#GCPConfidentialComputing

[govindhtech]

How GCP Confidential Computing develops trusted AI

How Confidential Computing Builds Trusted AI

GCP Confidential Computing changed enterprise cloud-based sensitive workload management. Google's hardware ecosystem development makes Confidential Computing compatible with privacy-preserving AI and multi-party data analytics, boosting adoption.

Google Cloud unveils its latest Confidential Computing advancements, illustrating how clients protect their most sensitive workloads, including AI.

Newest innovations

A glance at GKE Nodes and private virtual machines with NVIDIA H100 GPUs for AI applications.

A private Vertex AI Workbench preview

An overview of Confidential Space using the widely available NVIDIA H100 and Intel TDX GPUs.

General access, secret GKE Nodes on C3 systems with integrated acceleration and Intel TDX.

On AMD SEV-SNP N2D computers, confidential GKE nodes are available.

Private AMD SEV virtual machines on C4D machines preview

Speed up your Confidential Compute trip with Gemini Cloud Assist's preview.

Confidential Computing New Applications

Businesses using GCP Confidential Computing to unlock commercial breakthroughs affect all major industries.

AiGenomix

AiGenomix uses Google Cloud Confidential Computing to provide unique infectious disease surveillance, early cancer detection, and pharmaceuticals intelligence with a global network of public and private partners.

Google Ads

Google Ads uses confidential matching to link first-party customer data for marketing. Google Ads products are the first to employ confidential computing, and additional companies will follow suit.

Swift

Swift is using GCP Confidential Computing to power a money laundering detection model while protecting the privacy of some of the biggest banks.

Confidential Computing

Confidential VMs, GKE, Dataflow, Dataproc, and Space safeguard data in use.

Encrypt data during processing.

A simple, user-friendly implementation does not sacrifice performance.

Confident collaboration with data ownership

Benefits

A secrecy breakthrough

Confidential Virtual Machines, a breakthrough technology, lets customers secure their cloud data during processing.

Simply understood by all

Google Cloud lets users encrypt data in-app without changing app code or affecting speed.

Creating new possibilities

Confidential Computing opens new doors. Companies can collaborate while protecting data.

GCP Confidential Computing

Secret Computing Platform

Secret VMs

Cloud data can be protected by confidential virtual machines (VMs) encrypting data-in-use during processing. Modern AMD, Intel, and other CPUs have security protections for secret virtual machines. With GCP Confidential Computing, customers can trust their cloud data to remain private.

Google Cloud uses the Intel AMX CPU accelerator on the general-purpose C3 machine class for Confidential VMs for AI/ML applications. Confidential VMs on the C3 machine series secure AI data and models at the hardware level and boost deep learning and inference performance.

Confidential H100 GPU VMs

Companies can use AI and machine learning while securing sensitive data using accelerator-optimized A3 machine series with NVIDIA H100 GPUs' confidential virtual machines (VMs). With H100 GPUs, confidential virtual machines (VMs) protect data from GPU entry to output. This reduces the risk of privileged users or malicious actors gaining unauthorised access. Confidential VMs on the A3 machine line provide a trusted execution environment for AI applications, enabling businesses to collaborate more freely and securely.

Secret GKE Nodes

Confidential GKE Nodes encrypt GKE cluster data while retaining performance. The same technology powers Confidential GKE Nodes and VMs. This functionality lets you encrypt memory with processor-created, node-specific keys. Since the keys are produced in hardware during node formation and only live in the CPU, Google and other host nodes cannot access them.

Private Area

Confidentiality Space protects sensitive data while allowing businesses to aggregate and analyse it. Organisations can perform collaborative data analysis and machine learning (ML) model training with data security from all parties, including cloud service provider access. Confidential Space integration with Privacy Sandbox provides a trusted execution environment for privacy-preserving ad campaign analytics and retargeting post-cookie.

Secret Dataflow and Dataproc

Fully controlled Dataflow supports several machine learning and streaming analytics use cases at scale. Dataflow supports Compute Engine Confidential VMs, which allow inline memory encryption, for data pipelines.

Dataproc manages Spark, Hadoop, and other open source technologies and frameworks for huge data processing. Confidential Dataproc lets you secure inline memory using Compute Engine Confidential virtual machines. This improves security, especially for sensitive data.

All features

Using real-time encryption

Google Cloud users can encrypt data while using cloud services for hidden computing and AMD, Intel, and other CPU security capabilities. Customers may trust that their data will be encrypted while processing.

Elevate and move privately

Google Cloud wants to simplify GCP Confidential Computing. All your workloads—new and old—can run as Confidential VMs, making the transfer easy. You don't need to change app code to use Confidential VMs. Simple as checking one box.

Detecting sophisticated persistent assaults

GCP Confidential Computing enhances Shielded virtual machine rootkit and bootkit protection. This simplifies ensuring the integrity of your private virtual machine's operating system.

Enhanced creativity

GCP Confidential Computing could enable impossible computing scenarios. Cloud-based enterprises can now collaborate on critical data in secrecy.

Great performance

Regular N2D virtual machines function like confidential ones.

GCP Private Computing Cost

Confidential VM costs depend on machine kinds, persistent discs, and other resources.