3 posts
Phala Network has disclosed a security incident involving a Phala Cloud API vulnerability. While workload metadata was exposed, the core confidential computing environments remained secure.
Phala Cloud has successfully run GLM-5.2 with a 1 million token context on a single 8xH200 node. This breakthrough enables secure, large-scale AI inference within a trusted execution environment without sacrificing performance.
Phala Cloud introduces GPU TEE to enable confidential AI model deployment for regulated industries. By leveraging Intel TDX and NVIDIA H100/H200 hardware, enterprises can now run sensitive AI workloads with hardware-backed isolation and verifiable compute.
