NETINT Technologies about Building Localized OTT Networks

NETINT Technologies about Building Localized OTT Networks

Show Video

today I am joined by Steph vanderziel who is the founder of Jetstream and let me um say this about jet stream if you haven't heard of them or you're not familiar with the company um they they have some really interesting numbers that will probably pique your attention and some of these I know uh Steph is going to be talking to us about today but imagine this 40 percent TCO reduction is what they're able to give their their clients users of their network of their platform they have a hundred percent uptime they have been gdpr audited which is very significant especially if you're streaming and delivering content in Europe um there are 430 percent faster now this one even I had to raise my eyebrow a little bit and go hmm let's see how they do that so I know Steph's going to get into the details about how the system is built and uh and and what that actually means they support 8K that's pretty unique and they have integrated with nine cdns and so that's a little overview of jet stream and Steph welcome to the voices of video hey Mark thanks uh thanks for having me on your show really cool so I have to ask you a question did you really invent streaming as you claim tell us the story that's the claim yeah so we started producing our first live stream in 1994 so that's quite early in the internet days as well most people did not even know about the internet and we were doing live streaming and it wasn't full HD or HK or f4k it was a small postage size it was one frame per second but it was live and we don't know anyone else who was before us so yeah we we were the first as we hope to be able to claim to get give us a short overview of the company how did you jump from you know 1994 to uh 2023 and uh then you know we'll get into talking about you know I know a very interesting use case of course 1994 it wasn't the company we started the company I think in 2003 after pioneering for many years with live streaming we broke down the internet here in the Netherlands by having so many viewers so we overloaded the internet and so in 1997 we built our first CDN which was a temporarily CDN uh to do a large-scale life events and we broke them down after the event was over and then for the next event we built up a new CDN that's how we did it back then uh so we were first in production we were really going with cameras and encoders to locations doing the entire streaming but at one point I said that's not really scalable let's build a streaming platform we have so much knowledge and experience now let's build a streaming platform as a permanent platform room and that's what we did and I think the company grew 1800 in the first five years so it was you know boom streaming became hot when people got Broadband basically that's how it started and then we started to develop our own software because we were running you know Windows media real ice cast and QuickTime streaming servers and manually helping customers but because of the size we needed to develop something that automated provisioning customer provisioning live stream provisioning so that's what we developed a piece of software we called it Video Exchange because we could exchange videos with customers I don't work for orchestration stuff and Edge service deployment and all that stuff and it was really cool technology and we started to license it out to telcos so they could build their own cdns but at the a few after a few years we decided to go SAS all the way just host our software um uh and and scale it up and where we are today we run our own cloud it's called Jetson cloud of course and yes it's 430 faster we we tested uh customer streams uh full HD streams over multiple clouds and cdns against our own cloud and streams uh do you know the video chunks of when you know the one two or four seconds for instance yes the the download time from our cloud is uh more than 400 faster compared to regular clouds because we're not a generic Cloud we are an optimized streaming Cloud so we don't use virtualization the network package is optimized we have Hardware software integration with the caches so everything is done for tuning to to be able to burst out those video chunks as fast as possible at one point you you can ask what's the point of you know delivering video chunks at 50 speed instead of 25 speed or five speed uh but once you go into the 8K territory that's going to make sense you want to be able to burst out the very large pieces of video to those users um and another benefit is that uh when you can burst out those video chunks at the highest bit rate to the users faster the video players will analyze how fast they get get the chunks in so the chances are higher that you go to the highest bit rate in the ladder compared to other clouds and cdns I want to get into a use case that I know that you're Building Solutions around and and deploying that that I think is pretty fascinating and it's this whole idea of a localized Ott Network you know when I first heard the phrase from you and of course we're collaborating on on a couple projects I drew my own idea of what it actually meant uh but it turns out there's quite an interesting application so you know first of all tell us what is a localized Ott Network and you know why does it exist it's it's funny because you know when you talk about about Ott you talk about world domination right we have Netflix and HBO world and we reach every piece of the continent and that's what we try to help our