Edge computing takes a 'quantum leap'! AMD Launches New Family of FPGAs to Support Anti-Quantum Cryptography

On March 6, AMD announced the AMD Spartan™ UltraScale+™ FPGA family, the latest addition to AMD's broad portfolio of cost-optimized FPGAs and adaptive SoCs.

It's also just over a month since AMD introduced the industry's first VESA DisplayPort 2.1-compliant FPGA and adaptive SoC implementation on January 22nd.

Spartan UltraScale+ devices deliver cost and power-efficient performance for a wide range of edge I/O-intensive applications, provide the industry's highest I/O-to-logic cell ratio in FPGAs using 28nm and lower process technologies, reduce total power consumption by 30 percent compared to the previous generation, and include the most robust security features in AMD's cost-optimized portfolio.

Supports anti-quantum cryptography for enhanced security features

Optimized for the edge, Spartan UltraScale+ FPGAs offer high I/O counts and flexible interfaces that enable FPGAs to seamlessly integrate and efficiently connect with multiple devices or systems in response to the explosive growth of sensors and connected devices.

This family of FPGAs offering the industry's highest I/O to logic cell ratio, built on 28nm process technology and below, features up to 572 I/Os and up to 3.3V voltage support to enable any-to-any connectivity applications for edge sensing and control. The proven 16nm architecture and support for a wide range of packages, from as small as 10x10mm, provides high I/O density in an ultra-compact footprint. The broad AMD FPGA portfolio also provides scalability, starting with cost-optimized FPGAs all the way to mid-range and high-end products.

In addition, Spartan UltraScale+ FPGAs offer the most advanced security features in AMD's cost-optimized FPGA portfolio.

1) Protect IP: Anti-quantum cryptography is supported through NIST-approved algorithms, providing state-of-the-art IP protection against evolving cyber-attacks and threats. The physical non-clonable feature provides a unique fingerprint for each device for added security.

2) Tamper Resistance: PPK/SPK key support helps manage outdated or compromised security keys, while differential power analysis helps prevent side-channel attacks. These devices include permanent tamper penalties to further prevent misuse.

3) Maximize Uptime: Enhanced single-event rollover performance facilitates fast, secure provisioning and improves customer reliability.

The AMD Vivado™ Design Suite and Vitis™ Unified Software Platform support the entire AMD FPGA and Adaptive SoC portfolio, enabling hardware and software designers to take advantage of the productivity benefits of these tools and include IP from design to verification with a single designer cockpit.

The AMD Spartan UltraScale+ FPGA family sampling and evaluation suite is expected to be available in the first half of 2025. Documentation is available now and tool support begins in Q4 2024 from the AMD Vivado Design Suite.

"For more than 25 years, the Spartan FPGA family has helped to enable some of humanity's greatest achievements, from life-saving automated defibrillators to CERN particle gas pedals that have pushed the boundaries of human knowledge," said Kirk Saban, corporate vice president of AMD's Adaptive and Embedded Computing Business Unit. "The Spartan UltraScale+ family is based on proven 16nm technology, and the enhanced security and functionality, common design tools, and long product lifecycle further strengthens our market-leading FPGA portfolio and underscores our commitment to delivering cost-optimized products to our customers."

AMD and Advantech are longtime partners, and the latter has been driving the evolution of edge computing. In particular, leveraging AMD's leading technological capabilities in processors will help Advantech continue to enhance its high-performance edge computing solutions and develop a robust product roadmap that meets the future needs of customers and the market.

"By combining AMD's CPU processor technology, Xilinx's FPGA expertise, and Advantech's 40 years of experience in the embedded industry, both companies will further optimize the efficiency of advanced computing applications." Advantech Embedded IoT Platform Business Group Senior Manager Shih-Chiang Liao said.

Advantech, a global leader in Industrial IoT, continues to enhance its edge computing solutions. Advantech has teamed up with partner AMD to provide advanced processing cores to drive the industry towards high-performance computing, helping to empower diverse embedded and AIoT applications and accelerate the edge computing revolution.

In addition, Advantech expects AI-powered PCs to play a key role in the next three years. The company plans to launch more AMD-based hardware AI-accelerated solutions to continue expanding the diverse AI and high-performance computing market.

......Focus on future technology trends

Meanwhile, Gartner, Inc. predicts that by 2025, approximately 75% of enterprise data will be generated and processed at the edge outside of the data center/cloud, further emphasizing the need to implement high-performance computing at the edge.

S. Liao expects the demand for edge computing performance to increase significantly over the next 5 to 10 years in order to enable advanced high-end applications. This will drive hardware requirements in line with higher computational performance, more efficient multitasking capabilities, higher data transfer bandwidth, and the need for AI-assisted real-time decision making.

Despite the growing demand for high-performance edge computing, customers continue to face limitations with existing products that are unable to meet the computing needs of advanced applications.

To address this issue, Advantech has introduced the concept of "High Performance Edge Computing (HPEC)" based on edge computing.

Advantech's HPEC series further features a high-density processor as the core, support for a variety of expansion interfaces and I/O, high-bandwidth data transfers, and the ability to support or integrate high-end GPUs and AI acceleration cards, which can meet the growing demand for high-performance computing at the edge, with dimensions that are 30% to 50% smaller than those of traditional server products.

The series includes industrial motherboards AIMB-592 and AIMB-523 with AMD high-performance processors; system-on-module (SOM) SOM-E780; and the high-performance system AIR-520. In addition, Advantech provides software solutions, including the Edge AI SDK, to accelerate application integration efficiency.

In response to increasing edge workloads, "edge computing" continues to play a key role in the embedded Internet of Things (IoT) or Artificial Intelligence Internet of Things (AIoT) space. It further combines advances in AI and data computing and aims to bring higher performance computing (HPC) to the edge.

Edge computing technology brings several benefits. Edge computing eliminates the need to upload data to the cloud for processing and analysis, reducing the distance between data generation, processing, analysis and storage. Proximity to the data source minimizes data flow and latency, reduces security risks, and lowers transmission and management costs, resulting in more efficient data processing and operations.

On the other hand, advances in computing technology and standardized hardware and software architectures have made systems more scalable, increasing the flexibility for large-scale deployments across a variety of application areas while lowering the barriers to adopting new capabilities. Today, with the support of artificial intelligence, edge computing will be able to automatically adjust in real-time to demand and improve accuracy.

For more than 50 years, AMD has been driving innovation in high-performance computing, graphics and visualization technologies. In response to the hottest topic in the entire computer industry, AI PCs, AMD also demonstrated its leadership in this area of research and development.

On March 21st, AMD hosted the AI PC Innovation Summit in the presence of many partners, media and opinion leaders to showcase its momentum in the AI PC ecosystem in China. As the first company to introduce a dedicated AI engine, or NPU, into the x86 ecosystem, AMD has now shipped millions of systems and offers a diverse range of PC designs in the market, with the AMD RuiLong 8040 series of processors bringing leading-edge computing and AI experiences to the Chinese market.