I'm a (retired) analogue IC designer and have many happy (?) memories studying other companies IC's using optical microscopes in our engineering laboratory, to reverse engineering them. It was a good way to learn tricks of the trade. I hope some engineers have studied some of my creations! We also used manual micro-manipulators to electrically probe connections when studying a powered up die. We sometimes used a laser to cut tracks in the lab. We also occasionally used an ion-beam milling machine to slice into the depth of an IC so we could modify the silicon experimentally. Oh what fun we had!
Ion beam milling machine what a unique bit of gear! Something like that would have cost millions. What companies would even manufacturer a machine like that
I forget the name of the company we used. It's not a big machine. Same principle as an electron beam microscope I think, although the polarity is obviously reversed! A tall evacuated chamber along which one accelerates charged particles. I think you have a very hot "anode" source creating a plasma (if that is the right terminology) so ions can be sucked away by the electric field. magnetic fields deflect the beam. The ions are much heavier than electrons and abrade the surface they hit... If I recall correctly, there was even technology to deposit material at a lower energy so that one could add a conducting track, very slowly......
@@fakename3344 I think it was companay in Cambridge or maybe even a university facility, I live in the UK... If you do a google for "focussed ion beam" milling etc, you should find some resources, certainly in the USA and in Europe, probably Japan and China too..
I've always been impressed with those who could read disassembled code. but at minute 11:43 , Ken is disassembling SILICON and THEN disassembling the code that is in that ROM in that silicon! Oh as if that is not enough, he also wrote a simulator of the calculator he reverse engineered BY LOOKING AT THE SILICON!? WOW...
Actually there is a method on how to perform that... Decap, hotspot, curve tracing etc. its just depending on the ic function/s thats varying and talent is required.
I studied automation engineering and almost everything here was a part of the curriculum. I wish it was taught like this! Had so much difficulties in understanding the implementation of gates on chip that I couldn’t visualize it so had to learn them and I hated ‘just learning’ part. There’s no fun without seeing it practically. Great video
Good talk but what are they doing with the camera taking shots from the back of the room and switching to the talker when he is showing something on the screen. Whoever filmed and edited this, watch some defcon talks.
Early Mostek calculator devices (1970s) had fake contacts that would mess with reverse engineering and swiping the circuit. The fake contacts were not able to be detected under optical microscopy unless you were good at noting the direction of focus
crazy how much stuff is on these tiny pieces, and its getting even crazier by the day. I can't even imagine how much engineering hours and knowlege went into any of these things.
I mean, in a way, they're essentially the same thing. With Factorio, you're taking resources and converting them into other resources. With processors you're taking inputs and converting them into outputs.
If you have an archive of the raw footage from all of the cameras, it would greatly helpful to have the wide angle shot showing you pointing is really helpful. However, the slides are better resolution, but where he is pointing is lost. The lecture is very interesting, but when you don't see where he is pointing makes much more difficult to follow Thanks.
Yeah you can't win with the slides... laser pointer OR better quality image but you can't have both.... ideally we need speakers to use some kind of "virtual laser pointer" on the presentation laptop.
@@edgeeffect You can have both, it's called picture in picture. You can show slide, presenter and picture of where he's pointing at the same time. Then the virtual experience would in most cases actually be better than the real one. "Virtual laser pointer" as in mouse pointer? Yeah, would be nice if someone invented that.
When I was 11 or 12 years old I did a science fair thing where I got the chips out and let people look at them with a microscope. The way I came up with for getting the silicon out was simply by heating the package until it crumbled.
How about an old SGI Tezro? SGI doesn't exist anymore, and to our knowledge these machines aren't repairable using standard services. Reengineering the ICs may be the only way to service them, and emulation requires the same thing.
Currently in school for comp E and we're slowly going over the topics presented, so this talk is really fascinating (to see this stuff in practice!) Does anybody know how large were the teams that were designing these chips?
well, I started programming in the late 60s, but I can still understand like maybe up to 30 percent of this. I checked with a couple other people, sure enough, they couldn't understand a single word, so I'm happy enough.
Anybody here still remember the early days of 1980s when the Soviets stole the blueprint for Zilog Z8000 microprocessor and tried to reverse engineer it.
Am I the only one who saw "Instruction ROM" in the calculator chip and immediately thought, I'd love to experiment with that and see what changing values in it would do?
His good . the true beauty of transistors is the differance in opening, and closing . Some are Voltage controlled, some are Ampere controlled, some are Normally Open NO, others are NC, Normally Closed , before you apply voltage to them . NPN & PNP junctions .
