IBM's original name for "read only memory" was actually ROS (read only store). This was the name used on the System 360 machines. They were actually sheets of plastic with conductive things.
Babbage was a top-class mathematician, once Lucasian Professor of Mathematics at Cambridge, as were Newton, Dirac and Hawking (and Lt. Cdr. Data). He knew about binary, but he knew denary (base-10) was more compatible with round gears.
Yes, just came here from a video on the Difference Engine that explains that he did consider a number of bases for that. But larger bases are just much easier to work with with geared computation.
Mechanical machines that were theorised and built, not to far back in the past, operate in such a similar way to their modern counterparts that it makes you realise the basis of the technology used in modern computing is still very much in its infancy.
Well, there's Blaise Pascal and his decimal adding machine which was the basis for Babbage's Difference Engine, which in turn led to the Analytical Engine. There's also John Napier, who's discovery of logarithms let to the need for the Difference Engine. However, neither of these two had anything Turing complete other than what was between their own ears.
Judging by the size, it would be more like when gulls get caught in jets. To think, programmers might still be talking about debirding their programs if Babbage had completed this!
Second computer I used was an ICL and the central processor was called a mill. Each program was supposed to minimise "mill time". Interface was a Telex type machine which recorded my input on paper. There was one error message, it was "?". To make a program work you had to be a half decent typist. Some of the users had probably never even seen a keyboard except on a typewriter as they walked past. Mega-frustration. I still can't type properly.
Thanks for reminding me! Yes my first 3 computers were :(Ferranti) Mercury, (English Electric) KDF9 and then (ICL) 1906A and I *do* recall in looking at the machine usage stats at the bottom of my line-printer output (which determined the "cost" of the job) seeing the phrase "Total Mill Time"
What a magnificent machine! It really gives one a perspective on just how incredible the technology that we take for granted truly is. Thank you for another fascinating video!
I live in Texas and used to work at a hotel where I met an elderly Mexican tourist who was an admirer of Babbage.He introduced me to his work and said he even went to Great Britain to visit either his home or the museum about him.Its been so long ago it's a little fuzzy.This took place in March of 2005.
Can you imagine how they'd have to deal with component failures on this machine? It's bad enough with the vacuum tubes in things like ENIAC, which would all have to be shut down and left to cool before someone had to crawl in and find the blown tube, but that would be a relatively simple proposition compared with hunting down a gear or ratchet or spring which had failed, and they would fail at an alarming rate, given the complexity of the machine and the sheer friction generated in use. The room in which this would have been housed would have been incredibly noisy and hot.
Something tells me that just getting to the bad part would be 1000x harder than (say) replacing the failed monitor wiring harness on a laptop. Imagine tearing down a 50 wheel register to replace wheel #31.
Thanks for advertising Sydney Padua's book, I got it for myself for christmas based on the drawing in here (started reading it early but I think hot cocoa, blanket, fire, and book will happen a lot in the near future!).
I wish Computerphile upload more up to date videos. As technology is improving literally every day we need to be fed by the current and upcoming news. Of course some history is definitely curious to watch, the todays tech news are lacking massively in this channel.
I just recently got into computer science and I am kind of late to the game (I am 23 and already have a degree) but these videos inspire me to learn more. Thank you.
With the earlier Difference Engine, which has been recreated and it works, Babbage was very aware of the time it would take to ripple a carry through dozens of columns of gears. He created the brilliant "Anticipated Carry" mechanism, where any gear reaching "9" would trip a flag. When the number reached all 9s, they would all flip at once, in a single clock cycle, without rippling carries through the mechanism. I imagine he would have had a similar mechanism to place a number into memory without it having to propagate down the entire length of the store. He was obsessed with two things: accuracy and speed. As regards the former, the machine was able to detect an error and basically crash itself, and as regards the latter, he put numerous clever speed-ups into the Difference Engine.
I have not seen any error analysis of the Difference Engine. I made some investigations using Decimal in Python and my findings are that a restart has to be performed quite often. I calculated log10(x+1) with 31 digits and seven differences deep. To produce six digits I had to restart every 200th number. Would be nice to see this subject covered in Computerphile.
Charles Babbage must have been an incredible genius. He made all the plans for a fully mechanical computer long before electronic computers would be developed. Likely one of the main reasons this Analytical Engine was never built is that it would have cost an enormous amount of money to build such a large complex mechanical machine like this in the mid nineteenth century. Even today to build this mechanical computer would cost an enormous amount of money and would only be useful as a museum item, because we have modern electronic digital computers.
