yeah. the canadian, ai chip manufacturer. you know? that there's a bunch out there. and when we, were doing blockchain stuff a few years back, there was a bunch of different chip manufacturers that were building asics with the blockchain algorithm and the different algorithms right on the chip to do. so there's a lot of interesting chip stuff happening in canada when you start looking into it. yeah. so we'll have to see where this goes. so spracera has not fixed ai data centers yet, but has raised the serious money to rebuild one of the smallest and most fundamental components. if canada develops the technology, but most of the investment and customers remain elsewhere, is the canadian semiconductor sovereignty or just canadian r and d feeding a global supply chain? let us know what your thoughts are below and whether you're gonna check them out. we'd love to hear from you.