Advancing Performance in AI Networks
Fast, efficient, and cutting-edge: The AI Networking Revolution is here with game-changing products and solutions. The AMD Pensando™ platform utilizes advanced technology to deliver cloud, compute, network, storage, and security services at cloud scale.
AMD Advancing AI – Networking Technology Solutions Group - Analyst Briefing
Soni Jiandani, SVP of AMD’s Networking Technology Solutions Group speaks with leaders from IBM Cloud, Microsoft Azure, Oracle Cloud, Tensorwave, and Cisco about groundbreaking results and the future of networking for AI workloads.
Featured Products
Back-End Networking with AMD Pensando™ Pollara 400 Adapters
Designed to meet the unique demands of AI networking, the Pollara 400 NIC enables efficient, reliable, and high-speed data transfers, empowering businesses to fully harness the potential of AI cloud computing.
Features
- UEC-ready RDMA
- Adaptive packet spray
- Path Aware Congestion Avoidance
- Selective Loss Retransmission
- Advanced monitoring and telemetry capabilities
- Optimized for AI training and inference tasks
Benefits
- Open Ethernet compatibility for vendor choice
- Future-ready with Ultra Ethernet Consortium (UEC) support
- Fast Job Completion Times
- Accelerated AI workload performance
UEC Ready RDMA Outperforms RoCEv2
6x faster message completion time1
Estimated performance results
5x faster collective completion time1
Estimated performance results
Continuing to Raise the Bar with Robust Performance
The AMD Pensando Salina DPU is fully P4 programmable and optimized for minimal latency, jitter, and power requirements. Delivering incredible power efficiency while gaining significant throughput for P4 pipelines, the Salina DPU also preserves software compatibility with previous AMD Pensando™ DPUs, making it easy for customers to adopt.
Features
- Enhanced Observability
- Cloud Networking
- Advanced Security
- Storage Acceleration
- Encryption
Benefits
- 2x performance and scale vs our previous generation DPUs6
- Optimize cloud and AI front-end infrastructure performance
- Deterministic latency, low jitter
- Fully programmable P4 data path
Explore Cisco Products Built with AMD Pensando™ DPUs
Cisco N9300 Hypershield Switching
The Cisco Nexus 9300 platforms with AMD Pensando™ DPU technology and Hypershield deliver a comprehensive data center security architecture, enhancing compliance and reducing risk for enterprises by natively infusing Hypershield in the switch embedded with DPUs.
Cisco 8102 DPU Switch
The Cisco 8102-28FH-DPU-O is a switching platform, with embedded AMD Pensando DPUs, significantly enhances network services performance for high-demand workloads in Microsoft Azure.
Learn more about the use case of Microsoft Azure’s performance gains with AMD and Cisco.
HPE Aruba CX 10000 Smart Switch with AMD Pensando™ DPUs
The Aruba CX 10000 provides 800G of distributed stateful firewall for east-west traffic, zero trust segmentation, and pervasive telemetry.
Delivering Meaningful Outcomes for Our Customers
5X SDN Performance2
Multi-Service
Future Ready
100X CPS Gains3
Cost Savings
Bare-Metal and VSI
TCO Savings4
Regulated Industries
TCO Savings of $65M5
Cisco SiliconOne and AMD Pensando™ DPU network switch
Customer Experiences with AMD Pensando™ DPUs
Resources
Footnotes
- Yanfang Le, Rong Pan, Peter Newman, Jeremias Blendin, Abdul Kabbani, Vipin Jain, Raghava Sivaramu, Francis Matus. Case Study: STrack: A Reliable Multipath Transport for AI/ML Clusters . Results not specific to the AI NIC 400, actual performance may vary.
- Vithal Shirodkar, SVP, Software Development, OCI Networking: “This jointly developed solution has been instrumental to accelerating performance for the most demanding customer workloads. Improvements have been dramatic, with a 5X performance boost for SDN networking.” – published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- Gerald De Grace, Lead, Microsoft Azure Accelerated Connections: “Our collaboration with AMD using the AMD Pensando DPU based smart appliance and full-stack AMD software in production for Accelerated Connections has resulted in 100X improvement in CPS performance over our existing solution and will extend to broader tiers within the Azure network in the future.”
- Satinder Sethi, GM, IBM Cloud Infrastructure Services: “By collaborating with AMD and utilizing their programmable DPU, and turnkey SDN and security services software, IBM Cloud has been able to accelerate to market our award-winning secure high-performance “bare-metal servers for VPC” offering for VMware, Red Hat and SAP workloads, driving significant network throughput improvements for customers, and realizing 30% TCO savings from greater workload density and preserved CPU cores.” -- published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- Jayant Tulsiani, DXC VP Cloud and Infrastructure Services: “Previously what required hundreds of virtual and physical firewalls to meet our segmentation and compliance requirements, is now being delivered natively inline on the platform, which has been transformation to the TCO model of our business, resulting in projected TCO savings of ~$65M” -- published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- PEN-012: Measurements conducted by AMD Performance Labs as of Aug 27, 2024, on the current specification for the AMD Pensando™ Salina DPU accelerator designed with AMD Pensando™ 5nm process technology, projected to result in delivering 400Gb/s line-rate estimated performance. Estimated delivered results calculated for AMD Pensando™ Elba DPU designed with AMD Pensando 7nm process technology resulted in 200Gb/s line-rate performance. Actual results based on production silicon may vary.
Footnotes
- Yanfang Le, Rong Pan, Peter Newman, Jeremias Blendin, Abdul Kabbani, Vipin Jain, Raghava Sivaramu, Francis Matus. Case Study: STrack: A Reliable Multipath Transport for AI/ML Clusters . Results not specific to the AI NIC 400, actual performance may vary.
- Vithal Shirodkar, SVP, Software Development, OCI Networking: “This jointly developed solution has been instrumental to accelerating performance for the most demanding customer workloads. Improvements have been dramatic, with a 5X performance boost for SDN networking.” – published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- Gerald De Grace, Lead, Microsoft Azure Accelerated Connections: “Our collaboration with AMD using the AMD Pensando DPU based smart appliance and full-stack AMD software in production for Accelerated Connections has resulted in 100X improvement in CPS performance over our existing solution and will extend to broader tiers within the Azure network in the future.”
- Satinder Sethi, GM, IBM Cloud Infrastructure Services: “By collaborating with AMD and utilizing their programmable DPU, and turnkey SDN and security services software, IBM Cloud has been able to accelerate to market our award-winning secure high-performance “bare-metal servers for VPC” offering for VMware, Red Hat and SAP workloads, driving significant network throughput improvements for customers, and realizing 30% TCO savings from greater workload density and preserved CPU cores.” -- published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- Jayant Tulsiani, DXC VP Cloud and Infrastructure Services: “Previously what required hundreds of virtual and physical firewalls to meet our segmentation and compliance requirements, is now being delivered natively inline on the platform, which has been transformation to the TCO model of our business, resulting in projected TCO savings of ~$65M” -- published on amd.com: https://www.amd.com/en/solutions/data-center/data-center-ai-premiere/industry-support.html
- PEN-012: Measurements conducted by AMD Performance Labs as of Aug 27, 2024, on the current specification for the AMD Pensando™ Salina DPU accelerator designed with AMD Pensando™ 5nm process technology, projected to result in delivering 400Gb/s line-rate estimated performance. Estimated delivered results calculated for AMD Pensando™ Elba DPU designed with AMD Pensando 7nm process technology resulted in 200Gb/s line-rate performance. Actual results based on production silicon may vary.