customers as well because we know we have all these cdns on board so we can deliver into every every continent and if one CDN doesn't perform we can switch over but there is there are some gaps there I mean this is about global streaming but there's another market and that's localized markets you know what about real local remote areas like compounds in the desert um you know you can have like a thousand or two two thousand households in a location but the back hall connection can be really poor but still those people want to be able to watch a decent Ott stream not just that use case it's also offices hospitals uh you know hotels uh holiday Parks University dorms you know people want to watch television on their cell phone or iPad but if you know if 10 people are doing that that's fine for the local network but what if uh thousands are doing this and what if the background isn't really that strong to even get you know like 80 quality high quality full HD streams in so then you need local information local transcoding and local serving and that's what we've developed we call this deep edge Ott it's an it's a it's an appliance and we build it together with your technology with netint and it's a one reg unit uh so you don't have to buy like five huge servers anymore with CPU encoding it's a hardware accelerators uh solution with our software and it's actually it's got three components you know it's it's a live encoder it's a transfer to create each lesson Dash streams and it's a deep batch cache so it's a local cache so you can actually use it also to serve out the the streams to the local users you don't need an extra cache or cdna um and the challenge in those locations is that it's not just a remote location with a poor internet connection most of the time there is not that much local physical capacity and power as well so we needed a solution uh we actually had a customer coming to us and they said we we have these compounds and gated communities in our in our countries and we need to have local ID TV there but hundreds of stations per location um and if you talk about hundreds of streams you need to convert on a CPU basis you talk about deep investments in in hardware and CPU Powers right but also energy consumption and um the solution we created can convert 80 Live HD channels and just with this one new chance which is pretty cool it's like 20 faster uh a more efficient compared to CPU encoding and it can serve up to 10K you know 10 000 concurrent viewers with just this machine which is pretty cool and I think one of the most important USBC savings and you know we've been talking to these customers and they said yeah but if we need to buy this this Hardware all these servers of the CPUs there are no business case we will spend like 80k per location and the energy bill will be higher after so many years yeah that's right yeah I mean with the current energy price so that's really a challenge I think after three or five years you're spending more on energy than on physical encoder uh that's a that's a problem so the cool thing about this solution is that I'm just reading it 72 lower capex so per stream you spend 72 percent less on investment for encoding the death stream and that's just a hardware we're not we did not even take licensing you know software and go to licensing into into this equation that is using 89 less power compared to cpu-based encoding which is also really impressive wow and the physical footprint is just a one unit rack unit server compared to yeah yeah this is what the interface looks like I'm not sure if you can see it but on the left side you have a config page where you can create those channels you know you can get MPEG transport streams in or SRT or rtmp or you have a config file a Json which you can remotely Deploy on the server to update the channel configurations and on the right side it's got a built-in video player for each lesson Dash so you can preview if the stream is okay and there's a URL which you can publish in like a middleware service or on a portal or wherever you'd like and uh normally if you know if you start trading off quality for power consumption or costs uh you will you will go down quality but we checked it against the CPU encoded we already have and you know the the the the the hardware encoding quality was up to par with the software encoding which is really cool and it can go up to 4K encoding so yeah you can get ultra high definition quality out of it um and also addressing it by standard we will use h.264 because you know that's widely adopted but it can also do a ship65 which is even more efficient and the latency is really small we've seen sub three frames latency in the transcoding while with CPU encoding you to talk about seconds uh you know that's that yeah of course and of course with software encoding you can do some tuning like forward checking but then the latest will just explode um so yeah I think this is a really cool solution for OTG the live Solutions uh and there are actually three components in there the first one is everything we do we build it in kubernetes and Docker or container environments because we run our own cloud and everything we build on our cloud is contained by Docker so it's you know you have isolated self restarting services so it's as low maintenance High uptime and we have our software encoding Services running on top of that so they're running within the kubernetes environment and then we use the net in acceleration card so you know that the A6 to do the real high performance encoding and that's a great combination