Enterprise application software leader SAP is enhancing its partnership with Amazon Web Services (AWS) through innovative AI integrations. This collaboration starts with the AI Core platform, where SAP will incorporate support for Amazon Bedrock, enabling companies—particularly in regulated industries—to explore generative AI capabilities.
Additionally, SAP is committing to utilize AWS chips to elevate its infrastructure. The company plans to upgrade its Graviton chips and employ AWS Trainium and Inferentia to enhance its Business AI offerings.
“AWS was the first cloud provider certified to support the SAP portfolio, and today, thousands of enterprises utilize SAP solutions on AWS for optimal performance of their mission-critical applications,” stated Matt Garman, AWS’ incoming CEO. “Together, AWS and SAP are streamlining the integration of generative AI into core business data, enabling organizations to become more efficient, responsive, and sustainable.”
Expanding Generative AI Access for Enterprises
AI Core, part of SAP’s Business Technology Platform, facilitates standardized, scalable, and hyperscaler-agnostic management of AI assets. With the integration of Amazon Bedrock, enterprises can now leverage a range of large language models (LLMs), including Amazon Titan, Anthropic’s Claude 3 Opus, Cohere’s Command R models, Mistral 7B, Mixtral 8x7B, Meta’s Llama, stable diffusion options, and AI21 Labs.
“Our flexible and open architecture allows us to partner with top technologies to meet specific needs,” said Dr. Philipp Herzig, SAP’s Chief AI Officer. “Amazon Bedrock offers some of the best models that our customers demand, providing a diverse selection for both SAP developers and customers to create tailored experiences.”
This collaboration enables enterprise developers to optimize key business processes built on SAP. For instance, Amazon Bedrock can enhance compliance within SAP’s S/4HANA Central Finance by streamlining audits through automated bank statement verification.
Matt Wood, AWS’ Vice President of AI Products, noted, “I haven’t witnessed this level of engagement and innovation from customers around any technology since the early AWS days. Our enterprise clients are rigorously adopting generative AI, moving from advanced prototypes to production processes close to their core operations.”
While AWS solidifies its position in AI development, SAP’s integration is noteworthy due to its user base in regulated sectors. Wood mentions that organizations like the New York Stock Exchange, NASDAQ, and major pharmaceutical firms are rapidly adopting generative AI, driven by stringent compliance requirements. “These organizations possess high data quality and security, making their adoption of generative AI significantly easier than for other enterprises.”
He adds that regulated firms, often rich in textual data, can harness AI effectively for numerous internal applications, giving them a chance to accelerate their digital transformation compared to industries like media and hospitality.
SAP customers can access Amazon Bedrock via the Business Technology Platform (BTP) or the managed cloud offering, RISE with SAP.
Enhancing AI Capabilities with AWS Chips
In addition to its Amazon Bedrock collaboration, SAP is expanding its deployment of AWS chips for AI infrastructure. Specifically, HANA Cloud, which supports intelligent data applications, will transition from Graviton3 to Graviton4 chips. Furthermore, SAP plans to utilize AWS Trainium and Inferentia chips for training and deploying future AI products.
According to AWS, “Using Trainium and Inferentia2 instances in a proof of concept, SAP engineers were able to train and fine-tune generative AI LLMs in just two days, in contrast to 23 days with comparable Amazon EC2 instances. By leveraging Trainium’s specialized architecture for efficient ML model training, SAP can accelerate development without compromising accuracy.”
Herzig highlighted that the focus on purpose-built chips aims to reduce energy consumption, lower costs, and enhance performance. He considers Graviton4 chips a significant advancement for database and enterprise workloads, supporting SAP’s BTP, Datasphere, Analytics Cloud, and Cloud ALM solutions.
Ranking
