Targon (CRYPTO: SN4) — Company Overview, News & Financial Data
Targon is a confidential compute cloud platform and decentralized compute network engineered to support privacy-preserving artificial intelligence workloads.
Targon
About
Targon is a confidential compute cloud platform and decentralized compute network engineered to support privacy-preserving artificial intelligence workloads. Operated by Manifold Labs Inc., the network aggregates high-performance graphics processing units and central processing units within hardware-enforced trusted execution environments. Utilizing Intel TDX memory encryption and NVIDIA confidential computing technologies, it delivers zero-trust infrastructure for model training, inference, and agentic execution. Integrated as Subnet 4 on the Bittensor ecosystem, the network cryptographically validates compute performance and incentivizes hardware providers to supply confidential processing capacity for secure data operations across distributed cloud environments.
Snapshot
Operations
Products and/or services of Targon
- Confidential Virtual Machines: On-demand cloud virtual machine instances powered by Intel TDX and NVIDIA H100, H200, and B300 GPUs, providing hardware-encrypted compute infrastructure for secure AI model training and enterprise workload execution.
- Developer Sandboxes: Isolated runtime execution environments designed for agentic workflows, software evaluations, and secure code testing, allowing developers to run untrusted code without exposing host hardware, internal networks, or sensitive underlying data.
- Compute Supply Marketplace: A decentralized portal enabling independent data centers and hardware providers to connect GPU and CPU capacity to the network, routing idle compute resources to end users while earning cryptographic rewards.
- Hardware Attestation Architecture: A zero-trust security frame utilizing Intel TDX hardware-backed memory encryption and NVIDIA PCIe confidential computing to cryptographically verify hardware integrity and protect workloads in transit and at rest.
- Bittensor Subnet 4 Protocol: A decentralized consensus and validation layer operating as Subnet 4 on the Bittensor network, evaluating miner performance, benchmarking confidential AI inference queries, and distributing network emission rewards.
- High-Performance Compute Infrastructure: Scalable compute rentals offering pay-per-use access to high-density GPU and CPU clusters, optimized for continuous zero-downtime execution of distributed artificial intelligence training, model hosting, and inference workloads.