this is the schematics I can put it on full screen for you so you can see it better and so on the bottom there's just just an open x86 chassis with 10 of those A6 cards and I believe they're just using like seven watts per per car so the the is really low and then on top of that we run of course the operating system in kubernetes and an average nescoding process is an isolated container so if one crashes it doesn't bring down all the other channels and if it crashes the cover needed system automated restarts it within the buffer size of the user so you don't even see that there was a restart or a crash or whatever so it's uh it really rock solid and on very stable yeah and there's an edge sketch running on top of it and then of course we have some API and GUI stuff running on top of that so it's like a hybrid solution between Hardware accelerators and this software stack to do this stuff and this is basically the process I can I can put in a full screen as well on the left side you have the transport streams and big two and big4 or SRT or rtmp going coming in um uh we skated in software and then we decoded in hardware and encoded in hardware and then the packetizing is done in software again the chunks and the Manifest files are stored on the local storage with the DVR window if you need this and then there's an edge cache actually that's two for redundancy reasons with a lot of streaming optimizations like anti thundering Hertz you know not thundering protection smart caching stuff for live streaming yes yeah so it's one solution that can just do this and what I like about it is that you can cluster it you can have multiple of those servers in one site because if you go from 80 to like 160 channels you can just add a server and it will just automatically load balance it that's the beauty of kubernetes it will just say oh there's a new server let's just share the load of the encoding but also share the load of the viewers over the machines um and if you put in three you have a high availability server so one machine can entirely break down but everything will just automatically keep working and that's nice I mean you don't want to have an operator in a hotel or in every Hotel facility that you have to you know to restart the machine and it just needs yeah that's right um so yeah it's got scalability and high availability and then we also thought about orchestration because you know if you if you have one location that's fine but what about if what about two or five or Twenty or 100 we can monitor it from a central facility from our Cloud so you can have all have all these satellite encoding locations and then centrally you can deploy the Json file for the configuration so if they if you need to add new channels you don't have to call the hotel and say hey guys just add this Channel and put in this rtmp link whatever you just updated Json file push it and boom there is a new channel that's there it is wow and then we thought okay we can go even further because you know the machine is logging and creating log files and we can actually push those log files as a stream of data to our cloud and then process it to statistics so you can have uh of course yeah of course you know we have this um elastic system Running In This Cloud processing and jumping you're jumping away says going through all the log files and making sense out of the session so then you can have reports like how many views do I have for location what's the most popular Channel what's the average viewing time and from which cities or countries are people watching those streams and all the data is there as well um so you can think of kind of you know business models that okay which channels do I want to pay for because they're not watched anymore but the most popular channels get more you know you can build those kind of cool things and finally last but not least we can actually put it into a mix so let's say you have like two or three CDN serving out capacity we can add those machines uh at those locations to have extra local capacity so deep deep edge capacity and certain locations so if traffic comes from a certain hotel or uh from a compound or from uh you know whatever location like an office we can recognize down to the IP address that the traffic is coming from that location and then prioritize traffic through the DPAC so it's not just the the streams from the local machine that it comes through about it can also become an edge for the cloud and for the cdns so we can have a localized Edge so if we're running there we can also offload internet streams and only month videos so yeah that's that's the idea super powerful and there's there's a lot to unpack and what you just presented um thank you a fabulous overview by the way um one of the things just to point out and I think everybody got this but you know we're talking about 24 7 live channels so this is like linear streaming um this would be impressive if it were even file based you know VOD um but do you have the um the the notion that you can support both VOD and linear live on the same system do you have a different way that you handle that how do you handle you know file based uh delivery this deep edge Ott solution was built for live so it's it's you know the use case is we have so many live channels from satellite or cable we need to converted to Ott streams like hls and dash and so for sure sure sure but of course um for the transcoding part we of course we also do the transcoding in our Cloud I mean we have that's right because of the cloud DVR functionality so it's not just a DVR