Babbage was brilliant, even if he had never worked on a computer. He made important contributions to statistics, cryptography, engineering, and mathematics, but he considered his Babbage machine to be his life's work. And the analytical engine would have been his second attempt after failing to produce the first one. I don't think it was just an "accident" that his machine really would have worked. But it was truly before it's time. The cost would have been absurd and the benefits questionable.
I was lucky enough to see a presentation Sydney Padua made when she was in London. She went as far as to reproduce in 3D CAD software how certain parts of the machine would work. When you look at just a small step of the machine it becomes apparent that it's beautifully simple in some ways, however when you put it all together it's likely that the machine would crash constantly due to how many moving parts there were. Babbage was aware of this of course and planned to have a bell ring any time it jammed.
A quick look at Charles Babbage's biography suggests that he was indeed a complete genius, and he gave his gift to the world willingly. I like to think that perhaps, at night, he quietly, humbly trembled at the existential implications of infinite computational power. (The more likely reality is he just saw them as numbers. An exact analytical machine for pure math lovers.)
From what I know of History through podcasts and the likes, Babbage did indeed not really see the implications of what he'd built, but Ada Lovelace did absolutely see the far reaching implications of general-purpose computing, and she'd spend a lot of time speculating about it. Unfortunately this thing sorta got stuck in development hell. Could you imagine where we'd be if they'd got this thing built?
I have an old mechanical calculator which could add and subtract by cranking a handle, and then it could also multiply and divide by doing certain operations (made by a company named Walther, which also produces weapons). So the mechanical side of things is partially there. OK, it could only store 2 numbers, one of which was overwritten by the result, but it is already a marvel of mechanical engineering that nobody will be able to understand nowadays (unfortunately, it doesn't work right any more). Based on this type of rotating barrel mechanism, it would indeed be possible to build such an analytical engine, but of course it would need much more than that, like the mechanics for storing numbers in the "permanent" store and so on.
Done. I saw the hardware at the Computer History Museum in Mt. View, California, about two years ago.. It was on display for a year prior to being shipped to another museum.
That is called the difference engine. While they were both designed by Babbage, the difference engine is much smaller than the analytical engine, and had a specified purpose; namely to tabulate polynomial functions. I'm not sure if it is Turing compliant (it most likely is not), but it certainly isn't a general purpose computer like the analytical engine. It is a right step in the direction of eventually building the full analytical engine, though. Maybe someday we'll get to see the thing in operation, and perhaps even run some of Lovelace's programs on it.
I watch this video every once in awhile. Professor Brailsford is always a great person to listen to. I find myself wishing for a deeper look at the instruction set the Machine would have used. I keep coming back to the idea that the instruction cards could be bigger, contain more instructions. There is talk of the logic the machine could process, but not the instructions. I'd love to see a video about the minimum instruction set needed, and also how it could be extended to add more speed and capability. I'd like to see an instruction on the Program cards that asked the machine to advance to the next card, for an additional instruction set. The idea that an operator would swap out the stream of cards for a subroutine for Square Root or some such seems inefficient. Perhaps the Machine could have had a dedicated sub-program built in, and called when needed. That and a few other things that I'm pretty sure Babbage would have realized as improvements. Anyways, thanks for all the videos, always.
Today, we have MEMS - microelectromechanical systems - using the same technology like in the production of semiconductors, tiny mechanical components - wheel, cogs, rackets, etc. can be built to very small sizes and integrated with electronic components; in theory this technology could be adapted to make only microscopic mechanical parts so Babbage's computer could be built but it would be hundreds or thousands of times smaller...
This is probably THE most steampunk thing possible. A fully theoretically realised, Turing-complete, steam-powered computer that's the size of a locomotive I mean, Sweet Jeebus I hope this thing is built one day. This would be amazing to see in operation
I seem to remember my early adventures into computing in the early 80's was via a Teletype connection to an ICL mainframe. Once you had fanished your session it told you how much "Mill Time" your work had used and your storage use was in Buckets :-)
I know this machine was never completed, but I wonder how much would it cost to build a mechanical computing machine like this today with modern methods?
As a computophile, I read that Mary, the Bitish poet Lord Byron's daughter, was the first computer operator who said, "To operate, I went into the room. Such a huge machine it's!"
If you want to better understand this machine, I recommend you learn about the Differential Engine first. This engine is the same basic idea, but where the memory and the processing are separate components rather than shared. Sidney Pardua has also uploaded good videos in her RUclips channel telling details of both machines.
The book that basically started the whole steampunk thing is set in a Victorian England where Babbage's designs were made. I think it was called The Difference Engine.
Bit-Bit-Jump manages to be Turing complete without a conditional jump(if statement) I think this demonstrates that as an alternative to conditionals, being self modifying is sufficient for Turing Completeness.
