Big Compute - The Helicarrier: When Movies Ignore Natural Law

Episode Date: September 26, 2023

We’ve all been there – Sitting inside a movie theater, fully engrossed in the film, when suddenly something happens on screen that is so far-fetched that we suddenly remember we’re sitting in a theater seat. This is the experience Rescale’s Edward Hsu had back in 2012 when he was watching the movie, The Avengers. As an aviation enthusiast, everything was well and good until he saw the epic sea structure called the Helicarrier take to the skies, and it was blatantly clear to him that the incredible piece of high budget visual effects would probably – in fact – kill its own jet pilots.  And while it was just a movie, the scene lingered in Edward’s mind for years, until the day Sandeep Urankar – an engineer with a knack for computational investigations – was able to run simulations to prove if Edward’s hypothesis was true, and consequently fueling the question – how accurate should big budget movies really be when it comes to natural law?

You don't want the nerds coming after you. I don't, actually. So I'm really glad you're correcting me. It's just the nerds are going to nitpick, right? They're going to be like, well, you left out the computational part. And arguably, that's the most important part of it. Hi, everyone. I'm Jolie Hales.

And I'm Ernest DeLeon. And welcome to the Big Compute Podcast. Here we celebrate innovation in a world of virtually unlimited compute, and we do it one important story at a time. We talk about the stories behind scientists and engineers who are embracing the power of high-performance computing to better the lives of all of us. From the products we use every day to the technology of tomorrow, computational engineering plays a direct role in making it all happen, whether people know it or not. Hello, Ernest!

Greetings, fellow podcast host who is also not an AI. Wait, what do you mean also? Is there... Oh, you mean like you're not an AI? They're doing a lot more than just podcasting with AIs. But one of the things I have noticed is that there's definitely an increase in AI written things going around, like marketing emails. It's a kind of a dead giveaway that it's not normal. Like I guess humans speak, if you want to call it that. And now we got AI everywhere. We got humans using AI and it's like, I guess, human speak, if you want to call it that. And now we got AI everywhere. We got humans using AI and it's like, okay, well, how do we know that each of us are not AIs? That's

the question. Well, at least the saving grace, I think when it comes to the creative arts is yes, they can have AI do voiceover and host videos or podcasts, but it doesn't sound like a human with a soul to me. Like there's no variation. It's almost too perfect. Yeah, down the line, I think they're going to get better and then it'll get scarier. But for now, I think it's pretty obvious when you're listening to a podcast if it's actually a human or not. Thank goodness.

What are you talking about, Jolie? You can't tell the difference between us and people of your kind at all. But Ernest, to start this off, I have a question for you. Yes. And actually, now that I'm about to ask it, I think I'm probably asking the wrong person, being that this question involves taste in movies.

Oh, that's right up my aisle. We shall see. Okay, so here's the question. Have you ever been watching a film? Oh my gosh, you're seriously the wrong person. That's okay. Have you ever been watching a film and thought, wait a minute, that would never be able to happen in real life? And I'm not talking about like when it comes to something fantastical or like magical or something where the

full intent of the filmmaker is to actually take the audience to places outside of the realm of reality, but more of like when a filmmaker is attempting to show reality, but they fall short in their attempt. So this has happened many times, but the one that comes to mind specifically for me was the Robocop movie. I believe it's from the 80s, if I'm not mistaken. And there's a scene where like acid spills on this guy. Okay.

And he just like slowly starts melting. Like, you know, if you put salt on a snail or something and I'm sitting here watching this thing thinking, this is not at all how this works. But yes, I, I have had that happen. And also it happens every time I watch a Michael Bay film. So, well, yes, of course, Michael Bay, he's in a ballpark all his own, but RoboCop. So wait, that's an 80s film and you're an 80s kid. Does that mean that you had this thought when you were a child? No, it's when I watched it later. Oh yeah. I wasn't allowed to watch that as a child because it had a lot of violence in it. I don't remember. So was this not a kid movie? No,

no, no, no. This is very much an adult movie. Oh, you know what? I think this was in our AI episode. And I do remember I was mixing it up with there's another cop. Kindergarten cop. Something. With Arnold Schwarzenegger. Maybe I'm I think I'm mixing it up with kindergarten cop. Yeah, that one's a comedy. This movie's super violent. They're a little different. I wasn't allowed to watch it as a kid, but I watched it later, probably the mid to late 90s when I was already, I wouldn't call myself an adult. You're old enough to know how acid probably works. Old enough to know how acid works and was like, hmm, this just doesn't seem right. Yeah, I think that happens a lot, probably to most people where you're watching a movie and suddenly you get taken out of it. I think really good filmmakers prevent that kind of a thing

so that from the moment you sit down in the movie theater seats to the moment you stand up, you forget that you're just sitting in a theater, right? But the first time that I remember being exposed to these liberties of Hollywood's perception on natural law was when I was actually a kid watching Home Alone 2. Have you seen the Home Alone movies, at least the first two? The other ones don't really count. I've definitely seen them at least one in two.

Okay. I still watch the first one every year around Christmas. Oh, yeah. It's awesome. When the McAllister family left on their Christmas vacation. Did we miss the flight? No, you just made it. Yeah! They forgot one small thing.

Kevin! Home Alone. So when I was a kid, I remember I was sitting there, I was watching this movie at home during Christmas time with my family. And there was a scene that came up where Kevin throws bricks off the roof of the apartment building. And these bricks like fly down a few stories and they hit the burglar Harry in the face. Do you know what I'm talking about? Yes. What?

And I didn't really think anything of it except that it was funny. But I was sitting next to my dad and my dad is a doctor and he just smirked. And I remember he said that would actually kill the guy. Sure would. And then it led intoed. And I remember he said, that would actually kill the guy. Sure would. And then it led into this discussion about all the many ways that Marv and Harry would have either been killed or instantly hospitalized throughout the Home Alone movies. And then I remember thinking, well, wait a minute. That means that the things that we see in movies aren't real. It was like this revelation to me, right? And of course, after that, I could never really watch a movie, let alone the Home Alone movies, without noticing the many creative liberties that are taken

by the filmmakers. But then, of course, ironically, fast forward a couple decades and now I was the filmmaker seeing things from the other side. Although I'd like to think that I was pretty pragmatic in my approaches to natural law on screen. I didn't do a lot of visual effects movies in my time in Hollywood. So that helps, right? Because it's pretty easy to stay within natural law when you're doing like family films or comedies or narratives or even documentaries. But now I do film and video, as you know, for the tech industry. And if I put something on the screen that is incorrect, all of the tech and engineering nerds like come after me so i have to be especially

cognizant yeah the meme of the nerd pushing up his glasses by the bridge and saying actually immediately comes to mind when you say that and it's true and as a recovering engineer you have to forgive us jolie our entire lives are driven by specificity and precision. So it's hard for us to accept ambiguity. That makes total sense. And it's super endearing. I actually really like this, right? My wife tells me all the time that it's like a special skill to associate with an engineering type person on like a daily and normal basis. Like we're not normal, I guess, if there is a definition of normal. So I totally understand how if you put something that is factually or even just like incorrect and like the most what we would consider insignificant little way, there is somebody whose career is that thing specifically.

