Ktema SystemsCorpus HubOne node, any scaleRequest a demoCorpus View

One node, any scale

Bare metalKtema

A part with no operating system writes a record into its own flash and reads it back by its address after a reset.

The smallest host in the claim: a microcontroller with no operating system, 8 MB of flash and 64 kB of working memory, writing a record that survives a reset. The same build runs on every host above it, and a deployment can be a field of these nodes.

The ESP32-S3 module on its board, lit and connected.
The part the run was made on: an ESP32-S3 with 8 MB of flash and 64 kB of working memory, running from the board's own flash. Only the application and the partition table were written, so the shipped bootloader starts it.

What the board reported

Over the RAM mediumOver the flash region
Heap peak 2,384 B 1,648 B
Arena 65,552 B 65,552 B
Medium The engine's own map in RAM A 2 MiB partition at 0x130000
Records Two written and walked back in one boot Three from three earlier boots, each reached by its coordinate after a reset
Application 356,576 B in a 1 MiB slot 447,968 B
Bootloader Untouched Untouched

The flash run peaks lower than the RAM run because the region channel holds no map in memory. That is the same trade the host measurements show from the other side: the path that keeps nothing per record is the path that answers by reading the log.

What the board printed

The table above is the board's own console output, with each coordinate written in the engine's ASCII form. The first run kept its records in the engine's RAM medium; the second kept them in the flash partition and read them back after each reset.

Two console blocks. The first reports an arena of 65552 bytes, two facts and a heap peak of 2384 bytes. The second reports a 2 MiB partition at 0x130000, three records kept from earlier boots, and a heap peak of 1648 bytes.
The two runs as recorded. A coordinate is sixty characters: twenty axes of three, in the engine's own alphabet.
arena 65552 bytes, minimum heap 65536
write id=0011xj56j0ew0ar2ps1u60xz2jq2dr38q7eo17s4mc3a13tl1zp49s7td7au
facts=2
heap used=1168 peak=2384
ktema: ok

partition ktema at 0x130000, 2097152 bytes fact 0002tt8lr6vp1ao0dv3bt0m74o678b8im06r3m131p0qx1bh57n76i09v39i uptime=72ms fact 0003lv56v7fn44a86t5jk7e73m045x1tc0yg16w2l06p486j5ro5q92eo2bb uptime=224ms fact 0005bs1i50c77tb0ir7k611k7yt7v387j5fb1mp13s8lw4s723y6w56y96tm uptime=85ms kept from earlier boots: 3 wrote 00000l6dc1fy4em3pw8ah3j66bv2o85nz3zn0pk4we1ld2cw1br1im35h2zl heap used=0 peak=1648

The surface the firmware links

A part like this has a flash region, a write, and an erase. It has no key-value medium, so the C surface takes that shape rather than a layout: the region as a slice the engine can read, and two callbacks for the writes and the erases that the driver owns.

The region channel sits above it and carries the log, the segment index, the forward-only write rule, and the recovery of a torn tail. The firmware writes no format and keeps no index of its own, which is why the same application runs unchanged here and on the host.

What the board does not cover yet

Two cases are modelled against a slice and have not run on a part: a power cut in the middle of a write, and a region that has filled and retired its oldest segment. Until both run on the board, the durability claim covers the reopen that the run above performs, and does not cover those two.

Firmware
firmware/litewing, ESP-IDF
Entry point
ktema_open_over_storage, ABI 6
Recorded
2026-10-24