It sounds like 4:27 the "store" is a faster form of memory, we computer engineers called a "register" 4:40 the "Mill" sounds like what we call the ALU or arithmetic logic unit. 5:04 looms sound like the hard drive or tape storage equivalent of the modern computer.
Do you think pointless genius are a new thing (string theorists comes to mind) well, it isn't. Babbage's inability to finish one thing, and lovelace writing code for a machine never built instead of goddamn building one. Babbage, because inventing is useless if no one builds it, lovelace, because no point making the best ammunition possible for a gun that will never be made, and both because science and invention is fun!
lloydgush They mightn't have actually been built but later real implementations certainly drew on the theory that this and other such designs examined... ☺️
It's not pointless. He needed a lot of money to build it. And the best way to get investment was (and still is) to show that you already have detailed schematics and something to run on it. However, he was never able to persuade investors that it would actually work which, given its cost which the investors were unable to even estimate at that time (because they were not sure if it was possible to build with nexessary tolerances), only a small part of the mill was built in 1910, fourty years after Babbage's death.
To be fair, if this was his first such machine, I think finding investors would have been easy. It's just the problem that he had the reputation of not being able to deliver, as his Difference Engine was never finished. That, and he was just bad with money in general, which meant the investors he did find didn't actually matter that much.
I just adore the whole concept. But i wounder, could one argue that the "store" was more like cache, not ram? The Rom would be the jacquard cards. Ram would have been a printable version of the same cards? A HDD would probably have been more of a physical storage of some sort.
Thanks to the Professor for being the only person I've ever heard pronounce the name "Lovelace" correctly. However sad it may seem, it is supposed to sound like "Loveless".
Has anyone ever made a complete 3D / CAD-Based model of the analytical engine? Maybe a fly-through so you can see some of the different parts? I've founds lots of videos exposing small parts of it but that shows an artist's interpretation of the full machine. Let me know if you know of one. Thanks!
I was thinking I can just about keep up with what's going on with the machine, but after halfway it was just impossible. Reading the cards mechanically and storing their information? Transferring the information to the back of the machine? Systems to add things together with a preset thingy that separates numbers from adding. It's like recently I've thought I almost start to understand what a very basic computer does on low level, but when you present this engine, it's all gone.
There's actually a group that's trying to build it. They're still in the process of trying to figure out Babbage's plans so they can build a virtual model for testing.
Before I clicked to watch I thought "Will they actually pay respect to Ada Lovelace?" Respect was paid in the first 10 seconds, well done guys! :) Kind regards, Meta Custom Computers
Can anyone in the comments help? For some reason, when sped up, sections of youtube videos that have little to no audio, eg the part where computerphile is being typed, create MASSIVE lag. It only happens on my chrome book. PLEASE HELP
How do you do a non-destructive copy of one register to another? Racks and pinions I get, but wouldn't you need a way to sense when two 50 digit registers hold the same number? If you just decrement one register while incrementing the other until the first hits zero, you've destroyed the original.
?how many registers had the Mill-CPU ?were the registers named ?were there specialized registers... ?length of registers (in decimal-state-units... of course) ;-)
Its really bizarre to me this vid was released in 2016... And only has like 100,000 views... I think people fail to appreciate that the analytical engine was turing complete. It may have been decimal/mechanical but Babbage was so far ahead of his time on this one its not even funny. This is the grandaddy of computers wheres all the so called "computerphiles"?
How fascinating. A steampunk engine might break the computer. I don't know, steam seems very crude and uncontrollable. I'd love to see even a small version of of this running.
***** That's not actually true. Increased clock speeds don't necessarily increase heat output. Higher operating voltages, however, do, and to stably use higher clock speeds, higher voltages are generally necessary as well. The manufacturing process also makes a big difference, with smaller transistors generally requiring less voltage to perform equally. The reason modern AMD CPUs output more heat than their Intel counterparts is that AMD's offerings are based on larger manufacturing processes, commonly 32nm compared to 14nm.
If this thing was built and improved upon over time then most likely computers would be far more advanced now. But I suppose you can thank WWII for the advancement of that, unfortunate as that sounds.
I don't think so. Remember this was pre-Turing era. At that time it was not known that every solvable problem has algorithmic solution - aka. they didn't know computers can in principle actually compute anything. Also, modern computers are so powerful because they take advantage of technology that did not exist back then (specifically electronics). In terms of hardware computers would be at about the same level as today.
The limits are actually software, you can keep just adding more and more cores to the cpu. Electrical limits are also a reality, at least for the common person, but businesses/governments could mostly get around that. If binary computing at it's very nature limits programming possibilities, as opposed to say quantum computing, that I do not know.
