[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fjrv88jXoUgL7qlGlDdbwaAMz-MOV9ZtDzU-X9Bhbgp0":3},{"article":4,"iocs":50,"watch_terms":51,"solutions":52,"lesson":84},{"id":5,"title":6,"slug":7,"summary":8,"ai_summary":9,"brief":10,"url":11,"image_url":12,"published_at":13,"ingested_at":14,"relevance_score":15,"entities":16,"category_id":32,"category":33,"article_tags":37},"c27ba2a1-7d34-48eb-986a-989dcb3613c5","Confidential Computing on CPU and GPU Systems: How AI Data Centers Protect Data in Use","confidential-computing-on-cpu-and-gpu-systems-how-ai-data-centers-protect-data-i-441fb9","Modern AI runs on shared, high-performance infrastructure that processes enormous volumes of sensitive data and valuable model weights.…","Confidential computing addresses the challenge of protecting data while it's being processed in memory, a critical gap in traditional encryption methods. This technology utilizes Trusted Execution Environments (TEEs) on CPUs and, more recently, GPUs, to isolate sensitive computations and data from the underlying infrastructure. The integration of confidential computing directly onto GPUs, like NVIDIA's Hopper generation, is a significant advancement for AI data centers, ensuring that model weights and data remain protected during intensive computations.","Confidential computing protects sensitive data and AI models in use within data centers.","https:\u002F\u002Fhackread.com\u002Fconfidential-computing-cpu-gpu-systems-ai-data-centers\u002F","https:\u002F\u002Fhackread.com\u002Fwp-content\u002Fuploads\u002F2026\u002F07\u002Fconfidential-computing-cpu-gpu-systems-ai-data-centers.jpg","2026-07-28T11:24:52+00:00","2026-07-28T12:00:07.086532+00:00",7,[17,20,22,25,28,30],{"name":18,"type":19},"Confidential Computing","technology",{"name":21,"type":19},"Trusted Execution Environment",{"name":23,"type":24},"Hopper","product",{"name":26,"type":27},"NVIDIA","vendor",{"name":29,"type":19},"CPU",{"name":31,"type":19},"GPU","c70f3a41-2f0c-4608-870d-b8cbcd8be076",{"id":32,"icon":34,"name":35,"slug":36},null,"Cloud Security","cloud-security",[38,43,48],{"category":39},{"id":40,"icon":34,"name":41,"slug":42},"0493c7e9-989a-4692-b4e6-136f5ec09675","Cryptography","cryptography",{"category":44},{"id":45,"icon":34,"name":46,"slug":47},"839da5c1-3c34-47e2-9499-f7201640e3ac","AI Security","ai-security",{"category":49},{"id":32,"icon":34,"name":35,"slug":36},[],[26,23],{"controls":53,"vendors":83},[54],{"slug":55,"name":56,"problem":57,"products":58},"encryption-key-management","Encryption and key management","Stolen data stays unreadable, and the keys stay under your control.",[59,67,75],{"vendor_slug":60,"vendor_name":61,"vendor_logo":62,"product_slug":63,"product_name":64,"url":65,"description":66},"utimaco","Utimaco","\u002Fvendor-logos\u002Futimaco.png","utimaco-hsm","Utimaco General Purpose HSM","https:\u002F\u002Futimaco.com\u002Fdata-protection\u002Fgp-hsm","Hardware security modules for key generation, storage and cryptographic operations.",{"vendor_slug":68,"vendor_name":69,"vendor_logo":70,"product_slug":71,"product_name":72,"url":73,"description":74},"thales","Thales","\u002Fvendor-logos\u002Fthales.png","ciphertrust","CipherTrust Data Security Platform","https:\u002F\u002Fcpl.thalesgroup.com\u002Fencryption\u002Fdata-security-platform","Data discovery, encryption and central key management across on-premises and cloud.",{"vendor_slug":76,"vendor_name":77,"vendor_logo":78,"product_slug":79,"product_name":80,"url":81,"description":82},"hashicorp","HashiCorp","\u002Fvendor-logos\u002Fhashicorp.png","vault","HashiCorp Vault","https:\u002F\u002Fwww.hashicorp.com\u002Fen\u002Fproducts\u002Fvault","Secrets management with short-lived credentials and encryption as a service.",[],{"slug":85,"title":86,"body":87,"framework_refs":88,"framework_notes":99},"confidential-computing-protecting-data-in-use-with-tees-on-cpus-and-gpus","Confidential Computing: Protecting Data in Use with TEEs on CPUs and GPUs","Traditional encryption secures data at rest and in transit, but leaves data exposed in plaintext while being processed in memory — a critical vulnerability in AI data centers handling sensitive model weights and proprietary datasets. Confidential computing addresses this gap by leveraging Trusted Execution Environments (TEEs) to isolate sensitive computations from the host infrastructure, hypervisors, and even privileged administrators. The extension of TEEs to GPUs, such as NVIDIA's Hopper architecture, is particularly significant as AI workloads increasingly rely on GPU processing for large-scale model training and inference. Without these protections, any compromise of the underlying infrastructure — whether by a malicious insider, cloud provider breach, or supply chain attack — could expose the most sensitive data at its most vulnerable moment. Organizations running AI workloads must treat data-in-use protection as a first-class security requirement, not an afterthought.",[89,90,91,92,93,94,95,96,97,98],"NIST SP 800-111 (Storage Encryption)","NIST SP 800-223 (AI Risk Management)","NIST AI RMF 1.0 – Govern & Protect","CIS Control 3 – Data Protection","CIS Control 4 – Secure Configuration of Enterprise Assets","ISO\u002FIEC 27001 – A.8.24 (Use of Cryptography)","GDPR Article 25 – Data Protection by Design and by Default","GDPR Article 32 – Security of Processing","NIST SP 800-190 – Application Container Security","NIST Zero Trust Architecture SP 800-207",{}]