but it's also live to thought I mean because we have customers recording live streams and then offering them as animal videos or we also have a lot of customers in our Cloud who are not doing live streaming at all they just have an online video library like an Ott service or or or or a marketing video library for for an Enterprise and we use the same technology stack for this so we started with science you know we have our own cloud with the CPUs and we thought okay if we build this transcoding software we can utilize the CPU power in the cloud but we will switch to Hardware transporting right we will at least uh a lot of the transcoding will be done by Hardware in the future because it's so more efficient and scalable compared to uh CPU based encodies I was going to ask you about your journey from from CPU and software based you know encoding to hardware maybe you can give some more insights into what what really drove that you know in in our talking to the to the market to the ecosystem we find the usual a lot of people just say well it's just purely cost you know it's just our operational cost but but there's some other factors as well and I'm just curious you know for you was it just purely an economic uh decision or were there some other factors that drove you to even Explore hardware and then ultimately land on A6 as a software house of course Hardware is like we don't like appliances we don't like it exactly yeah yeah you're virtualized everything's virtualized so it's like well what are we going to do with Hardware actually like 10 years ago we went to a show in London uh where a lot of vendors tried to sell their appliances to the Taco industry and build CDs you know at Cisco with a stack of appliances it would cost like a million euros to build a stack of it's a pop and I went on stage and I said hey guys I have news for you the appliance is dead because now we have software controlling everything and software edges so forget about this so one half of the of the audience was turning white because those were the vendors and the other will help with the audience yeah like okay let's let's have a meeting so but yeah let's talk but you know things flow back and forth so now we're like okay there are some limitations in software um if you if you talk about scalability and costs so our arguments were not just cost Factor but also about scale which at the end of course again is cost um yeah that's right yeah but to to be able to do more of what transcoding and more especially for I mean what transcoding is not even that hard because it's not linear it doesn't have to be real time yeah I mean it can be faster than real time but not necessarily sure but for live streaming to have capacity for customers to say I need to start a live stream right now and it has to be transcoded too you know four or five bit rates yeah you need to have a lot of Headroom in your in your Cloud uh to to be able to do that yeah and we were like okay if we do that in software we will have to buy a lot of equipment which will just be eating dust 90 of the time and it will use a lot of energy um and we want to be as green as possible uh with a lot of creating the energy but also with a small footprint so we were like if we use these cards we will have much better scale and uh and we don't have to invest and over invest in the infrastructure for that you know those few days uh a week that customers are peaking with a lot of concurrent lighting conditions today all the internet streaming basically is h264. as soon as you need to go to h265 which is not that popular but if you want to go to av1 for instance which we think will be a dominant encoding format in the future you have to throw a lot of CPU power rank against it for encoding which basically brings the business case for for av1 and with Hardware accelerated like the A6 that's instantly becoming a feasible business case so very interesting point uh that you make is that um there is and and we talked to uh you know many of our customers and uh um you know companies who are seriously evaluating Hardware particularly Asics and you know behind closed doors um you know a lot of them do uh reflect that you know three three to five years ago if we'd approached them they would have just flat set you know go away like Hardware you know like like you know good luck uh and and and so there uh but you know the point is is it is now they're very uh very engaged and very leaning forward and and also what's uh really fascinating is is that it's not that it's just kind of a temporary situation of for example energy costs you know or even uh you know okay the CEO is mandated you know we need to trim 20 from the budget and so well how are we going to do that you know yes um yes those things sometimes start the conversation uh but there's so so many more benefits and like you point out about the advanced Kodak the next-gen Codex support you know the the industry has just gotten to a point now where codex the the Next Generation if you look at the complexity of VVC let's say just a um you know but all av1 as well so VVC over hevc or av1 over vp8 vp9 yeah the VPX codex it is um it is so significant and when you factor in you know even if you're able to get which on day one you never get the 50 promise Savings in and bitrate you know it's more like 20 or 30 percent but in time the codecs get optimized right but so even best case scenario let's say on day one you could get 50 savings the added compute cost through the complexity of the codec it negates all those savings and in some cases and then