27 ms·

Wow! Is the trick that we now have powerful microcomputers small enough to fit into a USB plug? That's pretty incredible technology. How many years ago did this

Wow! Is the trick that we now have powerful microcomputers small enough to fit into a USB plug? That's pretty incredible technology. How many years ago did this become possible? My IT security training is dated, I am aware of the risks of plugging in a random USB key, but just a cable from a helpful "coworker"? Yikes.

fragmede 5y ago
That level of miniaturization is far older than Apple's removal of the iPhone headphone jack in 2016, but the related lightning-to-audio jack dongle had a microcontroller with a DAC inside that you'd never think existed due to the form factor.
- varikin 5y ago
  Just for reference, here are a couple breakdowns of how impressive the iPhone headphone jack is. It is impressive how good it is and how tiny it is. https://www.ifixit.com/News/8448/apple-audio-adapter-teardown https://www.ifixit.com/News/8448/apple-audio-adapter-teardow... https://www.kenrockwell.com/apple/lightning-adapter-audio-quality.htm https://www.kenrockwell.com/apple/lightning-adapter-audio-qu...
  - duped 5y ago
    It would be more impressive if they could fit it inside the phone.
    - 1-more 5y ago
      remember project aria? Apple actually did that and everyone got mad!
    - ChuckNorris89 5y ago
      Funny enough, that shouldn't be a challenge for Apple since the world's slimmest smartphone ever made actually has a headphone jack[1], so Apple excuse of not enough space inside the phone is horse shit to force everyone into the Airpods ecosystem. [1]https://youtu.be/Ijssm6Qv5nE https://youtu.be/Ijssm6Qv5nE
      - djrogers 5y ago
        It was never thickness, it’s was depth.
        ChuckNorris89 5y ago
        What's the difference in this case? Volume is a multiplication of dimensions on all axis. If they didn't compromise on thickness that means they have more international volume available to reduce the other dimensions which they didn't as modern iPhones don't lack depth either, barring the mini, they're all quite large on all axis.
- trangus_1985 5y ago
  The Lightning-to-HDMI adapter is also an insane miniaturization. It runs a (very) stripped down version of iOS/darwin (not sure what apple counts it as) that is loaded in about a second when you plug in the phone, establishes a network connection, and streams compressed video frames over the network over USB to the HDMI. That's why when you use the iPhone HDMI adapter, everything looks a little bit compressed. Because it is.
  - gregsadetsky 5y ago
    I didn’t know about this! Thank you. Truly nuts: https://hackaday.com/2019/07/30/apple-lightning-video-adaptors-run-ios-dynamically-loaded/ https://hackaday.com/2019/07/30/apple-lightning-video-adapto...
  - crazygringo 5y ago
    Yup. It's actually an incredibly clever piece of tech that lets the iPad/iPhone get around the fact that lightning doesn't have enough bandwidth to transmit HDMI. For regular home/app views, it does hardware compression of the iPad's screen, outputs that over lightning, then the adapter decompresses it to raw HDMI. While for Netflix/etc. streams, it outputs the stream directly to the adapter to decompress, without quality loss. (And at full size as well, rather than double-letterboxed.) I still haven't figured out the magic of how apps like Netflix are able to do overlays of subtitles on top of the compressed video stream. Best I can tell, there must be a separate API for that, that gets sent in parallel.
    - Dylan16807 5y ago
      As far as working within the constraints of USB 2, it's a solid way to do things. And I'll even leave aside the issue of running a whole OS just to decode video. But this is dealing with 1080p output. That's not very intensive. A lightning port has two high speed data pairs. You don't even need USB 3 speeds to transmit HDMI over one of those data pairs, and then put in like a $2 redriver. If you set up double-sided output I think you could even have a passive lightning to HDMI adapter. But that's a side issue, my main point is the bandwidth available that makes these tradeoffs unnecessary.
NamTaf 5y ago
It’s a copy of the NSA ANT Coppermouth cables. That was part of the Snowden leak, so it’s been possible since at least then (the doc itself is circa 2008) if you have a three-letter name and a national security black budget. https://en.m.wikipedia.org/wiki/NSA_ANT_catalog https://en.m.wikipedia.org/wiki/NSA_ANT_catalog
- dylan604 5y ago
