Microsoft and Intel have stated that they are collaborating to develop differentiated capabilities and system-level optimizations in Azure, tailoring infrastructure to meet the demands of cloud and AI for performance, efficiency, scalability, and security.
Redmond, Washington, and Santa Clara, California, November 18 (Reuters/PR Newswire)—Microsoft (NASDAQ: MSFT) and Intel (NASDAQ: INTC) today announced an expansion of their decades-long partnership under a long-term agreement to advance several areas related to Azure infrastructure in response to growing cloud and AI demands. Microsoft is the first company to announce the use of Intel Foundry’s EMIB™ (Embedded Multi-die Interconnect Bridge) advanced packaging technology in its custom Azure chips. The agreement integrates Intel’s chips and EMIB capabilities with Microsoft’s cloud engineering, confirming multi-year supply of Intel® Xeon® processors—including next-generation server CPUs, codenamed Diamond Rapids—while expanding collaboration between the two companies on Azure’s roadmap and Intel-based Azure hardware systems.
Microsoft will also collaborate with Intel on Intel Xeon processors supporting Compute Express Link (CXL) to advance memory efficiency and system-level innovations in Azure, including Intel QuickAssist Technology (Intel QAT) and confidential computing. Both companies state that these efforts will further enhance existing memory and database performance, including:
- On Azure M series v4 virtual machines, using Intel Flat Memory Mode with CXL memory and Xeon 6 processors delivers up to 2.3x performance improvement and greater memory flexibility for demanding workloads.
- Through the first virtual machines on Azure to support Intel QuickAssist Technology (QAT) in the cloud, SQL backup speeds can be increased by up to 2.5 times, while enhancing performance and reliability.
- When confidential virtual machines migrate between hosts, Intel Trust Domain Extensions (Intel TDX) encrypt the virtual machine's memory and CPU register state.
Both parties stated that these investments aim to provide Microsoft customers with a more predictable, scalable, and resilient infrastructure foundation, enabling computing power better suited for demanding cloud and AI workloads, more efficient memory usage, and enhanced protection for sensitive workloads.
Saurabh Dighe, Corporate Vice President of Microsoft Azure Systems & Architecture, said: “Our customers need infrastructure built around their critical workloads. Our long-standing collaboration with Intel combines advanced packaging, custom computing, memory, and security capabilities to power Azure compute and drive continuous innovation across Azure. We are connecting chip and systems engineering to advance the industry and meet growing customer demand for Cloud + AI computing.”
Naga Chandrasekaran, Executive Vice President and General Manager of Intel Foundry Services, Chief Technology and Operations Officer, said: “Microsoft and Intel share a clear, aligned vision for the future of AI infrastructure—systems will become increasingly complex and require co-optimization across chips, packaging, compute, memory, and software. By leveraging our advanced packaging capabilities and EMIB to scale future generations of Azure infrastructure, Intel and Microsoft can combine the flexibility of heterogeneous integration with the manufacturing scale required in the AI era, delivering new levels of performance and economics for customers using Azure.”
About Microsoft
Microsoft is committed to driving digital transformation in the era of intelligent cloud and intelligent edge. The company’s mission is to empower every person and every organization on the planet to achieve more. Microsoft provides cloud platforms, AI capabilities, productivity tools, and business applications that help organizations innovate, enhance agility, and drive meaningful outcomes at scale.
About Intel
Intel (NASDAQ: INTC) designs and manufactures advanced semiconductors that connect and power the modern world. The company states that its engineers are creating new technologies every day to enhance and shape the future of computing, unlocking new possibilities for every customer. Learn more at intel.com.
- Standard_M76ds_2_v4: 76 vCPUs, 1,764 GB (1 TB DRAM + 1 TB CXL, Intel® Flat Memory Mode) vs. Standard_M64lds_v4: 64 vCPUs, 1,007 GB (DRAM only, two NUMA nodes) | Sandbox workload: cube-hypervisor microVM, 2 vCPUs/24 GB per sandbox, persistent Python interpreter with 12 GB dataset, 9 tools/13 steps per round, 6 seconds per step, 126 seconds per round. Tested on November 10, 2026. Results may vary.
- Intel® Xeon® Gold 8362; x86_64; ICX; 48 MiB L3 cache; 32 cores per socket; 2 sockets; Hyper-Threading enabled; 128 CPUs; Intel® Turbo Boost enabled; 2 NUMA nodes; 1024 GB installed memory (16×64 GB DDR4 2933 MT/s); 2048 KB hugepage size; Madvise Transparent Huge Pages enabled; Automatic NUMA Balancing enabled; BIOS 1.12.1; Microcode 0xd0003e7; Microsoft Windows Server 2025. Comparison system configuration: Microsoft Azure VM | CPU: Intel® Xeon® 6973P-C (Granite Rapids), 384 processors | Memory: 2.06 TB | NUMA: 6 nodes × 64 processors, ~352 GB per node | Operating System: Windows Server 2025 Datacenter | BIOS: Hyper-V UEFI v4.1 | Large Pages: 2 MB | CPU Frequency: 2.60 GHz | Microsoft SQL Server 2025. Test completed on November 2, 2026. Results may vary.