some and so it's this real conundrum that the industry is in because on one hand you know a we want to always be moving to uh you know more efficient codecs want to be pushing our bit rates down resolutions are only going up customer quality expectations are increasing and yet for many services and platforms they're kind of stuck but Hardware in particular Asics are the Breakthrough you know for that and so yeah we are seeing um a really phenomenal I I know of two um you you know file based very high profile premium streaming services right now they're on the cusp of of deploying hardware and it's in a use case that you know I think a lot of you know traditional uh streaming Engineers might say well no you can use CPU for that and guess what their conclusion is no we can't you know if if you if you have to and go to film once and then you can stream it to Millions uh for years then of course the business case for high uh high cost encoding is better we have a lot of customers in niches and they need very high quality they don't have these high order these large audiences yeah so then there's no business case for the encoding if you do it in software so um that's right there will still be a market for software encoding I mean the tuning you can do software will be better all right but we also I mean when we first tested those cars we were first you know the first thing we did was look at the Quality is it really good I mean if the quality is average or poor then then it's like substandard and we wanted to have a high quality so we were very pleased with with the quality I mean if you compare to GPU encoding that's bad compared to CP encoding but but with video things it's it's in the real Realms of cpu-based encoding which is cool just to reinforce that point because I I you have seen this um this data but perhaps our listeners have not or not everyone has um so you know what's also so important when we're talking about software versus you know versus Hardware is um like you said if you have the benefit of encoding a file and then serving it you know literally hundreds of millions of times and you know there's a few Platforms in the world that that have that benefit but again you know pay close attention what I just said there's a few Platforms in the world meaning the whole world there's like a few you know Netflix of course being at the top and you know there's there's the other ones we all know um but what's super interesting is is that when you go to live when you're talking live and you're encoding in software you can't operate the encoder so like x264x265 is a great reference point because pretty much everybody in the industry understands the speed levels and can kind of associate quality at you know at like slow or at medium or fast or faster or fastest and you know the various levels so um in in in our evaluation uh and you know whether you use objective or sub subjective quality metrics uh our live performance is on par with x265 medium and that's live and there are very few services that could afford to run x265 on CPU at medium and and you know just because the compute you know you're going to need a lot of cores to do that and so you're not going to get the density so true if you're only running a single Channel well again I suppose you could say okay fine I don't care you know I'm just going to put a a big AMD 64 core machine there and you know and everything's good right um but what's super interesting is is that the Asic is able to achieve medium uh medium x265 medium speed quality but live and provide a 20 bit rate reduction on top of that so not only can it reach the quality but deliver 20 percent fewer bits at the same time which is just and is that the h265 encoding as well in the in the Asics that's right I'm talking about hevc yeah yeah so yeah so it's uh you know I also think that's that's an important reference point but let's get back to talking about uh you know what you built so so what is next you know um talk to us about this is a very interesting use case um I know about the types of projects that you're deploying it into and the regions of the world and it absolutely makes sense and you know I I know you're going to do very very well uh with this solution but you know how how are you going to grow it how are you going to expand like what's on the horizon are there are there some other use cases some other applications that you know might either be be well known or maybe also novel you know that you're targeting we never foresaw that we actually would uh develop an appliance because that's not our business our business is South right uh but when this customer came to us and we said but this is an interesting use case uh and we had so much technology on the shelves which we already were using in our Cloud so basically we just you know it's a it's a micro Cloud running in an appliance that's that's what it is um so didn't take us that much time to develop this this product technically and as a product um so we now have this customer starting to test it they started testing the software solution now they want to test the hardware solution um uh I don't want to full-blown start selling it I want customer feedback I want them to operate it for a quite some time and then see how they like it and how we should tune it to you know to to have a good fit for it for them and then uh we will make lists of potential customers in in markets and maybe maybe we will do it and sell it ourselves or we will find a partner like someone who already has a sales force into certain markets uh some joint