I wonder machine computation being available 100 years before it actually became available would have altered the course of science and technology such that we would be much more technology advanced in the present. In principle, atom bombs and nuclear reactors could have been designed with even the tiny computing power made available by the analytical engine.
PleaseDontWatchThese I use a computer only for things that can be displayed visually like games and videos. Or research, which isn't possible without the internet, or word processing, which would be pointless on that machine. Everything I can think of would be impossible or impractical on a machine like that.
Just think of the machine like a really big scientific calculator without graphing, such as the ti84 that is very popular in high schools. It's primary use would be making computations that are guaranteed to be correct.
+ Ybalrid Actually, nautical tables for use by ships at sea. That was how he pitched his grant proposal to the Crown. There were for such tables in those days, as mentioned in this video, tables of corrections issued; then tables of corrections to the corrections; then tables of corrections to the corrections to the corrections! Babbage's vision was to automate the whole thing, so that the resulting printed tables would be error-free, from the outset! And yes, pure mathematical tables - logs, trig functions, etc. - would also be well-suited work for this contraption.
Of course. With a kilometer-long stack of punch cards, a ten kilometer long store, and one frame per month with a resolution of 10x10 punch card holes. And a couple of tonnes of fuel per month for the steam engine.
I just love listening to Professor Brailsford's voice. It's like a kindly uncle reading us bedtime stories about computers. :)
YES!
I know he just seems like he genuinely loves explaining computers.
I got an amazing picture with him two years ago :p
Must have been you who said "Hello" to me on the train from Lausanne to Montreux ?!
Nice!!!!
IBM's original name for "read only memory" was actually ROS (read only store). This was the name used on the System 360 machines. They were actually sheets of plastic with conductive things.
Babbage was a top-class mathematician, once Lucasian Professor of Mathematics at Cambridge, as were Newton, Dirac and Hawking (and Lt. Cdr. Data). He knew about binary, but he knew denary (base-10) was more compatible with round gears.
Yes, just came here from a video on the Difference Engine that explains that he did consider a number of bases for that. But larger bases are just much easier to work with with geared computation.
F___k Dirac. I've had nightmare's in uni because of him
"racks and pinions like you've never dreamed of in your worst nightmares" made me smile
I was reading this comment exactly when he was saying it.
Mechanical machines that were theorised and built, not to far back in the past, operate in such a similar way to their modern counterparts that it makes you realise the basis of the technology used in modern computing is still very much in its infancy.
I really like this guy and how he explains things. You can tell how knowledgeable he is, but he has a great sense of humor too.
Fascinating where we've been, how far we've come, and how far we've still left to go...
Sci-Twi Nobody likes a pessimist.
What is the point of building such a machine, if neon LEDs were not invented yet to make custom lighting of it?
They must have been time travellers
That's why it was never built.
This is basically a steampunk enthusiast's wet dream.
the first steampunk book was about the analytical engine, its called the difference engine.
I'm here so I can draw better steampunk sooo :D......
Finally I get the link between steampunk and the binary system.
Man imagine Babbage's ideal world of the future. Now *that* would be an amazing sci-fi novel
@@oraz. Kennethsjsdjddjjsjjshahsnnannabababsbsbbs
Always love the videos with Professor Brailsford
A thing that Prof. Brailsford considers old. You don't see many of those.
well, in computer science you can't really get any older than Babbage
Well, there's Blaise Pascal and his decimal adding machine which was the basis for Babbage's Difference Engine, which in turn led to the Analytical Engine. There's also John Napier, who's discovery of logarithms let to the need for the Difference Engine. However, neither of these two had anything Turing complete other than what was between their own ears.
Jonathon Payne If we forget Turing-completeness, why not go back to the (very impressive) Antikythera Mechanism!
:)
I bet debugging this would have been fun :P
Any bugs would be just crushed by the gears.
That's actually where the word bug comes from...
Bugs getting fried in the electronics of ancient, room sized computers.
Judging by the size, it would be more like when gulls get caught in jets. To think, programmers might still be talking about debirding their programs if Babbage had completed this!
actually no...... it was Admrial Grace Hopper that pulled a moth from a Relay. they were still mechanical at that point.
@@midnightrizer correction: electro-mechanical
2:00 Cats have been around computers from the very beginning.
Second computer I used was an ICL and the central processor was called a mill. Each program was supposed to minimise "mill time". Interface was a Telex type machine which recorded my input on paper. There was one error message, it was "?". To make a program work you had to be a half decent typist. Some of the users had probably never even seen a keyboard except on a typewriter as they walked past. Mega-frustration. I still can't type properly.