  Was it actually something Snowden released, or part of the second leaker's data that got attributed to Snowden?
- sushisource 5y ago
  Why is the NSA so good at coming up with sweet codenames for things? I swear it's someone's full time job there.
  - elzbardico 5y ago
    Far better names than Microsoft or IBM
    - hbn 5y ago
      Microsoft Integrated Microcomputer Plug-and-Play Cable Office 365 Extended Subscription for Workgroups 2021 Powered by Bing
      - echelon 5y ago
        Google Play Game Libraries for Play (a real product name) I can't find a link referencing it, sadly.
      - nocman 5y ago
        Affectionately known as "MIMPCO 365 ESWPB 2021" Hey! That's my Windows 10 product key! (not really, but felt like it could be).
dylan604 5y ago
In the leaks that happened around the time of Snowden's revelations, there were DVI/monitor cables that had transmitters that would broadcast the decrypted signal so that someone with the receiver nearby could "share" the screen. No software installed for security software to find. Just a cable that could be installed by the cleaning crew after hours.
moonchild 5y ago
See also http://tomu.im/ http://tomu.im/.
5faulker 5y ago
Seems to be very tech-involved though.
lrvick 5y ago
There is already an Arm Cortex M0+ in the end of every USB type C cable for power negotiation. It has a complex codebase and firmware update methods to migrate to new USB specs. Cheap cables don't even support signing so go to town tampering with stock cables if you are so inclined. Also an Attiny85 can fit inside just about anything. My favorite BadUSB hello world is using the Arduino HID library to make a Digispark toggle caps lock randomly with maybe 10 lines of code. Drives people nuts.
- megous 5y ago
  > There is already an Arm Cortex M0+ in the end of every USB type C cable for power negotiation. Only some of the Type-C cables. Normal 3A ones don't have anything like this.
city41 5y ago
On a similar note, as I understand it skimmers placed on ATM machines and the like are now so small they are almost impossible to detect.
kybernetikos 5y ago
> How many years ago did this become possible? I don't know, but I do know that back in 2013 you could get an ARM computer running linux and a webserver with wifi and 16Gb storage in a space the size of an SD card. That is still a bit too big to fit inside a USB plug without being obvious, but not by much. https://hackaday.com/2013/08/12/hacking-transcend-wifi-sd-cards/ https://hackaday.com/2013/08/12/hacking-transcend-wifi-sd-ca... Fitting the electronics inside the usb plug itself has been used for years in slimline usb memory sticks and in tiny readers for micro-sd too. I expect that this has been possible for nearly 10 years, but maybe just not commercially viable for consumers for most of that.
- leeoniya 5y ago
  > Fitting the electronics inside the usb plug itself has been used for years in slimline usb memory sticks and in tiny readers for micro-sd too. those readers usually heat up like crazy for any significant length transfers. good in a pinch, though.
- thrashh 5y ago
  Eye-Fi started as a company making Wi-Fi SD cards in 2005. Chips have been small for decades. What’s changed over the years is that it’s gotten cheaper and cheaper.
  - verytrivial 5y ago
    I'd say the total size of silicon has been small enough for this for at least 20 years. The level of integration between these components, skipping the need for packaging and interconnect is what's allowed these covert devices, probably driven by manufacturing optimization that makes is financially tractable to only make 100 units, so that bit is much cheaper yeah. Things like antennas and other bulky non-silicon passive components are probably the limiting factor now.
baybal2 5y ago
> now have powerful microcomputers small enough to fit into a USB plug Like 20 years ago? 32 bit micros were actually not that far behind low end CPUs until they start to fall off dramatically in the 90, and post-90nm age because perf was good enough. After nineties, CPU improvements were guided as much by software getting worse, slower, and more shoddily written, than genuine need for more raw computing power.
janekm 5y ago
Not just the microcontroller but also the wifi radio and antenna... But a USB-A socket is pretty big anyway. This isn't too different from the tiny wireless mouse dongles that have been around for a few decades.
andai 5y ago
In 2008 we had USB flash drives which extended only two millimeters beyond the laptop when you plugged them in, and Wifi dongles in the same form factor.