It's so true. And they're going to come for you. Around here, Oscar is known as actually, because he will insert himself into just about any conversation to add facts or correct grammar. Now I actually have a new appreciation, having been surrounded by, what, quote, your kind? Yeah, like we're aliens, right? Yeah. Yeah, and all of this is relevant to this discussion today, right ended up actually running computational simulations. Which brings me to my next question, Ernest. How familiar are you with the Avengers movies?

I'm a big fan of Iron Man and Hulk. I always have been since I was a little kid. So I'm pretty familiar. They did justice to the Iron Man character in the Avengers series. But man, did they do the Hulk wrong. I'm still mad about that. Yeah, I don't think I've seen the Hulk be done right since I was a kid and it was... Since Lou Ferrigno played the Hulk?

Yeah, yeah, yeah, basically. Don't make me angry. You wouldn't like me when I'm angry. But Marvel Universe as we know it today has been making these movies since 2007. Obviously there were Marvel movies before then, but they weren't technically part of the same Marvel Universe, right? Right. And since 2007, there have been 32 of these movies.

And there are also 11 more apparently in the works. So they're really diving into these Marvel movies. I think they bring in a lot of cash. And I was actually working for Disney when Disney bought Marvel. And since I left the company, they've already expanded that Marvel footprint inside the Disney parks, too. So like in California Adventure at the Disneyland Resort, which is where I used to have my office, they have this entire Marvel themed land now with rides and shows. And they even converted the Twilight Zone Tower of Terror into like a Guardians of the Galaxy themed ride. I mean, have you been to

the parks and seen any of this? Unfortunately, no. I haven't been to a Disney park since I was a kid. I do really want to go and I keep bringing it up to my wife, especially when they bought the Star Wars stuff. I wanted to go. But yeah, then again, I kind of want to wait until my kids are a little bit older so they can kind of appreciate what's actually there. I think you're smart to do that. So not only can they remember it, but they can actually participate in more than just the Fantasyland rides and the carousel. But yeah, so there's this whole Marvel themed area back there. And oh, have you seen the video of the Spider-Man stunt mannequin that like went awry during a show and it crashed into the building? I haven't seen anything about that.

Oh, my gosh. Okay, here's a link. All right, let me watch this thing. Karen, airbags, please! Well, that's what you take out insurance on set for. Oh, it's amazing. Clearly, even Disney can't get away from obeying natural laws, at least in live stunt shows. And Spider-Man is part of this Marvel Universe that we're talking about.

And, oh, did you see the last movie where the past Spider-Man actors, like, all came together in this multi-dimensional story type thing? No, I'm actually not a really big fan of Spider-Man. Wait, what about, like, Into the Spider-Verse, which is actually not Disney really big fan of Spider-Man. Wait, what about like Into the Spider-Verse, which is actually not Disney and it's not part of this technical Marvel universe, but it is Spider-Man. Did you ever see that? The animated one? No, it's again. Oh, it's so good. It's so good. You know, I've heard that from a lot of people and I like Spider-Man. It's just that I'm a man with very limited time. Touche. And so like if I'm going to spend it on something, 80 hours of going into Final

Fantasy 16. That's what's happening. Right. Which is exactly how you and I have spent our evenings playing this game and talking about it for like a few months now. And I totally have empathy for you a hundred percent. I was looking at this list of 32 movies and I think I've seen about a dozen of them and they were all before I had my first kid. Exactly. And I'm a filmmaker. Now I've seen about a dozen of them. And they were all before I had my first kid. And I'm a filmmaker. Now, I've seen all the Avengers mainline movies, minus the character specific ones. Oh, OK. So the ones where everybody gets together.

Where everybody gets together, right? The things that I have watched beyond the mainline and Hulk and Iron Man were literally when I was flying on an airplane between point A to point B just to pass the time. So I think we've established you and I have seen a number of the Marvel movies, but we're not like die hard, see every single one of them kind of fans. And for those of our listeners who are kind of like us, or even maybe less inclined to see Marvel movies, those movies include characters like Iron Man.

You said the Incredible Hulk, Thor. I think some of the Thor movies are kind of funny. Yeah, those are funny. Captain America, Guardians of the Galaxy, Ant-Man, Doctor Strange, which is kind of a personal favorite for me because I just think Benedict Cumberbatch is phenomenally talented. What else? Captain Marvel, Black Panther, although the actor who played Black Panther died, which is super sad, but I think they're doing some sequel stuff there. Black Widow and then even Deadpool and the Eternals now are part of this universe. I think it's probably safe to say that most

people out there who watch movies at all have probably seen at least one of these 32 Marvel movies. I'm sure there are some engineering types out there listening who haven't. Rescale's CEO among them actually. But even those people have at least heard of Marvel movies and characters, right? If they haven't heard much about these

movies, they've probably at least seen a Marvel commercial. Or if not a commercial, then maybe a billboard. And if not a billboard, then maybe an action figure toy or something, right? Which brings us to our story today. When the Helicarrier came out in the first Avengers movie, it was pretty cool, right? To see that thing take off. And I was like, wow, that's pretty badass. And then as soon as it takes off out of the water, you see where the fans are that lift up the helicarrier. And immediately, you know how sometimes when you see things in movies and the suspension of disbelief kind of breaks because you're like, OK, that doesn't work. That's Edward Su, chief product officer at Reskill, who was enjoying his time being pulled into the story of the 2012 Marvel hit The Avengers when suddenly he saw something that

didn't look right. And consequently, he was reminded that he was sitting in a theater. And before we get into exactly what Edward saw, we should explain what the Helicarrier is. And this is not to be confused with the company called Helicarrier at Helicarrier.com, which apparently is just a random company that uses helicopters to carry stuff. In case anyone tries to Google it, which I did. But according to Marvel Cinematic Universe Wiki, which is a real thing, helicarriers are, quote, advanced flying command centers that double as aircraft carriers. Specifically, helicarriers were designed to be capable of sustained, independently powered flight via four massive turbine engines that provide the lift needed to keep them in the air while also being able to retain seaborne capabilities. Their technology

was created in part with designs by Howard Stark, end quote. And who is Howard Stark, Mr. Iron Man aficionado? I think everyone knows who Howard Stark is. That's Tony Stark's father. Right. Tony Stark, who is Iron Man. Tony Stark's father. Exactly. But more generally speaking, a helicarrier is basically something that it looks just like an aircraft carrier. You know, the giant ships with runways for planes to be able to take off and land in the sea. Or to simplify even further, the five-peg ship in the game of Battleship. Yeah, there you go.

Which was always the easiest one to hit because of its sheer size on the game board. Which is why in real life, aircraft carriers never sail alone. There's an entire carrier group that escorts them. And the rest of that group, their entire job is to protect the carrier. Man, you must be so good at trivia games. You know all the things. Like, you must be so good at trivia games. You know all the things. Like, I would never know that.