Thanks for reminding me! Yes my first 3 computers were :(Ferranti) Mercury, (English Electric) KDF9 and then (ICL) 1906A and I *do* recall in looking at the machine usage stats at the bottom of my line-printer output
(which determined the "cost" of the job) seeing the phrase "Total Mill Time"
What a magnificent machine! It really gives one a perspective on just how incredible the technology that we take for granted truly is. Thank you for another fascinating video!
I live in Texas and used to work at a hotel where I met an elderly Mexican tourist who was an admirer of Babbage.He introduced me to his work and said he even went to Great Britain to visit either his home or the museum about him.Its been so long ago it's a little fuzzy.This took place in March of 2005.
Build that same wall already
Can you imagine how they'd have to deal with component failures on this machine? It's bad enough with the vacuum tubes in things like ENIAC, which would all have to be shut down and left to cool before someone had to crawl in and find the blown tube, but that would be a relatively simple proposition compared with hunting down a gear or ratchet or spring which had failed, and they would fail at an alarming rate, given the complexity of the machine and the sheer friction generated in use. The room in which this would have been housed would have been incredibly noisy and hot.
Something tells me that just getting to the bad part would be 1000x harder than (say) replacing the failed monitor wiring harness on a laptop. Imagine tearing down a 50 wheel register to replace wheel #31.
Turn it off and turn it back on again
Just put some WD-40 inside and you’re set.
Thanks for advertising Sydney Padua's book, I got it for myself for christmas based on the drawing in here (started reading it early but I think hot cocoa, blanket, fire, and book will happen a lot in the near future!).
I wish Computerphile upload more up to date videos. As technology is improving literally every day we need to be fed by the current and upcoming news. Of course some history is definitely curious to watch, the todays tech news are lacking massively in this channel.
I just recently got into computer science and I am kind of late to the game (I am 23 and already have a degree) but these videos inspire me to learn more. Thank you.
Professor Brailford is excellent at explaining these things, I always think these concepts are so confusing, but when he explains it it makes sense!
With the earlier Difference Engine, which has been recreated and it works, Babbage was very aware of the time it would take to ripple a carry through dozens of columns of gears. He created the brilliant "Anticipated Carry" mechanism, where any gear reaching "9" would trip a flag. When the number reached all 9s, they would all flip at once, in a single clock cycle, without rippling carries through the mechanism.
I imagine he would have had a similar mechanism to place a number into memory without it having to propagate down the entire length of the store. He was obsessed with two things: accuracy and speed. As regards the former, the machine was able to detect an error and basically crash itself, and as regards the latter, he put numerous clever speed-ups into the Difference Engine.
I have not seen any error analysis of the Difference Engine. I made some investigations using Decimal in Python and my findings are that a restart has to be performed quite often. I calculated log10(x+1) with 31 digits and seven differences deep. To produce six digits I had to restart every 200th number. Would be nice to see this subject covered in Computerphile.
Charles Babbage must have been an incredible genius. He made all the plans for a fully mechanical computer long before electronic computers would be developed. Likely one of the main reasons this Analytical Engine was never built is that it would have cost an enormous amount of money to build such a large complex mechanical machine like this in the mid nineteenth century. Even today to build this mechanical computer would cost an enormous amount of money and would only be useful as a museum item, because we have modern electronic digital computers.
Money is not the reason. Come on.
Babbage was brilliant, even if he had never worked on a computer. He made important contributions to statistics, cryptography, engineering, and mathematics, but he considered his Babbage machine to be his life's work. And the analytical engine would have been his second attempt after failing to produce the first one. I don't think it was just an "accident" that his machine really would have worked. But it was truly before it's time. The cost would have been absurd and the benefits questionable.
"Computer Earth" Narrated by Professor Brailsford. A rival for Planet Earth and Cosmos. Should it happen?
buddhavskungfu
I am not British but listening to Professor Brailsford reminds me of my own grandfather telling me stories when I was younger.
Someone needs to make this before I die so I can see it. I'm too inept to do it myself.
You're in luck. There's a group in the UK that's currently trying to figure out Babbage's plans so they can build it.
@@Roxor128 7 years later:
I was lucky enough to see a presentation Sydney Padua made when she was in London.
She went as far as to reproduce in 3D CAD software how certain parts of the machine would work.
When you look at just a small step of the machine it becomes apparent that it's beautifully simple in some ways, however when you put it all together it's likely that the machine would crash constantly due to how many moving parts there were.
Babbage was aware of this of course and planned to have a bell ring any time it jammed.
To drive such an exhibit in perpetual operation with a waterwheel at an old millhouse somewhere would be a very interesting endeavour to undertake.