fragmede 5y ago

That level of miniaturization is far older than Apple's removal of the iPhone headphone jack in 2016, but the related lightning-to-audio jack dongle had a microcontroller with a DAC inside that you'd never think existed due to the form factor.

varikin 5y ago
Just for reference, here are a couple breakdowns of how impressive the iPhone headphone jack is. It is impressive how good it is and how tiny it is. https://www.ifixit.com/News/8448/apple-audio-adapter-teardown https://www.ifixit.com/News/8448/apple-audio-adapter-teardow... https://www.kenrockwell.com/apple/lightning-adapter-audio-quality.htm https://www.kenrockwell.com/apple/lightning-adapter-audio-qu...
- duped 5y ago
  It would be more impressive if they could fit it inside the phone.
  - 1-more 5y ago
    remember project aria? Apple actually did that and everyone got mad!
  - ChuckNorris89 5y ago
    Funny enough, that shouldn't be a challenge for Apple since the world's slimmest smartphone ever made actually has a headphone jack[1], so Apple excuse of not enough space inside the phone is horse shit to force everyone into the Airpods ecosystem. [1]https://youtu.be/Ijssm6Qv5nE https://youtu.be/Ijssm6Qv5nE
    - djrogers 5y ago
      It was never thickness, it’s was depth.
      - ChuckNorris89 5y ago
        What's the difference in this case? Volume is a multiplication of dimensions on all axis. If they didn't compromise on thickness that means they have more international volume available to reduce the other dimensions which they didn't as modern iPhones don't lack depth either, barring the mini, they're all quite large on all axis.
trangus_1985 5y ago
The Lightning-to-HDMI adapter is also an insane miniaturization. It runs a (very) stripped down version of iOS/darwin (not sure what apple counts it as) that is loaded in about a second when you plug in the phone, establishes a network connection, and streams compressed video frames over the network over USB to the HDMI. That's why when you use the iPhone HDMI adapter, everything looks a little bit compressed. Because it is.
- gregsadetsky 5y ago
  I didn’t know about this! Thank you. Truly nuts: https://hackaday.com/2019/07/30/apple-lightning-video-adaptors-run-ios-dynamically-loaded/ https://hackaday.com/2019/07/30/apple-lightning-video-adapto...
- crazygringo 5y ago
  Yup. It's actually an incredibly clever piece of tech that lets the iPad/iPhone get around the fact that lightning doesn't have enough bandwidth to transmit HDMI. For regular home/app views, it does hardware compression of the iPad's screen, outputs that over lightning, then the adapter decompresses it to raw HDMI. While for Netflix/etc. streams, it outputs the stream directly to the adapter to decompress, without quality loss. (And at full size as well, rather than double-letterboxed.) I still haven't figured out the magic of how apps like Netflix are able to do overlays of subtitles on top of the compressed video stream. Best I can tell, there must be a separate API for that, that gets sent in parallel.
  - Dylan16807 5y ago
    As far as working within the constraints of USB 2, it's a solid way to do things. And I'll even leave aside the issue of running a whole OS just to decode video. But this is dealing with 1080p output. That's not very intensive. A lightning port has two high speed data pairs. You don't even need USB 3 speeds to transmit HDMI over one of those data pairs, and then put in like a $2 redriver. If you set up double-sided output I think you could even have a passive lightning to HDMI adapter. But that's a side issue, my main point is the bandwidth available that makes these tradeoffs unnecessary.

