A Trusted Execution Environment (TEE) is a secure area within a device's central processor isolated from the regular operating system and other applications. It provides a protected space for executing code and processing sensitive data with enhanced security guarantees. You can explore more about TEEs through this Perplexity thread.
Find research from the the CryptoSat team (SpaceComputer advisors Yan Michalevsky and Yonatan Winetraub) here: SpaceTEE: Secure and Tamper-Proof Computing in Space using CubeSats.
Although the current hardware is limited, we anticipate significant improvements in the near future. In the meantime, private benchmarks have demonstrated that the existing setup provides sufficient capacity for block production and the cryptographic signatures required for active validation.
True Random Number Generators (TRNGs) serve several critical purposes across different domains:
TRNGs are essential for cryptographic operations, where they generate encryption keys, nonces, and initial values to secure data transmission and storage. Their unpredictability makes them crucial for maintaining strong security protocols and protecting sensitive information
The gaming and gambling industries rely heavily on TRNGs to:
True random number generators are particularly valuable because they:
Our cosmicRNG relies on Muon-Ra: Quantum random number generation from cosmic rays which leveraging cosmic rays coming from deep space and our center of solar system.
This CosmicRNG can be used on various ceremonies where there is strict requirement for high quality entropy, i.e. trusted setup, something can't be manipulated. We will be expanding more on this and trying demonstrate with small mini-game. You can meanwhile learn more about HRNG
We currently utilize the Iridium network and will soon transition to Iridium NEXT, which is expected to support data speeds of up to 512 kbps via radio frequency (RF) communication.
In the future, we plan to adopt optical (laser) communication systems, which offer significantly higher bandwidth and lower data costs. Optical systems can deliver data rates up to 100 times faster than the fastest RF technologies—while RF typically peaks at around 1 Gbps, optical communications can exceed 10 Gbps.
For example, NASA's DSOC system recently achieved a transmission rate of 267 Mbps from a distance of 30 million kilometers aboard the Psyche spacecraft.
The current daily data rate per node is 250KB, though ongoing experiments suggest it could increase to 300MB with the existing setup.
Latency can vary significantly but remains predictable based on orbital mechanics. Transmitting 1KB to a specific node in the network currently takes anywhere from a few seconds to up to 30 minutes. In the future, we expect this process to become near-instantaneous.
Our current orbit time is approximately 90 minutes, and we fly at a speed of approximately 7km per second. Each orbit repeats approximately every ~ two weeks.
Join the official SpaceComputer Telegram group. We're terminally online and always happy to chat!
While we currently rely on the Iridium network, and Starlink integration will take time, we don't deny that in the future, we can leverage constellations from Kuiper and OneWeb. More satellite internet constellations will be available.
You can find the SpaceComputer media kit here.
We are currently in discussions with multiple satellite constellations, with over 100 satellites potentially available for SpaceComputer's use. This would significantly enhance our space computing capacity and exponentially increase network bandwidth.
While we aim to increase our network bandwidth beyond 1GB cumulatively—and eventually exceed 10GB per day—current limitations stem from the underlying hardware. We are actively working to overcome these constraints and expand bandwidth to support a wider range of applications. Think of it as the early Bitcoin era (2009), with a gradual evolution toward the greater capacity seen in Ethereum around 2014.
Theoretically, it is possible to extend the SpaceComputer network between Earth and Mars, based on a few key assumptions.
Light takes approximately 12.5 minutes to travel from Earth to Mars at their average distance of 225 million kilometers.
The actual travel time fluctuates depending on the planets' positions:
Closest Approach
Farthest Distance
If the adversary gets access to the satellite before launch, they could tamper with the spacecraft's hardware and software. One way to make the solution tamper-evident is to coat the computer with Kapton tape and measure the satellite's inertial moment before launch.
This measurement is compared to the observed rotation rate of the satellite once in orbit. This can be measured by tracking the antenna's range rate as the satellite revolves or via direct imaging of the satellite (depending on launch configuration).
By measuring the rotation rate changes, we can estimate the satellite's actual moment of inertia and verify that it is identical to the expected one measured from before launch. Tampering with the coated hardware or adding a new one inevitably changes the inertia moment.
Further analysis and simulations are required to assess this method's constraints. Although it is possible to reintroduce the original inertial moment, it is challenging to do so within the time window of a launch effort.
In addition, during a launch, the rocket's vibrations on the satellite are likely to shake apart and destroy any component that was not correctly installed and tested for vibrations.
You can read more about this in the SpaceTEE paper.
Space data centers can utilize high-intensity, uninterrupted solar power unhindered by day/night cycles, weather, and atmospheric losses (attenuation). This enables orders of magnitude lower marginal energy costs. Compared to their terrestrial counterparts, space-based data centers like SpaceComputer will eventually offer drastically lower operating costs. Learn more in this recently published study.
The cost comparison of operating a single 40 MW data center cluster for ten years on land versus in space is striking: $167 million on Earth versus just $8.2 million in space. This substantial difference highlights the economic advantage of space-based solutions. Source-Lumen Orbit Whitepaper
The global data center market size was estimated at USD 256.05 billion in 2024 and is projected to reach around USD 775.73 billion by 2034, growing at a CAGR of 11.72% - Source.