This machine is in a chart, at the very beginning to demonstrate Moore's law. Brailsford is really nice to listen to.
A quick look at Charles Babbage's biography suggests that he was indeed a complete genius, and he gave his gift to the world willingly. I like to think that perhaps, at night, he quietly, humbly trembled at the existential implications of infinite computational power.
(The more likely reality is he just saw them as numbers. An exact analytical machine for pure math lovers.)
From what I know of History through podcasts and the likes, Babbage did indeed not really see the implications of what he'd built, but Ada Lovelace did absolutely see the far reaching implications of general-purpose computing, and she'd spend a lot of time speculating about it. Unfortunately this thing sorta got stuck in development hell. Could you imagine where we'd be if they'd got this thing built?
I have an old mechanical calculator which could add and subtract by cranking a handle, and then it could also multiply and divide by doing certain operations (made by a company named Walther, which also produces weapons). So the mechanical side of things is partially there. OK, it could only store 2 numbers, one of which was overwritten by the result, but it is already a marvel of mechanical engineering that nobody will be able to understand nowadays (unfortunately, it doesn't work right any more). Based on this type of rotating barrel mechanism, it would indeed be possible to build such an analytical engine, but of course it would need much more than that, like the mechanics for storing numbers in the "permanent" store and so on.
Done. I saw the hardware at the Computer History Museum in Mt. View, California, about two years ago.. It was on display for a year prior to being shipped to another museum.
That is called the difference engine. While they were both designed by Babbage, the difference engine is much smaller than the analytical engine, and had a specified purpose; namely to tabulate polynomial functions. I'm not sure if it is Turing compliant (it most likely is not), but it certainly isn't a general purpose computer like the analytical engine.
It is a right step in the direction of eventually building the full analytical engine, though. Maybe someday we'll get to see the thing in operation, and perhaps even run some of Lovelace's programs on it.
Difference Engine No. 2. I've seen its twin at the Science Museum in London.
I watch this video every once in awhile. Professor Brailsford is always a great person to listen to. I find myself wishing for a deeper look at the instruction set the Machine would have used. I keep coming back to the idea that the instruction cards could be bigger, contain more instructions. There is talk of the logic the machine could process, but not the instructions. I'd love to see a video about the minimum instruction set needed, and also how it could be extended to add more speed and capability.
I'd like to see an instruction on the Program cards that asked the machine to advance to the next card, for an additional instruction set. The idea that an operator would swap out the stream of cards for a subroutine for Square Root or some such seems inefficient. Perhaps the Machine could have had a dedicated sub-program built in, and called when needed. That and a few other things that I'm pretty sure Babbage would have realized as improvements.
Anyways, thanks for all the videos, always.
Today, we have MEMS - microelectromechanical systems - using the same technology like in the production of semiconductors, tiny mechanical components - wheel, cogs, rackets, etc. can be built to very small sizes and integrated with electronic components; in theory this technology could be adapted to make only microscopic mechanical parts so Babbage's computer could be built but it would be hundreds or thousands of times smaller...
I really love this professor!
beatngu he's fantastic. I enjoy every video by him
I feel like this thing would break down pretty regularly with all of the mechanical parts.
This is probably THE most steampunk thing possible. A fully theoretically realised, Turing-complete, steam-powered computer that's the size of a locomotive
I mean, Sweet Jeebus I hope this thing is built one day. This would be amazing to see in operation
Father of Computer for a reason..........
A mechanical processor , he's the Ultimate Hipster
His machines were Turing complete before it was cool.
A mechanical processor, from the mechanical professor \o/
I seem to remember my early adventures into computing in the early 80's was via a Teletype connection to an ICL mainframe. Once you had fanished your session it told you how much "Mill Time" your work had used and your storage use was in Buckets :-)
Charles Babbage was the computer Genius of his time, if in any doubt Google "The Babbage Engine" . It truly is an astonishing feat of human invention.
I know this machine was never completed, but I wonder how much would it cost to build a mechanical computing machine like this today with modern methods?
$238,721
As a computophile, I read that Mary, the Bitish poet Lord Byron's daughter, was the first computer operator who said, "To operate, I went into the room. Such a huge machine it's!"
One can only imagine how Babbage would react knowing his dream would be vindicated and analytical engines would spread across the world
And imagine telling him he could get thousands of these on a thin wafer of silicone......
If you want to better understand this machine, I recommend you learn about the Differential Engine first. This engine is the same basic idea, but where the memory and the processing are separate components rather than shared.
Sidney Pardua has also uploaded good videos in her RUclips channel telling details of both machines.
That's insane. I would've loved to see the inside in action.
This is so cool and steampunky, It really reminds me of Wolfenstein (the new order) with those big steampunk machines.