marketing with you guys would be interesting that those are the things we would we would looking at we're looking at to to start selling this um we're not interested in selling Hardware right so we have the software stack so we will sell it as a turnkey solution but we will primarily focus on the the software license for it what parts of the world are you primarily focused in or or are you literally selling globally um well just Justine is European company with the European Focus so our customer base is in Europe uh but our audiences of course are around the world because you know we have these seniors who can deliver throughout the world so but our primary customer is in Europe you know I think it might be interesting to uh talk about from from your perspective um there's an interesting uh um you know challenge that when you talk about the video Distributors the platforms that are you know let's say a notch or two or even a few more notches below like like like a Netflix of the world or or the you know pay TV platforms you know some of the large you know what satellite or cable or even IPTV um and you know it's interesting because there's there's there's this tension between um the the you know content licensing costs the technology the business model and sometimes it's hard to get those kind of you you know proportions correct and I've seen over investment in in in all of those areas right but they're sort of like levers you know you need to get them in the right position or else it's going to be really difficult to make money you have been in the business for a long time you know you have seen the technology expansion and growth and you've seen the shifts and the changes and you know so do you have any insights um about either where you know the state of of streaming and and and the technology stack is today where it's going you know is there anything that you'd like to share technology wise or business gauge wise or both or both is is is is is fair game you know more and more it's my encouragement to engineers and and people who primarily live on the technical or the product side um to begin to get an understanding of the business you know because I think it makes us better Engineers it it allows us to build better products um and and and then you know always of course you know I think there's a benefit for the business folks to you know to have some understanding that technology but something that I think our audience is primarily going to be engineers and so if you have insights on the business side go for it it's it's interesting uh if you look at the business rise in the in the o2d market like the Netflix and HBO's uh hardly any of them are making money right and that's that's a challenge of course I mean uh there's a discussion going on between the telcos and the Ott companies uh uh at Mobile World Congress you know companies raised the question like why should not the Ott providers pay to tell cause a little money because you know they have to keep scaling up their infrastructure uh yeah I I I posted something on LinkedIn a few days ago and said yeah but if you look at this industry you have a lot of lost leaders not just in the Ott space but also in the vendors I mean name me a vendor that's really profitable right now not that many are and um actually most I love leading and and burning through a lot of money um it's both the otd vendors and the technology vendors are burning through a lot of money so basically they're subsidized by investors who hope that one of those companies will eventually you know get better better rates or better revenues and or be bought by another company that's not really healthy it's not sustainable um I think there are too many vendors in this market space and I think they their expectations haven't been too high some of those companies cannot meet their Expectations by the investors so they're getting into trouble that's bad of course um and um another interesting observation I made was you have all these cell calls and the tacos are in a crowded saturated Market but they're making money and the Ott vendors you know the Netflix and HBO's are also I think they're basically saturated I mean there's so many offerings and they have all the audiences and I don't think they can have more customers as they hope it's it's getting saturated but none of them are really making money so why why would then those go the Ott company start paying the telcos or who are actually profitable and those guys are not so that's an interesting that's an interesting perspective and I'm sure Mobile World Congress was not that popular I tend to agree with the basic premise though so yeah you know if you look at these these hyperscale Otto Defenders like uh Netflix and all those guys they have the budget to put in Edge caches within the tell girls to negotiate great private hearing deals and that's great so I don't think there's a real problem there is not a challenge yeah or newcomers in the industry who don't have the Deep Pockets and don't have the skill you won't be able to negotiate the same deals with the telcos you won't be able to get your ad servers in there and that means that you will get uh industry some really high large guys with great performance great capacity and low cost and Challengers with all kinds of uh thresholds to get into that market and that's uh I think it touches the discussion on net neutrality so how far shoot attackers be forced to also open up their networks for those guys what about any insights on the technology side you know and and it doesn't necessarily