Today’s workloads are erupting in size and complexity, driven by advances in computing technology and massive data growth. Artificial intelligence (AI) and advanced analytics are increasing the demands on traditional infrastructure. Organizations in key industries — such as financial services, healthcare, life sciences, manufacturing, and others — can gain exceptional value from AI by capturing deeper, broader insights from more of their available data.
However, these organizations can face significant roadblocks to bringing the power of data and AI into the core of their operations. They must develop a computing environment that exceeds the capabilities of traditional infrastructure to achieve greater performance, flexibility, and adaptability. High performance computing (HPC) technology is the key to their success.
AMD has introduced a new solution to accelerate the most data-intensive workloads. The AMD Alveo™ V80 compute accelerator brings massive hardware parallelism and memory bandwidth for workloads with large datasets. This flagship offering is designed for memory-bound applications that demand hardware adaptability to handle custom data types and custom data movement while scaling to manage large volumes at network line rates. The card offers the highest logic density, network throughput, and memory bandwidth in the AMD Alveo portfolioi:
The Alveo V80 offers critical advantages in performance without adding to IT complexity. Enabled by the AMD Vivado™ Design Suite, the card features familiar field-programmable gate array (FPGA) design flows to simplify bring-up. High-end FPGAs provide higher bandwidth than mainstream cards, allowing users to adapt hardware to their applications and enable custom access to memory in storage to make the most use of these FPGAs. This is a strategic option for organizations that are considering Versal™ HBM devices to eliminate processing and memory bottlenecks but also want a streamlined production flow.
To make adoption even simpler, the Alveo V80 production card simplifies integration tasks for faster time to market. AMD handles product lifecycle and inventory management, so organizations can focus on accelerating key workloads and generating value. Our management processes help ensure manufacturability and testability of packaged devices and build in quality and reliability right from the start. We develop and integrate silicon, board, and package technologies into a final product that is based on industry standard guidelines and qualified to meet customer expectations.
With these capabilities, the Alveo V80 is a powerful solution for memory-bound compute. Organizations can expect dramatic performance improvements in a range of workloads:
- HPC. Supporting custom data types with capability for hundreds of nodes, the network-attached Alveo V80 is suitable for HPC applications such as genomic sequencing, molecular dynamics, sensor processing, and astrophysics).
- AI compute. Efficient retrieval of diverse datasets, congestion algorithms, and simplified cluster architectures enable diverse AI applications, from large language models to recommendation engines.
- Database acceleration and analytics. Low-latency processing coupled with HBM for large datasets allows for energy efficiency, scalability, and faster insights for industry applications like fraud detection, public safety, healthcare analytics, and supply chain analysis.
- Networking and connectivity. Featuring built-in cryptographic engines and adaptable hardware for custom packet processing, the Alveo V80 accelerator is ideal for firewall and packet monitoring. Custom data movement also makes the card suitable for GPU clustering in data center networks.
- FinTech and blockchain. Massive parallelism supports complex applications that leverage vast amounts of data such as risk analysis, algorithmic trading/back-testing, option pricing, and Web3 blockchain cryptography.
- Computational storage. The Alveo V80 delivers highly effective storage capacity per server, making it ideal for compression in storage server nodes equipped with SSD storage drives.
Ready to see the Alveo V80 in action? Join us at SC24 on November 17th–22nd in Atlanta. Visit the AMD booth to meet with our experts and watch technology demos of our new and industry-leading solutions. Check out Hardware Zone 4 and Zone 3 for a cloud/Alveo V80 machine learning demo using a FIN network.
Let’s advance, together.
Footnotes
1 Based on AMD comparison of the specifications of the Alveo V80 vs the Alveo U55C accelerator cards in the publicly available AMD Alveo Product Selector Guide, April 2024. ALV-13.