The book that basically started the whole steampunk thing is set in a Victorian England where Babbage's designs were made. I think it was called The Difference Engine.
Well it is a steam-driven computer.
Bit-Bit-Jump manages to be Turing complete without a conditional jump(if statement) I think this demonstrates that as an alternative to conditionals, being self modifying is sufficient for Turing Completeness.
Ada Lovelace is a queen
It sounds like 4:27 the "store" is a faster form of memory, we computer engineers called a "register" 4:40 the "Mill" sounds like what we call the ALU or arithmetic logic unit. 5:04 looms sound like the hard drive or tape storage equivalent of the modern computer.
Its time for more computerphile! Yay!
I think "mill" was used to draw an analogy to flour or sawmills: a place where numbers are processed like wheat or wood.
Do you think pointless genius are a new thing (string theorists comes to mind) well, it isn't.
Babbage's inability to finish one thing, and lovelace writing code for a machine never built instead of goddamn building one.
Babbage, because inventing is useless if no one builds it, lovelace, because no point making the best ammunition possible for a gun that will never be made, and both because science and invention is fun!
lloydgush They mightn't have actually been built but later real implementations certainly drew on the theory that this and other such designs examined... ☺️
It's not pointless. He needed a lot of money to build it. And the best way to get investment was (and still is) to show that you already have detailed schematics and something to run on it. However, he was never able to persuade investors that it would actually work which, given its cost which the investors were unable to even estimate at that time (because they were not sure if it was possible to build with nexessary tolerances), only a small part of the mill was built in 1910, fourty years after Babbage's death.
To be fair, if this was his first such machine, I think finding investors would have been easy. It's just the problem that he had the reputation of not being able to deliver, as his Difference Engine was never finished. That, and he was just bad with money in general, which meant the investors he did find didn't actually matter that much.
9:58 Need to store some numbers. Babbage's Analytical machine go BRRRRRRR!!!!
I just adore the whole concept. But i wounder, could one argue that the "store" was more like cache, not ram? The Rom would be the jacquard cards. Ram would have been a printable version of the same cards? A HDD would probably have been more of a physical storage of some sort.
Yes, but the design incorporated a printer, so an automatic card punch would have been easy. I think it was planned, actually.
Amazing, and very well presented, thank you.
Cats on computers is a timeless phenomenon
Likely one of the greatest engineers that ever lived.
Thanks to the Professor for being the only person I've ever heard pronounce the name "Lovelace" correctly. However sad it may seem, it is supposed to sound like "Loveless".
Send a Fifth-Stage Navigator to demand details of the Emperor. 1:16 Navigator: “We have just folded space.”
Has anyone ever made a complete 3D / CAD-Based model of the analytical engine? Maybe a fly-through so you can see some of the different parts? I've founds lots of videos exposing small parts of it but that shows an artist's interpretation of the full machine. Let me know if you know of one. Thanks!
I was thinking I can just about keep up with what's going on with the machine, but after halfway it was just impossible. Reading the cards mechanically and storing their information? Transferring the information to the back of the machine? Systems to add things together with a preset thingy that separates numbers from adding. It's like recently I've thought I almost start to understand what a very basic computer does on low level, but when you present this engine, it's all gone.
Someone should start a kickstarter for this!
There's actually a group that's trying to build it. They're still in the process of trying to figure out Babbage's plans so they can build a virtual model for testing.
Before I clicked to watch I thought "Will they actually pay respect to Ada Lovelace?"
Respect was paid in the first 10 seconds, well done guys! :)
Kind regards,
Meta Custom Computers
No wonder Babbage never built it. I knew it was complex, and huge, but I honestly had no idea.
What I want to know is if you put into the machine wrong figures, will the right answers come out?
Can anyone in the comments help? For some reason, when sped up, sections of youtube videos that have little to no audio, eg the part where computerphile is being typed, create MASSIVE lag. It only happens on my chrome book. PLEASE HELP
New Mark 3 Analytical Engine with internal combustion drive. Get yours for only 999,999,999.95. Write now they're going fast!
How do you do a non-destructive copy of one register to another? Racks and pinions I get, but wouldn't you need a way to sense when two 50 digit registers hold the same number? If you just decrement one register while incrementing the other until the first hits zero, you've destroyed the original.
Even in the 1830s, kitties like to walk across computers.
Excellent topic guys.
?how many registers had the Mill-CPU
?were the registers named
?were there specialized registers...
?length of registers (in decimal-state-units... of course)
;-)
Its really bizarre to me this vid was released in 2016... And only has like 100,000 views... I think people fail to appreciate that the analytical engine was turing complete. It may have been decimal/mechanical but Babbage was so far ahead of his time on this one its not even funny. This is the grandaddy of computers wheres all the so called "computerphiles"?