NamTaf 5y ago

It’s a copy of the NSA ANT Coppermouth cables. That was part of the Snowden leak, so it’s been possible since at least then (the doc itself is circa 2008) if you have a three-letter name and a national security black budget. https://en.m.wikipedia.org/wiki/NSA_ANT_catalog https://en.m.wikipedia.org/wiki/NSA_ANT_catalog

dylan604 5y ago
Was it actually something Snowden released, or part of the second leaker's data that got attributed to Snowden?
sushisource 5y ago
Why is the NSA so good at coming up with sweet codenames for things? I swear it's someone's full time job there.
- elzbardico 5y ago
  Far better names than Microsoft or IBM
  - hbn 5y ago
    Microsoft Integrated Microcomputer Plug-and-Play Cable Office 365 Extended Subscription for Workgroups 2021 Powered by Bing
    - echelon 5y ago
      Google Play Game Libraries for Play (a real product name) I can't find a link referencing it, sadly.
    - nocman 5y ago
      Affectionately known as "MIMPCO 365 ESWPB 2021" Hey! That's my Windows 10 product key! (not really, but felt like it could be).

dylan604 5y ago

In the leaks that happened around the time of Snowden's revelations, there were DVI/monitor cables that had transmitters that would broadcast the decrypted signal so that someone with the receiver nearby could "share" the screen. No software installed for security software to find. Just a cable that could be installed by the cleaning crew after hours.

moonchild 5y ago

See also http://tomu.im/ http://tomu.im/.

5faulker 5y ago

Seems to be very tech-involved though.

lrvick 5y ago

There is already an Arm Cortex M0+ in the end of every USB type C cable for power negotiation. It has a complex codebase and firmware update methods to migrate to new USB specs. Cheap cables don't even support signing so go to town tampering with stock cables if you are so inclined. Also an Attiny85 can fit inside just about anything. My favorite BadUSB hello world is using the Arduino HID library to make a Digispark toggle caps lock randomly with maybe 10 lines of code. Drives people nuts.

megous 5y ago
> There is already an Arm Cortex M0+ in the end of every USB type C cable for power negotiation. Only some of the Type-C cables. Normal 3A ones don't have anything like this.

city41 5y ago

On a similar note, as I understand it skimmers placed on ATM machines and the like are now so small they are almost impossible to detect.

kybernetikos 5y ago

> How many years ago did this become possible? I don't know, but I do know that back in 2013 you could get an ARM computer running linux and a webserver with wifi and 16Gb storage in a space the size of an SD card. That is still a bit too big to fit inside a USB plug without being obvious, but not by much. https://hackaday.com/2013/08/12/hacking-transcend-wifi-sd-cards/ https://hackaday.com/2013/08/12/hacking-transcend-wifi-sd-ca... Fitting the electronics inside the usb plug itself has been used for years in slimline usb memory sticks and in tiny readers for micro-sd too. I expect that this has been possible for nearly 10 years, but maybe just not commercially viable for consumers for most of that.

leeoniya 5y ago
> Fitting the electronics inside the usb plug itself has been used for years in slimline usb memory sticks and in tiny readers for micro-sd too. those readers usually heat up like crazy for any significant length transfers. good in a pinch, though.
thrashh 5y ago
Eye-Fi started as a company making Wi-Fi SD cards in 2005. Chips have been small for decades. What’s changed over the years is that it’s gotten cheaper and cheaper.
- verytrivial 5y ago
  I'd say the total size of silicon has been small enough for this for at least 20 years. The level of integration between these components, skipping the need for packaging and interconnect is what's allowed these covert devices, probably driven by manufacturing optimization that makes is financially tractable to only make 100 units, so that bit is much cheaper yeah. Things like antennas and other bulky non-silicon passive components are probably the limiting factor now.

baybal2 5y ago

> now have powerful microcomputers small enough to fit into a USB plug Like 20 years ago? 32 bit micros were actually not that far behind low end CPUs until they start to fall off dramatically in the 90, and post-90nm age because perf was good enough. After nineties, CPU improvements were guided as much by software getting worse, slower, and more shoddily written, than genuine need for more raw computing power.

janekm 5y ago

Not just the microcontroller but also the wifi radio and antenna... But a USB-A socket is pretty big anyway. This isn't too different from the tiny wireless mouse dongles that have been around for a few decades.

andai 5y ago

In 2008 we had USB flash drives which extended only two millimeters beyond the laptop when you plugged them in, and Wifi dongles in the same form factor.