No, that just comes by virtue of being the grandson of a Navy Seabee. And only this particular Marvel aircraft carrier, the Helicarrier, not only does it float through the ocean, but it can also, which makes it unique, lift off from the water and fly, which we see on the big screen in that 2012 Avengers movie. Is this a submarine? Really? They want me in a submerged, pressurized metal container? Go ahead, watch it now. Hold check. Aye, sir. I would imagine the visual effects people just took a modern aircraft carrier

and just stuck some fans on it like you might have on a drone. That's our friend Edward again. But the problem is, as the helicarrier lifts off, you can see that there's a lift fan right at the end of a runway. And to know what Edward is talking about, let's visualize the helicarrier, which looks similar to a standard aircraft carrier before, of course, it lifts into the air from the ocean. So to simplify, the helicarrier is this giant ship with airplanes on it, and it has one giant runway that runs from end to end across, you know, the long distance of the craft,

I guess you could say, straight across the top, which is meant to be the takeoff runway, apparently allowing aircraft to take off right from the middle of the ocean. But then there's this like second runway, which is much shorter, and it's kind of stacked on top of one side of the takeoff runway, like a bunk bed or so. And it's angled from the main runway. And that's meant to be the landing runway and I should specify the helicarrier that we're talking about today is the one that shows up in the 2012 Avengers movie there's actually other versions of the helicarrier that have appeared all across comics and tv shows over time especially in animated tv shows where they didn't need some giant budget to

show the helicarrier right and I think the first time it actually showed up in pop culture in some way was in August of 1965 in a comic book. And then later versions of it have also replaced the fans that are used to lift it with some, like, hoverboard-type technology or something. So the one that we're talking about today is the 2012 movie version. But anyway, Marvel has diverted away from mirroring exact placement of runways from actual aircraft carriers a bit. For example, typical aircraft carriers don't usually have some double-decker, like stacked bunk bed type runways in real life.

But that's not what Edward took issue with when he saw the movie. Those runways are important because aircraft carriers, because I'm also a Top Gun fan, right? They launch off of catapults, but when they land, they need to recover in this airstrip, which they tilt at an angle because pilots oftentimes miss their landings and they have to come around. And so it occurred to me if any pilot missed their landing, they would immediately be over a giant lift fan

that's big enough to lift a very, very heavy aircraft carrier. So I would imagine a little fighter jet barely at landing speed trying to take off for another round is going to get sucked in and, you know, crash into the fan. The way that the landing runway is angled, it's pointing directly at one of those massive fans, which is fine if the airplane lands smoothly on its first attempt. But on an aircraft carrier, that's not always the case. On takeoffs, aircrafts use catapults because most planes do not have enough thrust to get to flight

velocity within the length of an aircraft carrier, right? And so the catapults, if you ever watch a video of these pilots when they're taking off they put their head back they're not even grabbing the stick they're like holding onto the plane and the plane gets launched and as soon as it launches they take control and that's how they take off and then when they land it's the same problem right it's going to be very difficult to touch the aircraft carrier in the exact same spot and make sure you have enough runway to use your brakes to slow yourself down. When a pilot is attempting to land on an aircraft carrier, the plane's brakes aren't enough to stop it in time. And it has this tail hook that's on the back of the plane that has to actually grab a cable that's on the aircraft carrier called an arrestor cable and that cable helps drag the plane to a stop before it rolls off the other side of the runway. But in order to make such a precise landing the pilot has to be going slow enough to be able

to be helped to a stop by this arrestor cable tail hook thing while also flying yet fast enough to take off immediately again if that hook doesn't catch the yet fast enough to take off immediately again if that hook doesn't catch the cable or like something is off during the landing. If you watch these documentaries or just see how planes land, they hit the deck and as soon as they hit the deck, they go full power and you can actually see the afterburners light up. The pilots assume they have to be prepared because if they slow down and they miss and then they wait till they understand that they overshot and then power up, it might be too late to gain the speed. So they actually all land, hit the deck and go full power on the engines. And having to take off immediately again after not nailing the landing on an aircraft carrier

is not an uncommon thing. Apparently, all Navy pilots at some point committed what they call a bolter, which is when they actually miss and they have to go around. Every single one. Yeah, I mean, not all the time, but at some point, everybody's done that. Which is fine on an aircraft carrier

because the pilots are trained for this. But the helicarrier is a different story because it introduces four giant drone-style fans attached to the sides, and one of those fans is sitting directly below the end of the landing runway, right under where a pilot would need to fly if they had to do one of those bolter maneuvers or, you know, had to take off again after not nailing a carrier landing. Imagine you're about to jump off a cliff as a hang glider. You probably want air that's flowing towards you and up. That's ideal, right? And now instead, imagine you have a giant fan

that's just sucking in air and generating enough force to lift at least a quarter of a Nimitz-class aircraft carrier. Right? Would you want to jump? It was clear to Edward as he sat in that theater in 2012 that the likelihood of an aircraft that boltered being pulled directly down into the fan to the pilot's demise was quite likely. Likely enough that it jolted him out of the story at that precise moment. A good friend of mine who's a former military guy said, oh, I had the exact same thought when I saw that movie.

So I'm not the only guy. There's other people out there that saw it and got really annoyed. I know you get all the way to this point in the movie and that's what gets you. I mean, this is the Avenger. Like everything's fake in this movie. That's a good point. I mean, this is the Avenger. Like, everything's fake in this movie. Yeah, that's a good point. I bet there were other moments that pulled him out of it, but this is the one that just really stuck with him, right?

Yeah, he's like, that fan, that just isn't right. The girlfriend in question is a shape-shifting light elf from New Asgard. Wow, Dennis. Okay, well, to be fair, I thought I was dating Megan Thee Stallion. But it was one thing to assume that the pilot would be sucked into the fan.

It was quite another thing to prove it. And that's the risk when brainiac nerds watch movies. Not just brainiacs, but... I'm also a huge aviation geek. As a young child traveling, I would always just run to the bathroom to see how the bathrooms were designed. Because in aircrafts, they're always optimized for space, right? And I'm also one of those guys.

There's many people like me. You can just look in the sky, see the silhouette of a plane, and can probably identify what make and model it is. Like, oh, how do you know that's a 787 versus a 777 versus an Airbus? So what do you get when you put an aviation geek in front of an aircraft carrier propelled into the air by drone fans that sit next to a landing runway? A decade without inner peace. OK, I'm just kidding. Well, this movie scene never quite left Edward's mind.

He has definitely kept himself otherwise occupied, not only with a busy career in tech leadership, but... I'm a dad with two young kids. I am a science geek and a nature lover. And I didn't know this, but apparently he's also a Star Trek nerd like you, Ernest. And you guys even like swap Star Trek discussions on a regular basis. Oh, we do. And it's funny because it's exactly like the type of discussions that you would think they are.

Ed will bring up a topic about Star Trek, whatever it is. Either he watched an episode or some random thing. And I always bring in like the philosophical angle of that specific episode that he's talking about it's it's just great yeah ed and i love star trek well i'm glad that you have someone to bounce trekkie nerddom off of because i can't really help you except for the old william shatner days with that amazing fight between trouble for tribbles must be thousands of them hundreds of thousands 1,771,561 that's's assuming one triple, multiplying with an average litter of 10, producing a new generation every 12 hours over a period of three days.

