z16 Inference Processing

IBM z16 

On April 5 IBM unveiled the z16, IBM’s next-generation system with an integrated on-chip AI accelerator—delivering latency-optimized inferencing. This innovation is designed to enable clients to analyze real-time transactions, at scale — for mission-critical workloads such as credit card processing, healthcare, and financial transactions. Building on IBM’s history of leading security, IBM z16 also is specifically designed to help protect against likely future threats that might be used to crack today’s and future encryption technologies.

IBM has been telegraphing the z16 for months, at least since the Telum chip was introduced toward the end of Aug.  IBM announced the IBM Telum Processor at HotChips; Telum will be the central processor chip for the next generation IBM Z, finally introduced this week as the z16, and for LinuxONE.  With AI-based inference built in. organizations that want help in preventing fraud in real-time, or other use cases, will welcome the new z16, which promises to deliver  in-transaction inference in real time and at scale.

The 7nm Telum microprocessor is ideal for organizations seeking AI-based insights from their data without compromising response time for high volume transactional workloads. To help meet these needs, IBM Telum is designed with a new dedicated on-chip accelerator for AI inferencing, to enable real time AI embedded directly into transactional workloads, alongside improvements for performance, security, and availability.

With the z16 innovations, “companies can increase decision velocity through inferencing right where their mission-critical data lives,” said Ric Lewis, SVP, IBM Systems. “This opens up opportunities to change the game in their respective industries so they will be positioned to deliver better customer experiences and business outcomes.”

Clearly the payoff from the z16 can come fast and be significant. Financial institutions worldwide repeatedly struggle with the impacts of fraudulent activities on their revenues and consumer interactions. According to a new study from IBM and Morning Consult “2022 IBM Global Financial Fraud Impact Report,” credit card fraud is the most common type of fraud among consumers in the seven countries surveyed.

With the z16 bringing together AI inferencing via its built-in Telum Processor, and with its highly secured and reliable high-volume transaction processing, banks for the first time, can analyze for fraud during the actual transactions on a massive scale, catching the bad guys before the transactions go through.The z16 can process 300 billion inference requests per day with just one millisecond of latency. For consumers, this could mean reducing the time and energy required to handle fraudulent transactions on their credit cards. For both merchants and card issuers, this could mean a reduction in revenue losses as consumers avoid the frustration associated with false declines.

And it doesn’t stop there. Other threats, such as tax fraud and organized retail theft are emerging as challenges for both governments and businesses to control. Real-time payments and alternative payment methods like cryptocurrencies are pushing the limits of traditional fraud detection techniques. Applying the new capabilities of IBM z16 to these industries can help create an entirely new class of use cases, including:

  • Loan approval: to speed approval of business or consumer loans
  • Clearing and settlement: to determine which trades and/or transactions may have a high-risk exposure before settlement
  • Federated learning for retail: to better model risk against fraud and theft

Clearly the z16 brings a potent set of security capabilities and other goodies to the Z. Am eager to see how much and how quickly organizations can start cashing in.

Alan Radding is DancingDinosaur, a veteran information technology analyst, writer, and ghostwriter. Follow DancingDinosaur on Twitter, @mainframeblog. 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: