Chipworks Blog
chipworks.com · Company blog · on HN since 2011
Chipworks, a semiconductor analysis firm, posts teardowns and chip-level reverse engineering of devices like iPhones and the Apple Pencil.
What HN saysHN commenters point to it as an example of serious chip reverse engineering and for iPhone teardowns.
34HN people shared it
23linked it in comments
16stories submitted
495story points
7 years agolast shared
2011people per year2026
Most-shared posts
- 1Apple iPhone 7 Teardown4 people · 6 points · first shared 2016 · 1 comments on HN
- 2Homepage3 people · first shared 2013
- 3Inside the iPhone 6 and iPhone 6 Plus2 people · 1 points · first shared 2014 · discussion on HN
- 4Inside the iPhone 5S - Chipworks2 people · 1 points · first shared 2013 · discussion on HN
- 5Hidden warning message found in Samsung’s Galaxy tablet1 people · 161 points · first shared 2011 · 25 comments on HN
- 6Inside the iPhone 6s1 people · 98 points · first shared 2015 · 24 comments on HN
- 7The Apple Pencil; a lot of semiconductor content in something so small1 people · 83 points · first shared 2016 · 47 comments on HN
- 8Apple TV "single core" A5 actually has two cores, one is off1 people · 69 points · first shared 2012 · 58 comments on HN
- 9A9 Is TSMC 16nm FinFET and Samsung Fabbed1 people · 38 points · first shared 2015 · 15 comments on HN
- 10Inside the Apple Watch: Technical Teardown1 people · 32 points · first shared 2015 · 3 comments on HN
- 11Inside the Sony PS41 people · 2 points · first shared 2013 · discussion on HN
- 12A Survey of Recent Image Sensor Pixel Structures1 people · 2 points · first shared 2012 · discussion on HN
- 13The Apple A5X versus the A5 and A4 – Big Is Beautiful1 people · 1 points · first shared 2012 · discussion on HN
- 14WiiU Graphics Processor Polysilicon Die Photo1 people · 1 points · first shared 2013 · discussion on HN
- 15Comparing the invensense and bosch accelerometers found in the iphone 61 people · first shared 2014
What HN people wrote when they linked it
solarexplorer · 2014-06-26
That depends on your definition of "seriously" I guess. But there is no need to stop at the plastic. Have a look what the guys at chipworks are doing: http://www.chipworks.com
Also the Russians are traditionally quite good at reverse engeneering ;-) and they _did_ have their own domestic chips 20 years ago.
BarsMonster · 2013-03-08
Everything is possible.
Cloning whole chip is probably economically not viable, but definitely possible (it's cheaper to design from scratch with modern automatic design tools), but retrieving secret keys is a popular task.
For a company having all the tools (electron microscope with bells and whistles, polishing machines, microprobes) it might take hundreds to thousands of human-hours (0.1-1+ million $).
Companies like http://www.chipworks.com/ can do that, if your request is legal.
pdknsk · 2013-08-30
ceejayoz · 2019-08-13
My dad's a radiologist, so access to scans isn't an issue, but the sticking point would be knowing what's normal and what's not.
That looks like a bog-standard USB cable, but Apple's Lightning and USB-C cables are far more complicated, with actual chips onboard.
https://www.chipworks.com/blog/systems-analysis-apple-lightning-usb-ca…
I don't think there's much possibility of me vetting one of these sorts of third-party cable based on an x-ray, or even a physical teardown.
londons_explore · 2018-12-10
Uncited rumour is iPhones for the last ~3 years have had special flashlight LED's (made by NXP?) which also embed an RGB colour sensor for exactly this reason. The LED's when reverse biased send out the results of the RGB sensor. It has been suggested the reason they didn't use an extra pin was to hide this anti-flicker feature from competitors who were sure to do a detailed teardown of the phone.
There are x-rays of the LED here: https://www.chipworks.com/about-chipworks/overview/blog/true-tone-flas…
johntb86 · 2018-10-11
> That's not true. Imaging ICs (or other structures) down to sub-nm resolution is much much easier than making them. Most physics labs will have no problem doing that..
For example, see Chipworks's analysis of Intel's 14nm process:
https://www.chipworks.com/about-chipworks/overview/blog/intel%E2%80%99…
raverbashing · 2018-02-01
> Plus the part about the lightning protocol makes no sense whatsoever.
Yes it does, unless the port has an analog audio channel
It's extra hardware that has to be on the dongle (at the expense of everything else) - for Audio I'm gessing it's USB https://www.chipworks.com/about-chipworks/overview/blog/systems-analys…
> Try blind testing 10 songs or so and see if you can figure out which one is played from the adapter
Would be an interesting test, I would begin by checking the max volume allowed
ajb · 2017-06-23
Semicnductor art: https://www.chipworks.com/about-chipworks/overview/blog/silicon-art-sh…
(Editted to add See also https://web.archive.org/web/20130917012514/http://www.chipworks.com/en…
nhaehnle · 2015-02-23
Here's a report of a scan of Intel's 14nm technology: http://www.chipworks.com/en/technical-competitive-analysis/resources/b…
A good example for the problem with naming conventions is that Intel's 14nm M1 has lower pitch (i.e., lower distance between wire centres) than the 14nm technology of competitors; at least as claimed by Intel some time last years and apparently confirmed by the linked article.
rancur · 2014-12-10
well, we're hitting process node limits (5, maybe 4nm) in 10 years so I think we better stop it.
http://www.chipworks.com/media/wpmu/uploads/blogs.dir/2/files/2012/04/…
You see those little dots? Those are individual silicon atoms from one of a few fins in the gate on Intel's 22nm node.