So while Edward stayed busy since he saw this movie in 2012, soon he found himself working for Rescale, which conveniently happened to provide the digital playground he needed to potentially lay his hypothesis to rest once and for all. And then the scene started creeping back up to the front of his mind. So the question becomes, does the pilot get sucked into this fan and die? And I would argue if the pilot's about to get sucked into this fan, there's probably no ejection he or she could do that would give the person the proper escape velocity. The whole thing is going to the fan, right?

And then alternatively, maybe the lift fan isn't sucking in quite as much air quite as fast and eject can actually power its way out, right? So that's where I think we need to do some basic simulation to understand what that looks like. But in order to run such simulations, Ed needed a more extensive knowledge of computational fluid dynamics software than what his career had so far required. I did some rough back of the envelope calculations, but you got to get a model of an F-35 because they had the F-35 on the helicarrier, right, in the movie. And you have to run the fluid dynamics of that at some extreme angles, right, to be able to, like, do the proper assessment. And while there were several people who were already employed at Rescale who had this

capability to be able to run such an experiment digitally, they were already incredibly busy with their own projects. So when Sandeep joined with his background, I thought, wow, finally we'll be able to answer this question. I'm everything sci-fi person. That mystery voice belongs to the name Edward just mentioned, Sandeep Urankar, who had just hired into Rescale as a product marketing manager 90 days prior. And before we dive into his professional skill set, here's a little bit about him off the clock. I have a dad to one daughter, nine-year-old. She's just discovering Harry Potter. So it's borderline sci-fi, it's magic and stuff. So give me Harry Potter. We spend time on that, building Harry Potter Legos and that kind of thing. We have a puppy who's just turning eight months now. A golden doodle, apparently, which from what I hear are awesome dogs.

I don't know about all of like these blended breeds. I've heard of poodles and I've heard of labradoodles, but I've never heard of a golden doodle. But I'm guessing it's a golden retriever mixed with a poodle or something like that. I think that's what it is. Golden doodle, a dog that is bred as a cross between a golden retriever and a poodle. So when it comes to Sandeep, he also went to the theater to watch The Avengers back in 2012. But he told us he had a little bit of a different reaction than Ed had had to the helicarrier scene. I was totally engrossed by the storyline and what was going on. I shut off my brain. I enjoyed the movie and came back home.

But from the moment Sandeep started interviewing for a position at Rescale, I shut off my brain. I enjoyed the movie and came back home. But from the moment Sandeep started interviewing for a position at Rescale, Edward had his eye on him because he knew, based on Sandeep's background, that Sandeep might actually be able to tackle this movie question that Edward had had for over a decade now. So right when Sandeep was hired, Edward approached him with this idea. It was in the very first week of joining, and I'm pinching myself. Is this what this job's going to be about? This is awesome. This is fantastic. So Sandeep dove right in, and it turns out that he was the right person for the job. Being the geek that I am with CAD and things, I had a CAD model of a F-35 lying around.

I had it. Even before all of this. I discovered this was going to be a project. And I love this because he just happened to have a CAD model of an F-35C aircraft just hanging out on his computer already for hobby purposes. I mean, I have to ask you, Ernest, is this a thing that engineers do? Do you have like a CAD model of the Starship Enterprise or something just floating around on a hard drive somewhere? Well, there's more than one Starship Enterprise. Oh, I don't know. But I do have a CAD model of several of the Starships, including several of the Enterprise retrofits. I also have a lot more than just CAD models, but, you know, Trekkies are different. What? You would.

So this is a thing, like a CAD model. It is. And remember that not only do I have CAD models, I have those things that predated the CAD models, which were the blueprint or engineering specifications that got drawn by the, I guess they were artists. I don't know. Back in the old. Are these digital files? No, they're paper. They're print.

Oh, you're talking about the actual. Oh. Right. Because all the way up through next generation, the designers were still using paper. So. Well, I think you've established that while I am impressed by Sandeep having a CAD model, apparently it's not that unique to people in this field. Not in this field.

That's for sure. Well, that's kind of awesome. And I don't know anything about airplanes, but when I look at pictures of an F-35, I will describe it for people like me who don't know anything about airplanes. It basically just looks like a typical Air Force or Navy one-seater jet, similar to the ones I see flying around my house on practice missions from a neighboring Air Force base. There was a scene in the movie where a plane was taking off over the heli fan. Where the pilot was not pulled into the fan to their death.

Lucky them. But for the sake of this investigation, Sandeep had to start with a few assumptions. We're going to assume it was an F-35C and it wasn't going to do an EVTOL type takeoff or a landing. It's going to do a traditional landing. So it's an F-35C airplane, which is, I guess, and Ernest, you know more about this. It's the Navy version of the aircraft that is meant for like takeoffs and landings on an aircraft carrier, right? So it takes off and it

lands traditionally like a jet would, as opposed to like a drone or helicopter type takeoff, right? That would be straight up and down vertically. This is like an airplane that needs thrust to move forward and then it can move into the air. And that ties back to like, so the Navy obviously has a thing called Osprey, which is a VTOL, right? Vertical Takeoff and Landing Aircraft, which ties back to an episode of this podcast.

We had a while back where we talked to somebody who'd had EVTOLs. EVTOLs. The task was to get to the velocity accelerations of forces that we'll experience as a result of the fan so that we could then do some back of the envelope calculations once we have some basic answers from the simulation to determine whether it will get sucked in or it will be slowing down so much that it would come to a halt. Because remember, planes can't do that when landing on an aircraft carrier because they need the right amount of speed to be able to take off again if necessary. We assumed that the F-35C would be going at 150 miles an hour or 67 meters a second. We assumed a lightweight configuration, not too many bombs on board and such. And based on the velocity, we assumed that it would have such. And based on the velocity,

we assumed that we'll have one second to fly over the fan, basically, because it is going fast. And if it can't break free in that one second, it gets sucked in. So we gave it one second for all this dynamics to happen. And to know what to assume about wind velocity, Sandeep turned to a really cool article in Wired, and it was written by this physics professor named Rhett Allain, I think is how you say his name, where this writer analyzed whether or not the helicarrier would even be able to really fly in reality. The professor had calculated 600 meters per second. I actually gave the aircraft much less. I said, OK, fine, assume a situation where it experienced only 134 meters per second air velocity.

So Sandeep actually gave the filmmakers quite a bit of the benefit of the doubt here, right? Reducing the wind velocity of the fan from 600 meters per second, as calculated by the physics professor, down to 134 meters per second. And there are a couple of practical reasons that he did that. So for one, supersonic speeds start around 344 meters per second, which is, you know, translated to 768 miles per hour. And when you're dealing with supersonic, things get a lot more complicated. So he didn't want to get into the supersonic realm if he could avoid it. And even more so, Sandeep reduced the wind velocity because the physics professor who did this analysis in Wired actually realized that in reality, the fans that were intended to carry the helicarrier into the air as they're shown in this movie weren't remotely close to the size that they needed to be to get the job done. It was like almost an order of magnitude in size. Basically, you would look at this thing from above and you'd see all fans.

