I am new to hardware engineering, pen testing and modification. I have some experience from the past with Arduino boards, bus pirate but I do code for computers regularly.
I broke out my bus pirate and Arduino for projects in robotics, pen testing hardware manipulation. I want to expand my kit/gear within hardware manipulation and testing. But I have a different budget now due to some issues. I see 2 devices that pop up under hardware pen testing, programming and diagnostics: ChipWhisperer and faultier. I was wondering what is the difference between CW as opposed faultier, and bus pirate? The big differences….am I missing out on function by choosing one over the other or is it a matter of quality and power/stability.
I have no experience with Faultier. It seems closer to our Nano in capabilities (and cost). I invite to you compare specs; if you have more specific questions by all means fire away. It may also be useful to compare the associated learning material; ours is here.
I am going to sound horrible but where do I find prices for the nano for a single base unit? I may be looking for it wrong, all I can find is the more expensive husky model and larger kits on 3rd party websites. I prefer to support the creator of the unit (you guys) rather than a 3rd party if possible . So if your company sells these nanos in single units direct that would be perfect.
Ok I have seen specs for all your CW series. Just was not sure if say the husky or the bigger kits have additional features or if they are simply more power and range. I primarily am expanding my penetration testing hardware. I was a coder in the 90s and decided to go down the pen test route after playing with things like flipper Z and SDRs. I note that there is not very many options for direct hardware manipulation beyond logic analysis, EMFI. Glitch. I am just starting out in the hardware area (beyond F0 and SDR) and have a bit of a budget at the moment . Will the smaller CW units provide enough function to start me out on basic things as far as functionality and versatility ?? Or would I be better off just waiting a while and get the big boy??
There is a lot to digest there. At a high level, Nano is a good learning platform. While it can be used on other targets, there are lots of limitations and it’s not the best tool for that.
Lite/Husky are much more similar. Their glitching capabilities are similar. Husky has much deeper storage (for longer captures), a much faster ADC, and many more triggering features – all of which can make it a much more useful tool for working with “real” targets.
Greml1n You have two or more options here. 1. something complex but more useful (chipwhisperer) 2. something simple but less useful.
Faultier can generate crowbar glitches with similar or even higher resolution than Chipwhisperer Lite, but the delay resolution (trigger offset) is limited compared to Chipwhisperer Lite’s capabilities, and this parameter is most important when sending glitches. An RP2040 or even an RP2350 is a good processor for generating glitches, but it will never match a real FPGA in terms of internal delay counters. Apparently, the Husky can generate internal delays using Artix counters up to 1.2 GHz. I have an old Chipwhisperer Lite, but I use a Sipeed Tang Nano FPGA programmed with my own code to generate glitches. I haven’t calculated exactly how much resolution I can get on the internal timer, but the code assumptions predict a resolution in picoseconds equal to the Artix A7, the only plus is that the Sipeed Tang Nano 4k is much cheaper.
Greml1n - Faultier should be enough to generate glitches in less complex processors, however if you come across something quite advanced you may need synchronization with the DUT clock to inject a crowbar glitch in a given clock cycle of a given narrow or disrupted operation, then fpga gowin or chipwhisperer will come in handy.
I have some doubts here. The ARM you’re using in the ChipWhisperer Nano is just an ARM. The RP2040 and RP2350 have functions similar to a programmable FPGA, which is mainly why they can generate precise pulse spikes on the GPIO outputs. For example, 55ns, 56ns, and 57ns, whereas another processor with a pure ARM core won’t be able to generate such precise pulses. It might be able to generate 40ns, 66ns, 81ns, etc. So, this is the main advantage of the RP2040 and RP2350, but their internal timers won’t be able to match a real FPGA’s trigger offset. This parameter is the most important one, and cwlite, cwhusky, can perform it.