AxoGrid

Biological computing. On demand.

AxoGrid is cloud infrastructure for biological computing.

Your code

Network

Our facility

The Grid is a population of BPU modules, cultured, powered, and monitored in our facility. You reach them over the network.

We operate

  • We culture and maintain the neurons.
  • We operate the hardware.
  • We manage the biological environment.

You run

  • You connect through software and train the network for your task.

No wet lab. No custom hardware stack. No months of integration.

The platform

We integrate all layers in the stack into one platform.

All five layers normally have to work together before you can run a single useful biological computing experiment. AxoGrid integrates them for you.

05 Experiment management You control this
04 Closed loop API You control this
03 BPU infrastructure We operate
02 Standardized biological environment We operate
01 Living neural networks We operate

Reproducibility

Low variance is key for reproducibility.

BPUs across AxoGrid are prepared and maintained under standardized conditions to produce consistent baseline activity across cultures.

That means changes in network behaviour can be more confidently associated with your experiment and training protocol rather than with differences between cultures.

Without standardization

Baseline activity per culture

Across AxoGrid

Baseline activity per culture

Diagram of the design goal, not measured data. Cultures are prepared and maintained to fall within a narrow activity band, so a change you observe is more likely to come from your protocol than from the culture.

Consistent biology. Comparable experiments.

BPUs inside the Grid

Two classes of biological compute.

Both run a 2D neuronal culture with spontaneous tonic firing. They differ in how much of that network you can reach at once.

AxoNano

For exploratory workloads

Accessible biological compute for smaller experiments.

60 channels across the culture

Channels
60 simultaneous stimulation & recording
Culture
2D neuronal culture
Activity
Spontaneous tonic firing
Inactivity
<10%
Throughput
~1,000 spikes/second
~1,000 spikes/second

AxoMini

For higher density experiments

Higher density compute for larger experiments.

2,000 channels across the same culture

Channels
2,000 simultaneous stimulation & recording
Culture
2D neuronal culture
Activity
Spontaneous tonic firing
Inactivity
<10%
Throughput
~40,000 spikes/second
~40,000 spikes/second

What the numbers mean

Channels
Each channel is a point where the system can stimulate the network and record from it. More channels means finer control over where you send signals, and a more detailed view of how the network responds.
Spikes per second
A spike is a single neuron firing. This is the rate of activity the system observes across the culture, so it sets how much signal you have to work with in a given run.
Inactivity
The share of channels not detecting activity. A lower figure means more of the array is in contact with a live, firing network.

Your next biological computing experiment doesn't need to start with building a lab.

Access living neural networks, BPU hardware, and the software required to train them through AxoGrid.

Bring the experiment. We'll provide the biological compute.

Request AxoGrid access