And then you see a tiny little body in the middle. That's how it would need to look. In fact, in the wired illustration of how big the fans should be, each fan is basically the size of the helicarrier. Each fan is basically the width of the entire helicarrier length. So there's quite a difference. Yes. If you've seen any of the modern commercial drones like the DJI ones or even the little toys you can buy like from Amazon or whatever, you'll notice that like the more weight that's on that thing, the bigger the propellers have to be. If you want to keep the number of propellers constant, right?

So like most drones have four, but you will see some would have six or eight for this very reason because they're carrying more weight. For a ship that was designed to displace water, right? So this thing is made out of iron steel. It's very heavy. And it's basically a floating city with an airport on top of it, right? That's a lot of mass, a huge amount of mass. So those propellers or fans in this case, they'd have to be massive to lift that thing out of the water. Even if you assume they had reduced the weight as much as possible, they'd

still have to be massive. Can you imagine seeing a helicarrier with like four city-sized propellers lifting it into the sky? I mean, it defeats the purpose of it at that point, but sure. I know. That'd be cool. Another assumption we had to make, remember, this is a fictional helicarrier. We didn't exactly know the vertical height between the fan and the deck. And so we looked at a few pictures and we eyeballed it and said, okay, 10 meters. We'll give it that much. And there are other issues with the helicarrier Sandeep had to ignore for the sake of this single investigation too.

One of those factors was what would probably happen to the hull of the entire carrier the moment it lifted from the ocean into the sky. Right. So remember, this is a ship, right? This is not an aircraft. This is not a helicopter. So it's designed to displace water, right? So it's designed to have external force pushing in against the hull and its job is to push out, right? Now it has ballast tanks that hold a lot of water and the hull and keel are not designed to resist the bending force, right? Mid-ship that would occur if a significant part of the ship were above or out of the water. Think of the Titanic when it went down, right? Oh, yeah. So without some kind of seriously crazy structural reinforcements,

which means more weight or some kind of material science magic. So you've now made the whole thing out of, I don't know, titanium or some other like ultra strong, ultra light material. The ship would fail midship when they tried to lift it out of the water and it would just break in half. So Sandeep had to ignore the fact that it would probably snap in two. And he had to remain focused on this one simple question instead, right? Would the plane and its pilot be sucked into the fan if they flew over it? And this is where computational simulation comes in, specifically computational fluid dynamics or CFD. Some things that you just can't do back of the hand, handle calculations and be,

even within ballpark range. That's where CFD software comes in. Engineers have been using it for a long time now, and it's just been getting more and more sophisticated. And it's reached a point, I think, where we're getting very good, accurate answers from CFD. Sandeep mentioned a few other technical specifications. We're going to assume laminar flow. So it's not going to be a vortex rotating flow that the plane is going to encounter because of the fan, right? Yeah, Ernest. Got to assume that laminar flow rather than the vortex.

Whatever the fluid dynamics guy says, I'm not one of those. So whatever he says, I believe it. Yeah, I'm with you. While these terms, I think, are old hat for someone who works regularly with this stuff like Sandeep, I smoothly or in regular paths as opposed to like turbulent flow where fluids undergo irregular fluctuations in mixing. So the internet tells me. So with this definition at hand, it makes sense that assuming laminar flow over turbulent flow would make the calculations at least much less complicated. Otherwise... It's going to be a tornado really that you have to go through, right?

So none of that, but pretty strong wind sucking the plane downward. Which, by the way, I should mention that Sandeep isn't even technically a fluids engineer. He's not? No. In graduate school, I did a bunch of blood flow modeling because I am a biomedical engineer. Mechanical undergraduate, graduate degree in biomedical engineering. So I was doing heart valves and blood flow and both the structure side and the fluid side. So I know how to do it, but 20 odd years had gone by.

I spent a lot of time in structures in the meantime, but hadn't really touched ACFD software for 20 years. So I had to ramp up quickly. But good news, there's been a lot of progress in both technology and usability of COD software, so I was able to ramp up quite quickly. But I still needed a little bit of engineering background. We're still at that point where you need to be an engineer still. You know, you can't still hand it off to a filmmaker. But with his brain and his background and the convenient CAD model of an F-35 that he had stored on his computer already, Sandeep was able to figure it all out, even when it involved a bit of complicated prep work. Then the job became, how can I feed the shape of an F-35C to a CMD software so then it could give me back some answers. So right now, there's a lot of tricky work that you have to do with the

CAD model itself, because the way these technologies evolved in CAE, the computer engineering and computer design, have been evolving separately and have been coming together too, but they're still a little bit separate. So I had to go into the CAD and clean it up a whole bunch so that it becomes friendlier for a CAE software, a CFD software to consume that. Ernest, have you ever worked with a CAD model? And oh, for those who are unfamiliar, CAD stands for computer aided design, which is just a fancy way, I think, of saying physical object in a digital or visual form, like a sort of geometric shape in a computer. Did I describe that right? Yeah, I've actually worked with them many times, but not ever for like work. I don't need them for work, but mostly for like to build things for hobbies or personal

enjoyment, like the Star Trek cab model, stuff like that. Right. Yeah. I've worked with 3D models a tad for like visual effects, but they're really just optimized for visual design, not for mathematic equations. That's the hobby part. That's why I had the F-35 flying around. I'm a Rhino user. A lot of architects use it for shaping and making cool looking shapes. And one of the things you can do is you can really smoothen things and extract the outer shape of the F-35C.

And that's exactly what I did. I smoothened it a lot to cover all the cracks and crevices that can be bad for the CFD solver. The Rhino software he's referring to is officially called Rhinoceros 3D, and it's used across the spectrum of industries from engineering to product design to graphics. It's used for like modifying 3D models, basically. And in this case, sandeep used it to smooth out the digital model of the airplane so that it didn't have like a bunch of unhelpful nooks bumps crevices and so forth that would complicate the calculations which he was able to do because he also conveniently had a copy of rhino on his computer so again the right guy for

this job and when i asked him how long it took to prep the CAD model in Rhino, he said, A couple of weeks because I'm a hobbyist. I'm not a professional industrial designer. I think an industrial designer could do what I did in an afternoon. And others could never do it. And after the CAD model was prepped, there was other prepping that Sandeep needed to do. First, you need to put it in a box. You literally have to enclose it and create the enclosure. Then that enclosure is going to be your control. Everything, all the mass, all the air that enters that box,

now you're going to be able to calculate the velocities off, the accelerations off. And it makes sense to put it in a box because you don't want to be solving for some infinite amount of space, right? The bigger the box, the more complicated the simulation, the more data points, blah, blah, blah. At least I would imagine. Sometimes thinking inside the box is the right thing to do. Got to get that dad joke in there. That's an awesome one, too. Took me a while to come up with that one.

Then you say, OK, I'm going to mesh everything. By meshing, engineers mean they're going to discretize it and break it up into really, really, really small bits and pieces where we can make measurements. And this is the grid-like molding we see in simulation graphics that represent, as I understand it, data points that are able to be calculated. And that's the hard part when it comes to meshing real life objects. They have lots of features, lots of gaps, lots of cavities. So we sort of reduce the complexity and make it more concept-like, which is a fair thing. If anybody is doing these calculations, they're probably doing it on a concept. And when you're in a concept stage, you don't have every last feature figured out.