Excellent amount of information. Respect. Just the Saliva sound was a bit bothersome. Sounded a bit kinky while under the impression of knowledge. You know.
These chips are not modern, but in fact very old. This would be nearly impossible to do with modern chips, as one of the commenters above explained very well
Silicon in an elevated enthalpy reaction(add heat) undergoes an elevation of the 3p orbitol, allowing for an enthalpy controlled reduction/oxidation reaction of the atom. So start there. Then make transistors...
I have a David and Mann x,y table from a microchip factory. Very accurate. It had a 110 volt motor on the x axis. So I wonder if it was for cutting the wafers up.
you can now inkjet print transistors. 1500nm is easily possible, with some tweaks to the printer's hardware you can reach 800-500nm. If you want to spend on a super high grade research printer... 160 nm is about the limit you can do . But you can do a one-off chip design on a piece of glass, so that's pretty impressive. not to mention makes custom chips very very cheap
I wish this were more accessible. Imagine if someone invented a machine that could scan a chip in real-time and get an oscilloscope-like trace at every point, and it was accessible enough that you don't need to be a big company or government to afford one. That would be the Holy Grail of hacking tools, and the end of manufacturers being able to hide secrets in products they sell without the owner being able to see them. But IDK how realistic that would be any time soon. :p
What about the GBA? It uses an ARM/Thumb core. Though obviously, yes, it contains a Z80 as well, for obvious reasons. (in fact, the Mega Drive/Genesis contains a z80 for basically the same reason; Technically it's the Audio co-processor, but I'm sure using it for that was an afterthought; The only reason they chose a z80 for that task is that they wanted it to play Master System games and thus needed one anyway. Might as well put it to work doing something else while it's there.) Still... Z80... Ehh. I'll stick with my 65xx family chips thanks. XD
I love the vintage processor chips, especially Z80, 8085, 805x series. Yeah I know you won't build a Pi clone with one of those but they are fun to work with!
I've found that very flat ceramic chips (which can't be chiseled) can be heated and will crumble, though occasionally with case residue on the die. The fumes are not good for breathing.
I think that we soon will have access to high power laser beams that can chisel away AND scan a chip in 3D and remodel them, as a by-product of the nascent 3D-printing technology.
I think the hardest if not impossible part of doing that is more complicated/modern chips easily contain many vertical levels like shelves. This way of reverse engineering integrated circuits only works with the surface layer.
Well theoretically its possible. But you should know how to completely etch given layer not touching the level below. And to know that you should sign NDA in Intel or AMD and work there for a while... SUCCS
There is a reason he works with old chips. They are much bigger and you can make out what's going on. Modern dies have all features really crammed together that you just can't tell wtf is going on. Way, way worse than these, if you can believe that.
I've been wondering how the control logic schematic looks like and how it interacts with the interrupts in the instruction rom( schematic) at transistor leve of the z80 and it's burning my neural circuit. Someone plz save me!
9:02 just for efficiency, would it not be better to use just 2 transistors to solve the NOR circuit the 3 transistor seems of no use theoretical it might be part of the drawment for the negation but why would you do that then
I always wondered if a working chip with a metal lid will continue to work with the lid popped off and exposed to ambient air. The lid would have to be removed carefully so as not to damage the bonding wires or the chip. Would light have any effect? How about a low power laser such as from a pointer? EPROMs continue to work even though the window is exposed (although erasure might occur after some time), so I guess yes?
it has to be exposed to UV light a strong UV light. Which is why some of these EEPROMs that are being pulled out of vintage computers from the 80's still have all of their EEPROM data intact.
I'm a (retired) analogue IC designer and have many happy (?) memories studying other companies IC's using optical microscopes in our engineering laboratory, to reverse engineering them. It was a good way to learn tricks of the trade. I hope some engineers have studied some of my creations! We also used manual micro-manipulators to electrically probe connections when studying a powered up die. We sometimes used a laser to cut tracks in the lab. We also occasionally used an ion-beam milling machine to slice into the depth of an IC so we could modify the silicon experimentally. Oh what fun we had!
Ion beam milling machine what a unique bit of gear! Something like that would have cost millions. What companies would even manufacturer a machine like that
I forget the name of the company we used. It's not a big machine. Same principle as an electron beam microscope I think, although the polarity is obviously reversed! A tall evacuated chamber along which one accelerates charged particles. I think you have a very hot "anode" source creating a plasma (if that is the right terminology) so ions can be sucked away by the electric field. magnetic fields deflect the beam. The ions are much heavier than electrons and abrade the surface they hit... If I recall correctly, there was even technology to deposit material at a lower energy so that one could add a conducting track, very slowly......
@@franklydude Would you happen to remember what it was? I'm really curious as to where I'd get the equipment to do this.
@@fakename3344 I think it was companay in Cambridge or maybe even a university facility, I live in the UK... If you do a google for "focussed ion beam" milling etc, you should find some resources, certainly in the USA and in Europe, probably Japan and China too..
@@franklydude
Yeah, they also have ion implantation (basically ion doping) machines.
I've always been impressed with those who could read disassembled code. but at minute 11:43 , Ken is disassembling SILICON and THEN disassembling the code that is in that ROM in that silicon! Oh as if that is not enough, he also wrote a simulator of the calculator he reverse engineered BY LOOKING AT THE SILICON!? WOW...
Anyone that had being paying attention to EE classes (VLSI, Computer Archtecture, Microprocessors etc.) can do that. It is a lot of work though.
Actually there is a method on how to perform that... Decap, hotspot, curve tracing etc. its just depending on the ic function/s thats varying and talent is required.
I’m sayin right... what a bad ass lol
what a royal waste of time though ...
You are watching videos, what a waste of time!
Jokes aside, any hobby may seem like a waste of time to others.
I studied automation engineering and almost everything here was a part of the curriculum. I wish it was taught like this!
Had so much difficulties in understanding the implementation of gates on chip that I couldn’t visualize it so had to learn them and I hated ‘just learning’ part. There’s no fun without seeing it practically. Great video
FFS, somebody give this guy a glass of water!
Banana noises.
so true lmao. Michael you hit right on the head
hahah i thought i was the only one who noticed that :D :D
it's like he's chewing on the very circuitry from his lecture.
actually those dry-mouth sounds are so pleasant for me... weird...
Laser pointer vs mouse pointer :-(
"this here..."
"these ..."
WHERE?!
Good talk but what are they doing with the camera taking shots from the back of the room and switching to the talker when he is showing something on the screen. Whoever filmed and edited this, watch some defcon talks.
Yeah the filming is abysmal. 95% of the time he's showing something with a laser pointer, you can't see what it is.
I really wanted to see what he meant by the "oh that's bad..." slide.
Couldn't watch...camera man destroyed all follow along to the educational value seeking to be given.
Hand that man a broom.
Ken is awesome
u too
My hero ❤️
Reverse engineering, the ultimate hacker skillz
If this the guy from the curious Marc videos?
I chuckled when I saw the instruction "waitno"
Early Mostek calculator devices (1970s) had fake contacts that would mess with reverse engineering and swiping the circuit. The fake contacts were not able to be detected under optical microscopy unless you were good at noting the direction of focus
*_THAT_* was freaking cool!!!
_So much better_ than textbook cartoons!
Simply put: Great overview of silicon features and excellent approach on chip decapping.
I need the laser pointer when you zoom in please! Because we have no idea where he is pointing to.
i actually learnt quite a lot
Phenomenal presentation! I want to thank that guy (and get him a bottle of water!)
It could've been even longer presentation and I still would watch it.
man you have to give the name of the music played at the end... so good
Very nice and interesting to see how those basic components and circuits are made. Great work!
crazy how much stuff is on these tiny pieces, and its getting even crazier by the day. I can't even imagine how much engineering hours and knowlege went into any of these things.
Absolutely brilliant. Thank you very much.
Thanks, had fun getting a better look into the transistor designs that actually make their way into an ic.
28:30 Looks more like a factorio base than an IC XD
this made me chuckle
I mean, in a way, they're essentially the same thing. With Factorio, you're taking resources and converting them into other resources. With processors you're taking inputs and converting them into outputs.
My digital logic professor brought me here. THANKS J DAWG
after a decades of innovation we still haven't figured out how to get a laser pointer on youtube
This guy should definitely write a book about this topic. I'd love to learn more about it.
2nd time i watch this, still wrecking my brain about that ALU trick.
If you have an archive of the raw footage from all of the cameras, it would greatly helpful to have the wide angle shot showing you pointing is really helpful. However, the slides are better resolution, but where he is pointing is lost. The lecture is very interesting, but when you don't see where he is pointing makes much more difficult to follow Thanks.
Yeah you can't win with the slides... laser pointer OR better quality image but you can't have both.... ideally we need speakers to use some kind of "virtual laser pointer" on the presentation laptop.
@@edgeeffect You can have both, it's called picture in picture. You can show slide, presenter and picture of where he's pointing at the same time. Then the virtual experience would in most cases actually be better than the real one.
"Virtual laser pointer" as in mouse pointer? Yeah, would be nice if someone invented that.
