Very interesting video! Thanks for sharing. Here is the video transcript (not perfect, but I think there is food for thought. Sorry for the poor editing):

and thank you for joining us today and uh could you please give us some of your thoughts and highlights of the major technology trends at OFC24 yeah I think um most of um uh what's going to be shown is going to be around uh 800 gig and the transition and what's going to be needed for 1.6 terabits in the future I think that um particularly with the um explosion of uh uh artificial intelligence what what we're seeing is and I think this is going to be very key for the industry as a whole is how's the supply chain going to cope with the volume of uh Optical interconnects that's going to be needed I mean there there obviously is a lot of hype on artificial intelligence but latest market research shows that you know the number of ports in the next few years of 800 gig and above is going to be um uh huge and that that really is driven by the fact that artificial intelligence um bandwidth is going to increase um 10 times in the next two years um from a from a volume perspective as well uh and if you just look at the frontend data center which we all know and have been dealing with for the last um 10 15 years that's only growing at uh two times uh in in the next two years according to latest Market um analyst reports so from that perspective um it's is the supply chain robust enough and that's some of the work that we've been doing we made an announcement yesterday of a strategic partnership with jabble and and really that's around uh being able to um bring silicon photonics really to the mainstream as a manufacturable product and uh jable obviously has a lot of experience in in high volume manufacturing from components all the way through to system level type of products and so that's one of the ways that we're trying to uh support the industry not only from a uh a data Comm and AI point of view but across other market segments as well and and for us really the the main aim is that when we deliver Wafers from our Foundry partner Tower is what the customers do with those Wafers at that point and where ja's expertise will come in is they will take uh the Wafers and they will uh put them into Optical sub assemblies depending what the customer wants and they'll be able to do that at scale and where we help we bring in our test and our reliability experience uh from uh the optical uh integration point of view all right so um in the transition to 400 gig uh things maybe didn't go as smoothly as some of the pundits had predicted at the beginning uh do you see the 800 gig transition coming along little f a little bit better yeah I think that 400 gig particularly because it was 100 Gig electrical interfaces um there was a lot of um issues around um how the system interfaced everything together which meant the DSP the Optics and and and the high-end chips that sit behind it and a lot of work had to go into that uh from a signal Integrity point of view from a firmware point of view to enable uh optically compliant Links at 400 gig I think what you've seen and what we're seeing is that that from my experience that took uh some of the bigger companies one to two years to get that all uh resolved I think what's happening is uh 800 gig is starting earlier because um because of the same issues - POET is currently at 800g - because you've doubled the number of channels basically at 100 Gig but I think as we go to 200 gig people are starting that early because the way that all this interoperates with each other at 200 100 Gig that's going to be an even bigger challenge so I think that's why you're seeing a lot of hype on 1.6 right now because that work has to be done and that will take time before you can start deploying systems in in in mass volume and how about with the lasers emls or vixel (vcsel) um so from a vixel point of view I think the 200 gig uh vixel are a very uh big Tech technology challenge I don't I don't think anybody would deny that if you talk to most of the vixel manufacturers I think emls um yep there are some very good emls out there the issue is is cant emls at 200 gig meet the the requirements from a volume perspective and then if you look at the way the industry has traditionally built uh pluggable modules I think what you'll see is that um you know the yields are going to be low initially which then puts even extra pressure on are there enough emls in the world - POET can mix laser sources - so I think that um you know if you talk to some of the bigger players particularly in the AI space they'll tell you that one of the concerns they have is can they get a diverse supply chain which would enable um you know them to get the volume that they have and you know from discussions that we've been having and what we're hearing in the industry 800 gig is a challenge to get the volume that you need and that's where coming in with some different Technologies using different architectures and things like that could enable you to diversify that supply chain somewhat as well okay and uh one more question on the supply side then uh of course there's a lot of government funding coming along um is that going to impact this segment of the market in in any you know significant way in the next few years um I I think so I think most of the foundaries a lot of companies have gone in to go particularly from the US government's chips sacks um I think this is going to be a key technology in the future whether or not you put heterogeneous integration in or not um and I think that um what we're going to see as we go on is that uh more and more foundaries will come on to to look at uh deploying uh more silicon photonics I I think the the other thing that is also a very major Trend in the industry and there was um there was an announcement I think overnight that uh you know the uh lpo MSA has been formed and I think that's also going to be a very Hot Topic at ofc - Remember, MSA's founder mentioned that POET Technologies was one of the early adopters? LINK - this year it was last year I think it'll be hotter now um um and mainly because there's a lot of work that has to be done to enable interoperability uh between different manufacturers models uh modules sorry and and and that the reason for that is is you're dealing with a completely different electrical interface uh coming out of the plugable uh because it's going to be more analog than digital because they've removed components from the module and and you know we we did that at 10 G and that hard problem to solve so you can imagine what it's going to be like when you're dealing with 100 gigabit type and 200 gigb type interfaces so that work got to be done and it's going to take a little while to do do you know it's that on the receive side or on the transmit side or both it's going to be on both yeah okay interesting one one final question and that's 1.6 T um any any uh observations on where we're headed next with that um I think that what you're going to see is 200 gig Lanes um uh so you'll see that most of the merchant IC suppliers who Supply dsps and um and the analog Electronics will all be dealing with the transition from 100 Gig iio to 200 gig I think you have to go in that direction to minimize the amount of uh obviously Optics that are in in in a pluggable module and to be able to fit in the form factors that are available today - POET is selling OSFP and QSFP-DD solutions - so I think you'll see a lot of that at the show particularly from the merchant suppliers I mean we're doing demos at the show with synopsis with their 200 gig uh cies in fact it's like an lpo type demonstration uh using our our um 200 gig EML uh or sorry our 200 gig EA modulator that's inside uh a silicon pick and that's married up uh to uh the synopsis 224 gabit uh seris so that'll be on their Booth being demoed and then I think that um what we'll see is that uh there'll be a lot more drive to that 200 gig interface as we as we go through time and you can imagine what we're going to have to be doing when we go to the next speed above 1.6 terabit all right well fantastic looking forward to seeing you at at ofc 2024

Just a reminder that POET is transitioning from DML to EML, as stated in the POET Technologies' Letter to Shareholders issued on December 5, 2023.

The development included an overhaul of the interposer architecture to expand platform compatibility to Externally Modulated Lasers (EMLs). Over 90% of 800G modules in production today use EML lasers and we expect the expansion of the platform capability to EMLs to have wide appeal with a number of module manufacturers who have already committed to that path.

he Infinity family of engines enables the use multiple chiplets engineered to operate in combination and within the size constraints of current pluggable module designs. We believe that providing chip-scale integration capability to EMLs will be an unparalleled achievement in the industry and would set us apart from even the largest established competitors in the market. LINK