So we went a little bit backwards in terms of engineering process, but that's fair. What we ended up doing was simulating a concept of a F-35C, where we retained the shape, the control surfaces accurately. So the forces that we would get out of this F-35C would be more or less accurate. To which I say, hooray, Sandeep! Yep, I always have a saying, don't let perfect be the enemy of good. Then it was a matter of applying boundary conditions.

Okay, now let's tell the CFD software how much inlet velocity would be of the air entering that box we created. So that's the velocity of the plane moving forward, right? So give it that inlet velocity would be of the air entering that box we created. So that's the velocity of the plane moving forward, right? So give it that inlet velocity. The downward velocity that we had calculated or assumed on 34 meters per second is what we applied. Again, that much wind would be flowing downwards at that velocity

through that box that we created. Then we let the CFD software do its thing. From supersonic jets to personalized medicine, industry leaders are bringing incredible innovations to market with unprecedented speed and efficiency by using Rescale, high-performance computing built for the cloud. Rescale is an easy-to-use computational science and engineering platform that delivers intelligent full-stack automation and performance optimization that empowers engineers while giving IT security and control. As a proud sponsor of the Big Compute podcast, Rescale would especially like to say thank you to all the scientists and engineers out there who are working to make a difference for all of us.

Rescale, high performance computing built for the cloud. Try it for yourself at rescale.com slash bcpodcast. From that point, the software figured out how much air is hitting the controlled surfaces of the plane given the established boundary conditions. By basically solving for Newton's law of space using certain equations, it was able to calculate the force that the plane would experience, lift, drag, everything. Then reported the downward force. Ready for the results?

I bet they're smashing. That's another one. So good. 1.77 tenders to the power 7 Newtons. That's a huge amount of force to experience as the plane flies over that path. And furthermore... Simple Newton's equations. Force is equal to mass into acceleration. If you know the mass of the aircraft, you know the force is experienced.

You can back calculate the acceleration. And it was going to experience 1,300 meters per second square of acceleration because of this force. Gravity is only 9.8 meters per second square. If you look at Sandeep's numbers, that puts the plane's acceleration due to aerodynamic force at 1,334 meters per second. And if we divide that by the gravitational pull on Earth, which is 9.8 meters per second, we get a g-force of 136.1. And to put this into perspective,

for those who don't have g-forces memorized, most human beings can experience up to four, maybe six G-forces, which they might do on like a really intense roller coaster or something. In fact, the highest G-force roller coasters in the world reach only up to 6.5 Gs, right? That would be the Tower of Terror roller coaster in South Africa, which is not to be confused by the Disney ride. And the Moonsault Scramble roller coaster in Japan. Every other roller coaster in the world is apparently less than 6 Gs, and most max out far under that number. Although, fun fact, according to Roller Coaster Wiki, because again, those are Wiki for everything these days, there was a wooden roller coaster built in 1895 that only stood for seven years at the Coney Island Park in New York. And apparently it reached a whopping 12 Gs,

which is so much force that it reportedly injured a lot of people's necks and it was eventually discontinued due to lack of popularity for some reason. Yeah, we actually, so similar story. In San Antonio, Texas, we have this thing called Six Flags Over Texas. Yeah. And they built this roller coaster there called the Rattler. When they originally built it, this was, I don't remember, in the 80s or 90s, similar thing happened. They had like a very severe angle at a certain point.

Oh, no. That caused a lot of people to have whiplash injuries. And they ended up having to change the angle of that part of the roller coaster. Really? Yeah. And I rode that thing I don't know how many times when I was a kid before they changed the angle. You know what? Given that it was built in the 80s, it was probably a curve track that didn't angle like 45 degrees. Because when a roller coaster turns, the track actually has to tilt.

Right. Or it's just the worst for everybody riding in it. I know this as somebody who plays a lot of roller coaster simulation games. I am an expert. So this roller coaster back in 1895 that had the 12 G's in it, it was called Flip Flap Railway, which totally sounds appropriate. And apparently it's considered the first looping roller coaster in the United States. And just to help listeners picture this thing, it was this wooden track with a little two-seater car. So only two people rode it at a time. And it started with this drop that just went directly into a small circle loop. And that's

where the whiplash occurred. Apparently the G-forces were so extreme because the coaster's loop was a circle, for one, and it was only 25 feet in diameter. So it's pretty small. And that is juxtaposed next to today's loops, which are much larger. Right. And they're also shaped like a teardrop, which apparently is much easier on the G's. The only G-force I know about is G-Force Now. You mean like the NVIDIA thing? But yeah, roller coasters really aren't my thing. When I was a kid, I used to go on them all the time. Yeah.

Loved them. But nowadays, I don't think I'd ever get on one again. Yeah, I'm the same way. I loved roller coasters. And then my body decided I can't handle them anymore. That's pretty much what it comes down to. Like they'd make me sick or feel injured. And I black out with G's. There's some Six Flags Magic Mountains coasters. There was one, I think, called Tatsu. I think it was a laying down roller coaster. And I would just completely

lose my vision for 45 seconds. Like I couldn't see anything. Right. They're not my thing. And then I finally decided I'm like, why am I doing this to myself? Exactly. This doesn't seem like it's good for me. Yeah. I digress. G-force now. That's what's important. G-force now. But yeah, if you cut to today, roller coasters basically across the globe, they max out at 6 G's or six and a half at the very highest. Right. But then if you look at fighter pilots. So they're asked to handle even more than four to six Gs. Fighter pilots have to be able to handle around nine Gs given only for a moment or two, like a couple seconds.

But I still imagine that nine Gs for a couple seconds would probably feel like an eternity if you're even able to remain conscious. Yeah. And I know that in certain aircraft, they have special flight suits that if they have to pull high G maneuvers will like squeeze on their legs and stuff to really, yeah. To keep like the blood from pooling in a certain part of the body. But the F-35 is actually so advanced at this point, because it's a new aircraft, right, that if it needs to do an aggressive evasive maneuver, so say there's either an air-to-air or a ground-to-air missile that's locked onto it, and the aircraft knows that in order to evade that, it needs to pull a high-G maneuver that is more than the pilot can withstand, it knows where that threshold is, where the pilot's going to black out. And it just takes over full autonomy at that point and evades whatever it needs to. And will basically preserve itself until the pilot is able to retake control of the aircraft. Man, our technology is getting so cool.

When it comes to weapons of war, we're the best. Is this a good thing? That sounds terrible, doesn't it? I sound like an arms salesman. This guy is an arms dealer. Made his bank selling weapons to the bad guys. I'm the good one. So if you take that nice little fighter pilot, right, that's able to tolerate 9Gs,

and you station them on the helicarrier, and then you apply a whopping 136Gs to that pilot. Oh my gosh, that's more than 15 times the maximum he should be able to handle, by the way. I'm willing to bet that the helicarrier fan hardly matters at that point. That pilot is going to be in bad shape either way, if any shape at all. I mean, is a blob a shape? Yeah, I suppose anything that has mass has a shape. But in this case, like, is it a recognizable shape? Probably not. Once you have the acceleration, you can calculate the velocity. Once you have the velocity, you can calculate the distance. And so the conclusion was that the plane was going to plunge 67 meters in one second. A 67-meter plunge in one second, or 219 feet in one second,

or the height of a 15-story building in one second. And remember, it was estimated that the fan on the helicarrier was only 10 meters below where the plane should fly. So, sushi. Pilot's definitely going to die. There ain't no other way to swing it. That airplane would be sucked down into the helicarrier fan faster than anyone would probably even be able to consciously process. The funny thing about all this is it's just a movie, right?