Guys like him and the designers are out of this world, seriously extra terrestrial ufos.....wow.... way over my head..........
shame I cant see the lazer pointer
When I was 11 or 12 years old I did a science fair thing where I got the chips out and let people look at them with a microscope. The way I came up with for getting the silicon out was simply by heating the package until it crumbled.
metallitech hey that's what i do!
I don't think I'd wanna buy a POSFET OR a POS transistor. this is a very interesting video.
How about an old SGI Tezro? SGI doesn't exist anymore, and to our knowledge these machines aren't repairable using standard services. Reengineering the ICs may be the only way to service them, and emulation requires the same thing.
Currently in school for comp E and we're slowly going over the topics presented, so this talk is really fascinating (to see this stuff in practice!)
Does anybody know how large were the teams that were designing these chips?
Incredible only in America such geniuses self develop
well, I started programming in the late 60s, but I can still understand like maybe up to 30 percent of this. I checked with a couple other people, sure enough, they couldn't understand a single word, so I'm happy enough.
:)
Anybody here still remember the early days of 1980s when the Soviets stole the blueprint for Zilog Z8000 microprocessor and tried to reverse engineer it.
I dont. Got any info? Im interested
Thank you! This is the kind of information I always tryed to find. Great links!
Am I the only one who saw "Instruction ROM" in the calculator chip and immediately thought, I'd love to experiment with that and see what changing values in it would do?
Beautiful work.
I watch these videos with no background in CS, i'm just like wtf how is this even possible?!?
No, it is possible. If you study semiconductor related knowledge, it will be piece of cake.
His good . the true beauty of transistors is the differance in opening, and closing . Some are Voltage controlled, some are Ampere controlled, some are Normally Open NO, others are NC, Normally Closed , before you apply voltage to them . NPN & PNP junctions .
Excellent amount of information. Respect. Just the Saliva sound was a bit bothersome. Sounded a bit kinky while under the impression of knowledge. You know.
You sounded really nervous. Relax, it was a fascinating talk and your presentation was excellent.
Very interesting what you do Ken. Thank You.
Thank for sharing this! Really appreciate it!
This Guy is a master of the modern reverse Engineering, a God...
These chips are not modern, but in fact very old. This would be nearly impossible to do with modern chips, as one of the commenters above explained very well
Nice outro music. Sounds very 303-ish :-)
Can you next reverse engineer Apple M1, bit by bit? ;)
It'd be nice if we could actually see what he's pointing at when he keeps saying which parts "run here"
Merci pour la vidéo j'aime beaucoup les recherches
كتبغيو البحث ولا كيعجبكم غير تشوفو الأخرين كيقومو بأمور خارقة بحال هكا,,العربان أش عندكم ماديرو
Silicon in an elevated enthalpy reaction(add heat) undergoes an elevation of the 3p orbitol, allowing for an enthalpy controlled reduction/oxidation reaction of the atom. So start there. Then make transistors...
where did the cool opening and closing music come from? anyone know who made it?
bro, u r amazing, that is soo cool
,nd the earth is burning,
MikesElectricStuff does videos deencapsulating chips which is cool, but havent seen or forgot ever seeing anyone show a transistor-level look! Cool
I have a David and Mann x,y table from a microchip factory. Very accurate. It had a 110 volt motor on the x axis. So I wonder if it was for cutting the wafers up.
very nice inside of very popular chips. I always wonder how do they work still so tinny.
Oh hello Ken! Seen you before hanging around with Curious Marc.
It was a pleasant surprise to recognise someone. He definitely is a reverse engineering king.
Fascinating talk, learned a few things new things it. Thank you so much.
Absolutely fascinating! Thanks a lot!
Hmm, chip manufacturers are playing factorio, but trying to make the base as small as possible. That's a tough one!
We need people like that to rebuild civilization after collapse.
you can now inkjet print transistors. 1500nm is easily possible, with some tweaks to the printer's hardware you can reach 800-500nm. If you want to spend on a super high grade research printer... 160 nm is about the limit you can do . But you can do a one-off chip design on a piece of glass, so that's pretty impressive. not to mention makes custom chips very very cheap
This blew my mind
Rawr, Love the chip die art!
this is crazy, awesome, breathtaking
I wish this were more accessible. Imagine if someone invented a machine that could scan a chip in real-time and get an oscilloscope-like trace at every point, and it was accessible enough that you don't need to be a big company or government to afford one. That would be the Holy Grail of hacking tools, and the end of manufacturers being able to hide secrets in products they sell without the owner being able to see them. But IDK how realistic that would be any time soon. :p
This looks like it takes a lot of patience. But that's okay, well worth it in the end.
we should b extremely thankful 4da existence of XEROX ALTO cuz it was a precursor 2all modern Pc's.