Imagine miniature steampunk computers everywhere
Really impressive and complicated to understand how this would work. Now, off subject, doesn't it look like a Guild Navigator tank?
Fasinating Video !! :D
How fascinating. A steampunk engine might break the computer. I don't know, steam seems very crude and uncontrollable. I'd love to see even a small version of of this running.
My Intel chip? No, professor Brailsford, I'm strictly AMD when it comes to CPUs...
So you are the one causing the forest fires?
Speedyjens No, because in my gaming rig I have a K10 chip, not Bulldozer. I only have Bulldozer in my HTPC, and that's a 25W Sempron...
> I'm strictly AMD when it comes to mills
FTFY
it gets hotter cause it has better performance/ higher clock speed
***** That's not actually true. Increased clock speeds don't necessarily increase heat output. Higher operating voltages, however, do, and to stably use higher clock speeds, higher voltages are generally necessary as well. The manufacturing process also makes a big difference, with smaller transistors generally requiring less voltage to perform equally. The reason modern AMD CPUs output more heat than their Intel counterparts is that AMD's offerings are based on larger manufacturing processes, commonly 32nm compared to 14nm.
Was building the machine using my minds eye, especially the data transfer part
Can you make a VR version that would work? Cheers, G - London
Love this guy!
If this thing was built and improved upon over time then most likely computers would be far more advanced now.
But I suppose you can thank WWII for the advancement of that, unfortunate as that sounds.
I don't think so. Remember this was pre-Turing era. At that time it was not known that every solvable problem has algorithmic solution - aka. they didn't know computers can in principle actually compute anything. Also, modern computers are so powerful because they take advantage of technology that did not exist back then (specifically electronics). In terms of hardware computers would be at about the same level as today.
honestly i doubt it, we are past the rapid developments and are hitting limits with normal processors.
The limits are actually software, you can keep just adding more and more cores to the cpu. Electrical limits are also a reality, at least for the common person, but businesses/governments could mostly get around that. If binary computing at it's very nature limits programming possibilities, as opposed to say quantum computing, that I do not know.
@@donaldjuan1729 Only some workloads can be mass parallelized like that. Many others aren't affected at all by slapping on another CPU.
I wonder machine computation being available 100 years before it actually became available would have altered the course of science and technology such that we would be much more technology advanced in the present. In principle, atom bombs and nuclear reactors could have been designed with even the tiny computing power made available by the analytical engine.
but can it run crysis?
Reaaaaaaally slowly
I expected this comment.
"It is only a question of cards and time"
of course not, crisis runs on binary. also i don't know if it can manage function
It's Turing Complete, so yes
is a machine so incredible!!!
This is amazing!
DdI you just analyse the analytical engine?
Building it would be the ultimate steampunk project :-)
"In typical victorian style, it's not called a databus."
Sydney Pasha's graphic novel "The Thrilling Adventures of Babbage and Lovelace" is hilarious.
What purpose would this thing have served if it was built? I really can't think of any useful applications.
mandaloin printing logarithmic tables was the original intention behind babadge works I think.
mandaloin it's a computer. it would do computer things.
PleaseDontWatchThese I use a computer only for things that can be displayed visually like games and videos. Or research, which isn't possible without the internet, or word processing, which would be pointless on that machine. Everything I can think of would be impossible or impractical on a machine like that.
Just think of the machine like a really big scientific calculator without graphing, such as the ti84 that is very popular in high schools.
It's primary use would be making computations that are guaranteed to be correct.
+ Ybalrid
Actually, nautical tables for use by ships at sea. That was how he pitched his grant proposal to the Crown. There were for such tables in those days, as mentioned in this video, tables of corrections issued; then tables of corrections to the corrections; then tables of corrections to the corrections to the corrections!
Babbage's vision was to automate the whole thing, so that the resulting printed tables would be error-free, from the outset!
And yes, pure mathematical tables - logs, trig functions, etc. - would also be well-suited work for this contraption.
WOW.. that would be roughly 20% bigger than a standard 20ft container :)
did it have RGB lights?
Do an episode on the ABC computer please!
My thoughts exactly.
Can it run crysis
Of course. With a kilometer-long stack of punch cards, a ten kilometer long store, and one frame per month with a resolution of 10x10 punch card holes. And a couple of tonnes of fuel per month for the steam engine.
I want to see a marbled computer. :D Rube Goldberg memory! XD
Neat! Think if Babbage had FLODAC tech, patent 3190554, the pipe organ folks could have built his engines?
really great
8080 views, fun coincedence