They could have easily made people like me and my one other friend that got bothered by this. Just by easily putting the lift fans in the other parts of the carrier, they could have avoided this whole thing. Yeah. So S.H.I.E.L.D. has a whole bunch of different aircraft. You remember they had like that big like military transport aircraft and then they had like those little jets that could land on this whole thing. Yeah. So S.H.I.E.L.D. has a whole bunch of different aircraft. You remember they had like that big like military transport aircraft. And then they had like those little jets that could land on top of it.

Remember? So it's like, why do we even have them? Well, that's what I was trying to figure out. What's the need for this? Who knows, right? You could maybe make the argument that it's so non-S.H.I.E.L.D. aircraft can actually land on it. Is there any purpose for that? So that's what I was trying to figure out.

I was like, I thought that S.H.I.E.L.D. had aircraft that didn't need runways, so to speak. Yeah, that's like VTOL. Right. But this thing was obviously in the water and had the ability to go into the air. So they thought that this was necessary. Right. So there must have been a reason. It was also created in 1965 and maybe they didn't actually think that far ahead. You're right. This was created in 1965. And obviously the Avengers movies, because they're made in 2007 and later, have some additional things that didn't exist in the original comics. So the two don't really align, but I can understand like this was a cool concept back in the sixties for them to put in here. So I think that's how this happened. Right. Cause they were thinking like, what do we have in the 60s that isn't an analog to, you know, something amazing?

It's like, oh, we're going to put an aircraft carrier, but it can fly. Yeah. That's awesome. Yeah, the writers of 1965, this was the best idea ever. 2023, maybe not as impressive. No. But regardless, Sandeep did get some results on this hypothesis. And as far as the tools that he used. I happen to be familiar with Gradle CFD. It's a nice tool and has a nice

sort of wizard-like user interface. It walks you through all the necessary steps for doing a CFD simulation. So that was the choice of software for me. And it generated nice visuals because it has a strong post-processing capability as well. So we were able to showcase our final answer nicely. Of the many softwares that are available on Rescale on cloud, Videl CFD is one of them. So it was available as the Elastic Workstation, the way to use. Meaning I could spin up a workstation on cloud from anywhere.

And as it so happened, as Sandeep was starting to near the end of his investigation... It was a weekend and I was away at a lake house and my wife decided to sleep in. And I was like, huh, why don't I just finish the simulation? So from the lake house, I was able to quickly finish the simulation, submit it, and actually went out on the kayak afterwards. Knowing that the computational power he needed to solve this wasn't extreme, right? He picked out eight cores on a pretty standard architecture on the Rescale platform, submitted

the job from his lake house, and then went out on the kayak and enjoyed time with his family, and then checked his results on Monday when he was back in the office. Once the more text-based solver runs, where you can even spit out these lines and numbers almost in text format. Once you're done with all that, when you know that everything is run properly, now you have to visualize this. And not everybody's laptop is gonna have good quality video cards to be able

to do good rendering of this. And what I had accessible to me was the really high-end graphics cards and GPUs that I can then use to run a very specialized computational load, specifically for visualizing this. I generated a photorealistic image from a lake house because I had access to those GPUs and those graphics cards. The Rescale Elastic Cloud Workstation is essentially a virtual machine in the cloud that you log into. You could play video games on it. You could do whatever you want on it, right?

But it just so happens, because it's Rescale, that we give you access to all the science and engineering software that are available, and we give you access to run that software on any of the specialized computing architectures available in our catalog. Right, so, Sandeep's next project was some structural analysis.

He gets to fire up whatever software that's best for that, using the hardware that's best for that, in whichever geography he wants to be in. And then the following week, he was doing a weather simulation for a different project. Exact same thing. So while Sandeep used a standard hardware choice to submit the job,

he made a different choice for the visualization. When I came home and I said, okay, now I'm to do a visualization, generate a photorealistic image and possibly a photorealistic animation, I chose a higher-end GPU configuration. Because since Rescale can connect you to pretty much whatever software and hardware you want through the power of the cloud, you can optimize your architectures according to the job you're trying to do.

If you were to do this on-prem premise, then you would end up buying another machine just to do visualization and just to do the solver parts. You end up with two. When the flexibility I have here is that once my project is done, I can go and pick another architecture that suits my needs for that project and then use it on demand and then close, shut down.

And maybe two months from then, I would have yet other requirements for yet another project, and I could just use it on demand. There's a few parts to this. The first is the ability to choose any software you want. So initially, when we were playing around with this, I was experimenting and trying to get this done on Ansys Fluent. Then we kind of thought maybe we'll do it on ANSYS Discovery.

And then when Sandeep kind of got on board, he was more comfortable with some other tools, so we went in that direction. What you didn't hear us talk about throughout this whole conversation is, oh, what software do we have access to? And how do we set up that software? Because we just assume that whatever we need is available, and we'll just get at it.

And the same is true for hardware. You know, notice in our conversation on this today, that software because we just assume that whatever we need is available and we'll just get at it. And the same is true for hardware. You know, notice in our conversation on this today, we did talk about, hey, we're going to use a GPU when we need it. We'll use something else when we want that. But it's not like there's a whole IT team behind this to help Sandeep set up his infrastructure. We just have the Rescale platform taking care of all of this. And so just having that flexibility to, I think one person I spoke with said, Rescale is like Jay Leno's garage, right? Drive any car whenever you want and just switch to another car when you want to.

So we basically got to do that with all the software and hardware that's available in the cloud. That's kind of one angle. The other angle is just being able to do it from anywhere. Right? And I think it's like many people take it for granted that we should be able to work anywhere, but for some organizations and some companies, people are still working using the workstation machines

they have in the office, and they have to go in there to use those machines, right? So we're providing that level of flexibility on what tools people use, where people work from, so they can really just focus on solving the question they're trying to answer, whether it's something serious, like solving a real health crisis, or just tooling around with the Avengers helicarrier. And Ed and Sandeep are looking to run more fun projects like these

using computational simulation to solve for interesting topics that people are just curious about. In fact, I think they have another blog out recently that's all about the physics of drafting and like whether or not it actually helps make driving greener. I think it makes it a lot greener if you have like a little hook you can launch from the front of your car to just lash onto the back bumper of the 18-wheeler. And then you're super green. I don't think that's just drafting, though.