Thank you. I wanted to ask about the power of the microscope used
Great job! Thank you for sharing :D
This Guy is a Beast with Silicon! Wow!
Beto O'Rourke?
Brilliant :)
My sentiments exactly.
Just realized he wrote the IR library for my garage stereo
Great lecture
if this guy can reverse engineer code from silicon surely he can invent a microphone that removes the mushy banana sound from his mouth.
im a big fan of yours since curious marc channel !!!!
this channel is awesome
Is chip reversing used in software reverse engineering?
It's Master Ken!!!
Nice explaination
1:49 Apparently it's not visible in the video.
“Any Z80 fans out there?”
Yes, me! Me! Me!
Right...the game boy advance..tho..
What about the GBA? It uses an ARM/Thumb core. Though obviously, yes, it contains a Z80 as well, for obvious reasons.
(in fact, the Mega Drive/Genesis contains a z80 for basically the same reason; Technically it's the Audio co-processor, but I'm sure using it for that was an afterthought; The only reason they chose a z80 for that task is that they wanted it to play Master System games and thus needed one anyway. Might as well put it to work doing something else while it's there.)
Still... Z80... Ehh.
I'll stick with my 65xx family chips thanks. XD
Ditto!
I had to pet the Z80 I have in front of me when he said that.
I love the vintage processor chips, especially Z80, 8085, 805x series. Yeah I know you won't build a Pi clone with one of those but they are fun to work with!
It would be so nice to "test" Sat TV again :) and reunite guys like Raton, Penga, Silverman etc,etc and learn from them again. I'm to old to start lol
this sounds interesting, can you tell more about it?
Not unusual cur mirror it saves allot of space. Beautiful design
I've found that very flat ceramic chips (which can't be chiseled) can be heated and will crumble, though occasionally with case residue on the die. The fumes are not good for breathing.
I think that we soon will have access to high power laser beams that can chisel away AND scan a chip in 3D and remodel them, as a by-product of the nascent 3D-printing technology.
I like how just one guy laughed at the POS joke. Probably an engineer.
Probably a salesperson.
Fantastic - very informative.
For YT videos, I suggest you show the diagram next to the actual screen. I can't see that laser dot on my screen to sync with his explanation lol.
master Ken
absolutely brilliant
So how much work would it be to reverse engineer an i5 and it's microcode to see if the NSA realy does have a backdoor at that level?
Hundreds of thousands of man hours.
AMD too now
I think the hardest if not impossible part of doing that is more complicated/modern chips easily contain many vertical levels like shelves. This way of reverse engineering integrated circuits only works with the surface layer.
Well theoretically its possible. But you should know how to completely etch given layer not touching the level below. And to know that you should sign NDA in Intel or AMD and work there for a while... SUCCS
I bet it won't take long before someone trains an AI to solve those kind of problems.
Who's the end theme song from? I need that track badly..
thank you!
Fantastic stuff. Is there anything like a program that can be fed a die photo and more or less make sense of it?
that's how you create Skynet
There is a reason he works with old chips. They are much bigger and you can make out what's going on. Modern dies have all features really crammed together that you just can't tell wtf is going on. Way, way worse than these, if you can believe that.
what's more he doesn't even want to deal with acid, kinda limits the options
In the past 40 years, dropping acid has been for hippies and dope-heads. In the future it will be an activity for hackers! :P
Just so happens I have 2 Z80s right in front of me. Just loose ones. Not in any circuit. I had to pet one when he mentioned it.
I've been wondering how the control logic schematic looks like and how it interacts with the interrupts in the instruction rom( schematic) at transistor leve of the z80 and it's burning my neural circuit. Someone plz save me!
Excellent!
9:02 just for efficiency, would it not be better to use just 2 transistors to solve the NOR circuit
the 3 transistor seems of no use theoretical it might be part of the drawment for the negation but why would you do that then
I always wondered if a working chip with a metal lid will continue to work with the lid popped off and exposed to ambient air. The lid would have to be removed carefully so as not to damage the bonding wires or the chip. Would light have any effect? How about a low power laser such as from a pointer? EPROMs continue to work even though the window is exposed (although erasure might occur after some time), so I guess yes?
it has to be exposed to UV light a strong UV light. Which is why some of these EEPROMs that are being pulled out of vintage computers from the 80's still have all of their EEPROM data intact.
thanks
Cool video, thanks :)