Also, I Am A C Programmer. I Think Everything Should Be Written In C. C Is Close To The Hardware And Is Faster And Use Less Resource Compared To The JavaSlow and WebKittenz and CSS. But I Like CSS Because It Has C In It. I Am A C Programmer.
Get back to me when we have compressed ZRAM implementations and Same Page Merging everywhere to save memory on each Chrome process and each JS interpreter instance. Don't you want to do the green thing?
unwiredben · 2014-09-26
http://www.chipworks.com/en/technical-competitive-analysis/resources/b… is the source URL, the EETimes article is a repost
joezydeco · 2014-09-24
Chipworks decapped the M7 on the iphone5 and found it to be a custom model of the LPC18xx from NXP
http://www.chipworks.com/en/technical-competitive-analysis/resources/b…
http://www.nxp.com/products/microcontrollers/cortex_m3/lpc1800/
M4v3R · 2014-09-23
Not really, it's a little more complicated as Chipworks described:
http://www.chipworks.com/en/technical-competitive-analysis/resources/b…
ghshephard · 2014-09-20
"Asked whether NXP supplies Apple with its NFC chips for the new iPhone, the company declined to comment. Apple suppliers are sworn to secrecy."
From: http://www.chipworks.com/en/technical-competitive-analysis/resources/b…
"Having now depotted the 65V10 device, we can positively identify the NFC controller as the PN548, a previously unannounced variant of the PN544 and PN547 devices by NXP. Industry sources identify this as a design that was tweaked specifically for Apple by NXP. If this is true, Apple has had access to this chip for over 18 months, since the mask date (the date appearing on the die itself and indicating the year the design was finalized) is 2012. Indeed, the die markings of the PN548 vary only slightly from those of the PN54"
wmf · 2013-11-19
Caches and GPUs have extra capacity that is used to tolerate defects. Cell had 8 SPEs but shipped with 7 working. It's possible that the PS4 has 20 CUs and ships with 18 working; this would increase yield dramatically. (Update: Yep, 20 CUs: http://www.chipworks.com/en/technical-competitive-analysis/resources/b… )
jgeorge · 2013-09-04
http://www.chipworks.com/en/technical-competitive-analysis/resources/b…
Presumably Apple can tell that the USB cables are "fake" by either the exclusion of the small security bits in the cable's chip, or perhaps that large enough quantities of "fake" cables all use the same cloned chip from one original cable somewhere.
The cable may work fine, but if the OS determines that it's not a valid security chip at the end of the cable, you get the "if you electrocute yourself don't sue us" disclaimer warning.
femto · 2013-07-27
It's not quite that consumer friendly, but there are companies that provide a service like this [1]. I gather it is not cheap. Does anyone know what the rough cost is?
[1] http://www.chipworks.com/en/technical-competitive-analysis/reverse-eng…
X4 · 2013-07-01
OK, I just busted your theory. Sorry. :) It has 16G
The LSM330 has an user-selectable full scale acceleration range of ±2 g/±4 g/±6 g/±8 g/±16 g and angular rate range of ±250/±500/±2000 d.
Samsung Galaxy S4 accelerometers/gyro Datasheet:
http://www.st.com/web/en/catalog/sense_power/FM89/SC1448/PF253882
Here's a FULL Teardown of the insides of the Samsung Galaxy S4. http://www.chipworks.com/blog/recentteardowns/2013/04/25/inside-the-sa…
StringyBob · 2013-03-08
Can't let a post like this go by without name-checking a couple of other sites:
Visual 6502 for taking such photos and then emulating, or should I say simulating, the cpu from the images - in JavaScript: http://visual6502.org/
Chipworks for commercial silicon analysis of big modern system on chips (source of some photos of recent apple silicon): http://www.chipworks.com/blog/technologyblog/2013/01/21/the-straight-d…
sparky · 2013-03-04
How does this prediction jibe with the statement that a larger sensor will always be beneficial, especially in low light?
Will smartphones be able to include much larger sensors in the future?
Is there some new physics that obviates the need for a large sensor?
The 5D Mark III has > 24x the sensor area of an iPhone 5, and most smartphones are even worse off [0] [1]
[0] http://en.wikipedia.org/wiki/List_of_large_sensor_interchangeable-lens…
[1] http://www.chipworks.com/blog/recentteardowns/2012/09/21/apple-iphone-…
mturmon · 2012-04-12
Agree.
It does seem like the images are real, because apparently the site is in the business of selling very high-res versions of such chips:
http://www.chipworks.com/en/technical-competitive-analysis/report-form…
"Die photos are available in the Chipworks Report Store at top metal and lower metal/poly"
Steko · 2012-03-16
Chipworks actually confirms it's 45 nm now.
http://www.chipworks.com/en/technical-competitive-analysis/resources/r…
nickpsecurity · 2016-10-09
The proprietary companies already do this looking for hidden stuff. Mostly patent violations, though, so they can scheme some licensing revenue out of the product they're inspecting. Here's the company that does much of it:
http://www.chipworks.com/about-chipworks/overview/blog/apple-iphone-7-…
In that one, you'll see the other end of the process where many mobile chips will get extremely specialized. Notice this one has several RF chips instead of an integrated one to attempt to hit some size/price/energy/performance tradeoff. Plus Apple sells premium stuff whether they can put in more, nicer parts. Notice there's 10 different chips involved in just the RF front-end part. :)
Here's a nice PDF with illustrations of the RE processes themselves plus what things look like:
https://www.iacr.org/archive/ches2009/57470361/57470361.pdf
I like Fig 3 especially as a lot of people forget there's circuitry inside their PCB's rather than just on top or bottom. That cell phone has 9 layers of wiring. :)