I think that's pretty much them pulling you at that point. It sure is. Again, I happen to have a cat part of my own car. Cat model. How many cat models does he have? And I know that they were even talking about trying to simulate whether or not it's possible to actually nuke a hurricane. As we may have heard claimed not all that long ago by a certain public figure. You can nuke it. Nothing prevents you from nuking it. Just remember that

these things draw their energy from heat. So the outcome might not be what you think it's going to be. Well, we need to simulate it to know for sure, though, Ernest. Oh, of course, of course. But if anybody listening to this has an idea, something maybe interesting that you'd like to see solved with computational simulation, please literally let us know by sending an email to info at bigcompute.org. We will pass that message on to Ed and Sandeep

and see what they come up with. So they're actually looking for these ideas. So please send them in. One of the reasons we wanted to do this project and the other projects is that we're not trying to tell the world that analyzing the helicarrier through CFD or nuking a hurricane

is solving any real problems for the world, right? But what we are trying to make sure people understand is that increasingly, if you have any curiosities about anything that's scientific or physics based in nature, there are tools out there that are easily accessible and there are automation platforms like Rescale that helps you to answer those questions. And Ed had some particular advice for filmmakers. If you are making a movie, please try to use common sense and get the right consultants involved to make sure the story is believable as possible.

I'm like totally torn on this, right? So the pragmatic side of me totally agrees with Ed. Well, the creative filmmaker in me has these flashbacks to science people who would complain about the movie we made because maybe we, I don't know, used a bird sound effect that wasn't native to the location we were showing in a scene or something. It's funny because like I can see both sides of that argument. But as someone who has run P&L in different organizations in my life, I understand that everything revolves around money and budget. Right. It's like how many consultants can you afford? How many can you afford, right? Especially if like you're the foremost expert on the yellow eyed warbler in Southern Australia, right? Bird.

Okay. How much is it going to cost to get this person to tell me about the bird sound, right? So there's just too many, right? There are just too many. So is that a real bird? No, I just made it up. But watch, someone's going to message. Someone's going to write in.

Actually, there's a red-eyed warbler that is native to Southern Australia. That's what's going to happen to us. But no, that's the point is like, I get it on the big ticket things. You should have consultants on things that have to deal with like maybe cultures, traditions, or things you're not familiar with. But like, is this a native species of animal to this subcontinent that we're putting the noise effect of this movie for? No, that's just not going to happen.

Yeah. And I guess the answer is to just try to get as much right as you can with the budget that you have, maybe. I mean, it takes more effort, more spend a lot of times. This is the reason I love like those low budget, like B movies, because- Those movies are terrible. Because they have no budget.

So there are no consultants. They just have to do it themselves. And there are no expectations. Like no one's going to cry because- That's the thing. It's the expectations, right? It's like- They're low.

Everybody expects them to suck. And so they do. The fact that they released the thing is a success story, right? It's like. They're low. Everybody expects them to suck. And so they do. The fact that they released the thing is a success story, right? Like that's. Can you imagine somebody calling in and being like, there's not actually sharks and tornadoes. Oh, I'm sure somebody did it. Technically, a shark could not survive in a tornado. Exactly.

That's a great movie series, too. It's stupid, but it's great. It's so dumb. If you guys remember the beginning of Armageddon, that movie. That movie's terrible, by the way. It's not a good movie. Just for the record, it's awful. So Ernest probably loves it, is my guess. Well, I think at the very beginning of the movie, it talked about many, many millions of years ago,

some comet hit Earth, right? And it didn't show Pangea. It showed like the continents as they are today. So you immediately know this movie, even before it starts, the science is just going to be total bullshit, right? And then that, of course, is later confirmed when they decide to dock with a mere space station. And somehow, just before they choose to dock, they choose to spin it. Which is, you know, I'm not an aerospace engineer, but that's got to be the most idiotic thing you do before you try to dock with something. It's just to get it to start spinning. Yeah, Armageddon, first of all, was an amazing movie. Anything Bruce Willis is in.

No, it was not. Anything Bruce Willis is in is good, right? That is not true. But Armageddon was not. Anything Bruce Willis is in is good, right? That is not true. But Armageddon was awesome. If for nothing more than Aerosmith released an awesome single just for that movie. I don't want to close my eyes. Yeah. Don't want to fall asleep.

That was it, right? That's the one. Yeah, I love that movie. I know it's, you know, oh oh wait steve buscemi come on like yeah he's in it steve buscemi billy bop thornton's in there owen wilson michael clark duncan come on my gosh affleck live tyler bruce this was like an a yeah it was it was a huge movie when it was released but that doesn't mean it was good who directed it though who was the director is Is it Michael Bay?

It better not be. Find out. I'm looking. I'm looking right now. It's going to be Michael Bay. Please, no. I'll have to take back everything bad I ever said about him. Director Michael Bay and producer Jerry Bruckheimer. It's Michael Bay and you like

that movie. What? Well, alright, so then we've now determined that Armageddon was the peak of Michael Bay and you like that movie. What? Well, all right. So then we've now determined that Armageddon was the peak of Michael Bay's career. And it's been downhill ever since. That was a very low peak, my friend. Call it what you want, right? It was like an anthill. It was a budget of $140 million that grossed $553.7 million at the box office.

So that's why he kept making movies. Yeah, that's what happened. He ended up, it was the highest grossing film of 1998. However, it received mixed reviews from critics. How? How could that even be possible? Mixed reviews on Armageddon? I mean, you have like every A-list actor who was popular in 1998 in one film.

Seriously, that is an A-list cast. And it still only cost you $140 million. Yeah, things are so different now than they were back then. $140 million with an A-list cast like that? No. Yeah, that's The Rock's take by himself. Yeah, I was going to say, that's like one actor now. Yeah.

Like one A-list actor. Like one Dwayne Johnson. That's what you would get for that money now that i think back on this the reason was because it was 1998 michael bay didn't have access to all the crutches he uses today but it was still bad you're arguing like it was good oh it was amazing no it was amazing it look i'm not gonna say it deserved like awards or anything like that. It was not like a cinematic masterpiece. Lord of the Rings, Armageddon.

As for Sandeep, well, he acknowledges that today an engineer like him is necessary to do these kinds of computational investigations, who would then need to pass the conclusions on to a filmmaker, per se, in the future, he believes that a day is coming where the filmmakers will be able to access that kind of conclusion on their own. What AIML will enable is the wrapping of that physics into a form, which is much more palatable to someone who doesn't have that expertise, that engineering training or software experience, all of that. And that's really exciting. And so that more and more people can start making more intelligent engineering decisions.

It could be for films or real engineering use cases, or maybe a hobby project at home. It doesn't matter which. I think more of us can start consuming the insights of physics. And I think Rescale is doing its part. We have the core types and architectures that are necessary for those kinds of analysis. We're starting to look at those kinds of workloads. And that's getting really exciting. That's going to do it for this episode of the Big Compute podcast. To read Sandeep and Ed's blog that covers the details behind this helicarrier investigation, you can find it at rescale.com slash blog,

or you can head over to the episode notes page on bigcompute.org, where we will also link you to the physics professor's blog as well as the movie clips and content that we've mentioned here on the show. And you can support us by leaving a rating or review on Apple Podcasts, Google Podcasts, or wherever you get your podcasts. Or better yet, telling a friend about us. Yes, that makes a difference. And with that, thank you for joining us today. And remember everyone,

configure multi-factor authentication everywhere and always use 321 Backup. Stay safe out there.