About

We build the silicon that lets machines think for themselves.

NeuralPeak is a small team of chip designers and systems engineers who believe edge AI deserves purpose-built hardware — not a server accelerator squeezed into a smaller box.

Our story

It started with accelerators that kept getting lost.

Dr. Elena Marchetti spent the better part of fifteen years designing neural accelerators — the kind that end up as one block inside someone else's system-on-chip, budgeted down until they barely matter. The pattern was always the same: the accelerator worked, and then the power budget, the die area, and the schedule all carved it back.

At a compiler workshop in 2020, she met Dr. Tomas Reyes, who'd been building the software side of the same story — squeezing models onto hardware that was never quite built for them. The two of them kept returning to the same question: what would the silicon look like if it were designed for edge inference from the first floorplan, not retrofitted?

"Edge AI kept getting the leftover silicon. We decided to build the silicon first, and let everything else follow."

In 2021 they founded NeuralPeak. The first year was a single test chip and a lot of late nights; the second year brought NP-1 back from the fab, and it worked the way they'd hoped. Today the company designs, tapes out, and ships its own neural processors — and stays deliberately small enough that the people answering your questions are the ones who drew the floorplan.

Team

The people behind the die.

A team of 24, most of them silicon and systems engineers. Here are the ones you'll actually talk to.

EM
Dr. Elena Marchetti
Co-founder · CEO
Fifteen years in mixed-signal and neural accelerator design. Still reviews the top-level floorplan herself.
TR
Dr. Tomas Reyes
Co-founder · Chief Architect
Compilers and systems. Owns the instruction set and the PeakStack toolchain end to end.
AB
Aisha Bello
VP of Silicon
Leads physical design and tape-outs. Ran back-end teams on two prior generations of edge silicon.
MC
Marcus Chen
VP of Engineering
Owns bring-up, evaluation boards, and the benchmark harness — the numbers come from his lab.
SL
Sofia Lindqvist
Head of Product & Partnerships
Turns partner requirements into silicon requirements, and vice versa. Your first call.
JW
Jonas Weber
Head of Compiler
Leads the quantization and scheduling work behind PeakStack. Obsessed with "one pass to flash."
Timeline

Five years, one continuous tape-out.

2021

Founded

NeuralPeak is founded by Elena Marchetti and Tomas Reyes. First prototype neural-engine test chip tapes out later that year.

2022

NP-1 first silicon

First silicon returns from the fab and passes bring-up. The architecture holds up — the tiled compute array runs as simulated.

2023

NP-1 in production

First design wins ship in robotics and smart-camera systems. PeakStack reaches a stable 1.0 release.

2024

NP-C1 & PeakStack 2.0

The compact vision processor launches, and the toolchain gains automatic sparsity and mixed-precision policies.

2025

NP-2 tape-out

Our next-generation die is taped out, and the partner program is formalized around evaluation kits and reference designs.

2026

NP-2 first silicon

NP-2 comes back from the fab and passes bring-up. PeakStack 2.1 brings transformer and on-device LLM inference to the family.

Contact

Talk to us about what you're building.

Whether you're evaluating silicon for a product or just curious about the architecture, we read every message and reply from the engineering side.

Sample requests
samples@neuralpeak.cloud

For teams evaluating NP-1 or NP-C1. Tell us your use case and we'll route it to the right engineer.

Support
support@neuralpeak.cloud

For existing partners with boards in the field. Escalates directly to our silicon team.

Press & media
press@neuralpeak.cloud

Benchmark methodology, product imagery, and background for coverage.