have to be codecs or you know encoding but is there anything that you're seeing any Trends any requests that are coming in from you know these telcos and these operators even if they're you know tier two tier three or working in smaller markets that might be interesting the the markets we work for are medium-sized customers typically who need very complicated workflows and that's what we solved with what we call our mixed solution where you can you know start an easy way and then go under the hood and start tuning and tweaking features which you can to do with regular video platforms and then if you go to the expert level you can plug in your own transcoders your own players you know build your own work for yourself using our our stack of uh of workflow appreciation tools and streaming features um that's cool uh that I I that's we we get a lot of positive feedback on this because a lot of people who enter the streaming industry and they're not necessarily have to they have data not necessarily they are otd provided it can also be like an Enterprise who needs to do some online video stuff platform that wants to do something with video or live streaming um there are two choices in the market either you go to a video platform and that's easy you know you sign up for uh not that much money you you can upload your videos and they do everything for you get a video player and you can publish it on your website um that's uh but a lot of those companies get stuck there because they cannot optimize their encoding quality they cannot optimize the player they cannot optimize the multi-seating distribution they go don't get access to the data they need so they get stuck with those platforms and the step to go to building your own your own stack of technology is Extreme because then you start to have to hire uh Cloud export streaming experts who you know go to Azure or AWS and start configuring all these modules and sticking everything together will take a lot of time a lot of money to build that and then you have your own your own your home brewing streaming platform which does what you need but what about tomorrow what about you want to go from h.264 to av1 or whatever you're stuck again because that's the problem so um uh that's what why we also claim to have this 40 cost reduction it's not just in traffic or transcoding costs it's also in operational costs and something that people underestimate how much time it takes to build and maintain a streaming platform and how much expertise you actually need in-house to do that just because you can grab an open source Library you know which is super powerful right and and I think it's a wonderful thing you know it's great that we have ffmpeg is developed you know as it is and x264 is you know let's face it it's an amazing encoder it really is and x265 a great encoders so this is all wonderful but you know it's one thing to deploy it at a moment in time it's another thing to maintain that platform over over over the life of a service and I think that is the point that gets missed by serve you know by services and companies and platforms who um by All rights do have Smart engineers and and they do have the talent to be able to build it so it's not that they can't but it's that maintaining it for the life of the service that's the part they miss so it's one thing to build it have it work today have it work tomorrow have it work next month and even for the rest of this year it's another thing to have that same service Rock Solid you know uh in in 2025 you know when maybe our user base has scaled 10 20 30 40x 100x you know hopefully you know over what it is or was today so if you look at uh uh equality I mean most of the content on the web is like HD or full HD but what about 4K and what does it mean for you what does it mean for you you support AK you know that's coming one day so hopefully there are also two types of uh of Worlds coming together we have broadcasting traditional broadcasting and we have Internet and on the internet we are used to having a very Dynamic Solutions I mean we have to change protocols you have to change infrastructures and what what is working today may not be working tomorrow uh while in broadcasting people are used to build systems for life you know you build it you don't touch it yeah um and yeah and especially the the people who have this more like traditional broadcast attitude of engineering infrastructures they're having a real hard time to understand the Dynamics of the internet because as I said tomorrow we can have I mean I'm not I I saw that Safari from uh from Apple uh the new latest beta would would actually uh introduce av1 so then overnight that's right overnight this industry can change and then how will you change your encoding and if you have you thought about the the the effects on your CDN and Origins if you start you know if you migrate from h264 to av1 probably not and then it will break or you're logging will break or your monitoring so you have to think about more future proof things and by the way I also started Apple pooled av1 out of the latest Beta release so I'm I'm really curious what's going to happen there well Steph um this this has been a really amazing discussion and you know I want to thank you for joining the voices of video and sharing uh all of your insights and what you built uh with us and with the audience so um thank you uh thank you Mark and why don't you tell everybody where they can go to learn more about jet stream oh it's simple it's jet-stream.com well thank you

2023-04-19 13:11

Show Video

Other news