Microsoft Corp. today added two sets of virtual machines to its Azure public cloud that are designed to facilitate confidential computing, an increasingly popular approach to improving the security of enterprise applications.

Confidential computing is a cybersecurity method whereby a server’s central processing unit turns a portion of the onboard memory into an isolated data environment. The data in this isolated  environment can only be accessed by the software that uses it. Neither the cloud operator that owns the server nor the operating system running on the machine have the ability to read or modify the data.

Microsoft’s first new set of virtual machines for confidential computing use cases is known as the DCdsv3 series. It’s based on Intel Corp.’s latest, third-generation Xeon processors for servers. The other virtual machine series is known as the DCasv5/ECasv5 line and uses Advanced Micro Devices Inc. silicon.

The Intel-powered DCdsv3 virtual machines provide confidential computing features using a technology called SGX that ships with the chipmaker’s server CPUs. The  third-generation Xeon processors feature an improved version of SGX. One of the biggest upgrades compared with earlier processors is an enhanced implementation of the Enclave Page Cache, a component that helps store the code and data an application uses as part of a confidential computing workflow. 

“The size of the Enclave Page Cache (EPC) for Intel SGX has increased 1500x to support much larger workloads” Microsoft engineer Michael McReynolds explained in a blog post. Intel’s third-generation Xeon processors, also known as the Ice Lake series, feature numerous other improvements as well. The chips  are made using the company’s latest 10-nanometer manufacturing process and offer more performance than previous-generation Xeon chips.

Also new is a feature that enables organizations to encrypt each of their virtual machines’ memory with a separate encryption key. “With this generation, we’re enabling customers to encrypt their VM with separate and unique keys using Intel Total Memory Encryption – Multi Key, which enables always-on encryption and provides protection against tenants on the same node,” McReynolds detailed. “Leveraging both, customers using Intel SGX get confidential computing in application enclaves and additional protection for all software in their VMs.”

The other new set of virtual machines that Microsoft debuted today for confidential computing use cases is based on AMD’s Epyc 7003 processor series. AMD introduced the chips earlier this year. 

The company based all the chips in the Epyc 7003 lineup on its latest seven-nanometer Zen 3 core design, which on average provides 19% more instructions per cycle than its predecessor. The chips feature as many as 64 cores that can run 128 processes between them. Microsoft’s new virtual machines create a confidential computing environment using a technology called SEV-SNP that AMD provides as a built-in feature of the Epyc 7003 series. 

“These AMD EPYC-CPU powered Azure VMs are fully encrypted at runtime, fulfilling the promise of confidential computing by protecting your data even when it is in use,” explained Microsoft Azure Chief Technology Officer Mark Russinovich. “The encryption keys used for VM encryption are generated, and safeguarded, by a dedicated secure processor on the EPYC CPU. This helps ensure that no one, even cloud administrators — and by extension the workloads, apps, or data in the VMs — have access to these encryption keys.

Russinovich said Microsoft will make all its confidential virtual machines available in Azure Kubernetes Service as an infrastructure option. Azure Kubernetes Service is Microsoft’s managed implementation of the Kubernetes framework for managing software container applications. 

Photo: Microsoft