for our final video here at Super Micro headquarters in beautiful Silicon Valley California I've got something you might not expect you probably think of Super Micro as a server designer and manufacturer but they actually have the capacity to build up to 3500 fully integrated plug andplay rack Solutions per month of which a thousand of them can be liquid cooled how by having their development manufacturing and validation all in their Silicon Valley campus every rack scale build starts with talking to the customer and understanding their needs after that super micro can draw from their extremely broad portfolio of compute Power Cooling and networking products to build the perfect one piece that rolls into place that rolls into place and can be plugged right into the wall of course achieving that level of integration doesn't happen overnight and every rack scale solution starts with what you see behind me a barebones skeleton of a rack you choose your own network blocks compute blocks storage blocks whatever works best for you they'll Build It Up Network it up and then it goes over to where things get really cool building a server it's nothing without validation and to that end super micro built their l11 l12 1 megawatt testing facility behind me it's got the cooling and power capacity to test everything from an individual server to a rack scale solution to a super computer all in this roof I can tell you you can actually probably see my hair blowing around from all the HVAC that they've got running in here this place is absolutely wild of course I wouldn't blame you for not being sure exactly what an l11 or l12 testing facility is so to cover that for you I've got my friend Minaj here from Super Micro who's going to explain it hi hi lonus so yeah you want to know more about the 11 test right I do what does that mean so when it comes to building a supercomputer one of the main challenges that we face is that all the nodes hundreds if not thousands of nodes have to have the same bios and firmware configuration and Os settings as well so what we do in our l11 testing is that we have developed our fully automated unique L1 test that we can run simult ously on thousands of nodes and run those consistency checks to make sure that all the nodes are talking to each other in the most optimized way super cool but what is l12 then okay so l12 is where you run your actual applications on the cluster level and make sure that by the time I actually receive the system it is fully validated and everything should be performing exactly as I expect literally all you do is plug in the rack and start running your application on day one that's it and all you need to do is go check out our entire series of H3 13 videos on super micro's landing page where we cover everything from Super micro's Hyper to Cloud DC to Giant 8 GPU compute servers it's worth checking out