Jane Street's ASIC puzzle: 400 submissions, one hidden Star Battle chip
Results from the Jane Street hardware puzzle
Jane Street's August challenge asked solvers to reverse-engineer a small chip from its GDS layout alone. About 400 submissions arrived from over 30 countries, using tools like KLayout, Yosys, and Z3, with custom code in Python, Rust, C++, and more. The chip turned out to be a hardware checker for an 11x11 Star Battle puzzle, emitting "TWO STARS" on success. The post details clever solutions, including SAT solving, FPGA synthesis, and even a Minecraft recreation.
A row of marks on an otherwise unused layer below the die spells PER ARENAM AD ASTRA in Morse code: “through the sand, to the stars”
- babush
I asked Claude to "fuzz the circuit" and, while it didn't work, it basically ended up going the Z3/SMT route and solving it in ~3 hours. This is after the writeups were already out, but still, as a software reverser, pretty wild. Writeup: https://www.babush.me/how-not-to-solve-jane-streets-asic-puz...
- kayson
The EE/IC Design part of me loves this for trying to build a community around what is typically a very opaque and secretive industry. The social justice / millenial part of me hates this for being funded by high frequency trading, of all things. What a colossal waste of resources.
- ryontikygod
Every time I see a Jane Street puzzle of this like I wonder in which way you can improve your problem solving skills for these things and how to extrapolate patterns from other things. I have a gut feeling that this is not only just grinding puzzles, there has to be a way to improve this. Curious if people have some insights on this
- amelius
Were AI models able to find the solution and easter eggs?
- IshKebab
> The hardware team at Jane Street designs FPGAs and ASICs that run some of the fastest trading systems in the world, and the day-to-day work is full of puzzles just like this one: staring at a sea of gates, timing reports, or waveforms and slowly teasing out what’s really going on.
Although not really any more. AI is really good at debugging failing tests so I no longer have to spend hours - sometimes it was even days - staring at waves and instruction traces.
Tbh that's one thing I won't miss.