The advantages of space-based data centers are:
Estimating the percentage of the market that could be addressable by space-based solutions given unique advantages and the fact that space-based data centers are still a nascent concept, we can conservatively estimate that they could potentially capture 5-10% of the overall data center market in the long term.
Using the projected market size for 2034 (USD 775.73 billion) and assuming a 5-10% potential market share: TAM for space-based data centers (2034) = USD 775.73 billion * (5% to 10%)= USD 38.79 billion to USD 77.57 billion
We will collaborate with space partners to verify our proof of authenticity in orbit and are broadly interested in remote attestation methods.
Specifically, we will work with Iridium—a trusted authority in location verification—to authenticate our signals from space. By leveraging the modem's transmission location, which changes rapidly across Earth's continents in under 90 minutes, we can demonstrate orbital velocity as proof of location. This speed is only achievable in space, serving as a strong validator. We aim to work with Iridium to generate signed, verifiable transmission data and to make this resource publicly accessible, as it is currently limited to private satellite contractors.
Verification through amateur RF antenna is impossible, even through public SatNogs infrastructure.
Verification would require covering the costs of obtaining a license and gaining access to AWS Ground Station or, alternatively, setting up advanced antenna infrastructure to enable two-way transmission.
Our satellite catalog numbers, known as NORAD ID, are the following.
The world is approaching a major shift in launch economics, driven by the emergence of partially or fully reusable heavy-lift launch vehicles. These new systems are expected to bring launch costs down to around $5 million per mission over the long term. This dramatic reduction is set to democratize access to space, enabling a broader range of industries to take advantage of space-based technologies. With a payload capacity of up to 100 tons to Low Earth Orbit (LEO) or Sun-Synchronous Orbit (SSO), this equates to roughly $30 per kilogram.
Costs could drop to as low as $10 per kilogram.
At these price points, launch costs are no longer a primary driver for orbital data centers.
SpaceComputer's current launch costs can reach up to $1 million per satellite, but we anticipate a steady decline, with a target to reduce costs to around $100,000 per unit.
We are leveraging the power-efficient Arm v8 64-bit architecture, which aligns with our current power budget. As hardware advances, we expect to support increasingly complex space-based computation.
Current technical specifications include:
While ARM and RISC-V differ in their architectural approaches to security, it could be feasible. For now, we are conducting experiments using the ARM v8 hardware already deployed in orbit. However, any switch would require extensive testing to ensure reliability and justify the transition.
While we are actively working on implementing a SpaceTEE prototype, we are also exploring the use of specialized hardware in orbit, including ARM TrustZones, to significantly enhance our security model. The timeline for obtaining hardware with ARM TrustZone is still being finalized, and we are currently working to get clearer estimates.
Learn more about TrustZone:
Kessler Syndrome is a theoretical cascade effect where space debris in low Earth orbit (LEO) creates a self-sustaining cycle of collisions. Named after NASA scientist Donald J. Kessler who proposed it in 1978, this phenomenon has become increasingly relevant to our modern space activities.
There is currently a huge amount of debris in Earth's orbit:
Potential solutions are proposed:
You can learn more about Kessler Syndrome in this video.
Space traffic is becoming increasingly congested, with thousands of satellites orbiting Earth and raising the risk of potential collisions. The main challenge is predicting and preventing these collisions while preserving the confidentiality of sensitive satellite position data.
Many governments and satellite operators are reluctant to disclose exact satellite locations due to safety, security, or political concerns.
In 2008, the Iridium 33 and Kosmos-2251 satellites collided due to a lack of precise trajectory disclosure by governments. According to publicly available data, the satellites should have passed each other with a buffer of 584 meters.
The solution involves using Sharemind MPC technology to perform collision detection calculations without revealing the actual positions of satellites. This allows satellite operators to:
The technology enables satellite operators to receive collision warnings while keeping their exact satellite positions private. This balance between security and safety represents a significant advancement in space traffic management
Read more about how current PoC works in wild to encrypt trajectories.
We can enable immediate communication between satellites, allowing collision courses to be calculated directly within the SpaceComputer network, all within Earth's orbit. This approach enhances security by eliminating the need to transmit sensitive data outside the satellite network.
With the SpaceComputer Proof-of-Location entropy protocal, we can assign constantly-in-orbit satellites with the task of verifying the location of a particular device on Earth.
SpaceComputer is a blockchain that utilizes available computing resources on satellites in orbit. It supports a range of applications, including secure data storage and decentralized computing. You can read about specific use cases on our website.
We aim to train and run large language models (LLMs) in orbit and are actively working to secure additional computational capacity. In the near future, running LLMs on space-based data centers and networks like SpaceComputer will become feasible.
On the SpaceComputer blockchain, it will be possible to run virtually any dApp, whether UTXO-based, EVM, or SVM. Read more about use cases currently in development here.
Currently, space computation capacity is limited compared to traditional data center resources. However, the future of space computing is promising. Operating a space-based data center in orbit will ultimately be far more feasible thanks to infinite access to solar energy and a more robust theoretical security model.
We recently talked at DeCompute 2024 in Singapore and during Trusted Espresso. The slides are available here.