the afellow scallars this is tomin it perpers with doctor carroge oni feed last talk e bas no simulations being able to similate snow nar computers is not now its been possible for a few years now for instance dis legendary dis ne paper from twenty thirteen was capable of performing that so why do researchers rite no papers on this topic well because the twenty thirteen paper had its limitations lest talk about them while looking at some really sweet footage from a no paper limitation number one for previous works is thet their no similations were seenfully expensive most of them took nat sacons and sometimes nat even minute that half an hour pefrain yes that means all nitter similations they were also natas good in fractoring interactions seimilating powder snow and evalenches were also out of reach antionow you see seemilating snow is quite challenging it clumps up deforms braves and hardens under conpression and even my favorit face change from fluid to snow these are all really challenging to similate properly and in a moment hou will see that this no method is capable of even more beyond that let's start with riction first with turn on this no machine and then engaged ta wipers dhat looks wonderful and now mea requests seeing some tiremarks tare you go this looks really good so how bout taking the closeer look at this phenomenon evb here is the boundary friction qorofficient and it is a peramiter that can be chosen freely my us so latsey what tat looks like if we inisialize this to al ow value wol get very little fiction and if we crank this up look things get a great deal more sticky a big clamp of snow also breaks upart in a spec tacular manner also showcasing compression and fracturing beyond the boundary fiction effect we just looked at oh my this is beautiful okay now this is a computergraphic spaper if you're a seasond fellows colourd you know what this means it means that it is time to put some virtual bonyes into the oven and thes is a great example for rule number one for watching physics simulations which is that weh tdiscuss the physic sport and not of visuals so why are these bonies blue well let's chuck hem into the oven and find out a hah they are color coaded for temperature look may start from minus one hundred hauseiears tas the blue and we see the collor's change as the approach izero degrees sasius an this point they don't yet start mouting but they are ora dy falling apart so it was indeed a good design decision to show the temperatures because it has us exactly what is going on here without this we woll be expecting mapting well can we see that mouthingg inection to you bett no hold on the eur papers and bring forth the soft i sow mat this machine can not only create an exquisite desert for computer graphic researchers but also showcases the individual countributions of dhis no ectique one by one there is the matting yes a da little frosting and tare you go bonapity now as we feared many of these larger carce simulations required computing the physics for millions of particles so how long does the take when we need millions of particles with typically have to wait a fuw minutes perfame but if we have a smaller sene we can get away with thise computations in a few seconds or frame a goodness we went from hours perfame to seconds perfane in just one paper out standing work and also way hes second if we are talking millions of particulars i wonder how much memory takes to keep track of them blatsy wo this is very appealing i was expecting a fewl gigerbites yet it only asks for a fraction of it the capple handed magbites so with this haf devalue proposition it is no wonder that theys paper has been accepted to the cigarette comference this is the only pi gold madal of computergraphic research if you will huge congratulations to the authors of thes paper this was quiete an experience what a time to be alive this apissolt have been supported by lamba gpu cloud if you're looking for inexpansive cloud gpuse for ai check out lamda gplo cloud the recently laged quadro arti xic thousand arti ex pe thousand and v one hundred instances and hold on to your papers because lamda gpu clouds can cast less than half of aws and ajier plas the arthe only cloud service with forty eal gigobite party as eih thousand joine researchers at organizations like apple amit and catack in using lamduck loud instances work stations or servers make sure to go to lembalab dot coms lash papers to sign up for one of their amazing gpuw instances today are thanks to lamda for their long standing support and for happing us make better videos for you faxks or watching and for your generous support and i see you next time