New chip design leverages deep learning inferencing to greatly improve fraud interception and address other financial challenges
At the annual Hot Chips conference, IBM unveiled details of the upcoming new IBM Telum Processor, designed to bring deep learning inference to enterprise workloads and help address fraud in real-time. Telum is IBM’s first processor that contains on-chip acceleration for AI inferencing while a transaction is taking place.
Developed in three years, the breakthrough on-chip hardware acceleration will help customers achieve business insights at scale across banking, finance, trading, insurance applications and customer interactions. A Telum-based system is planned for the first half of 2022.
According to a survey by Morning Consult, 90% of respondents feel the ability to build and run AI projects wherever their data resides is important. With respect to it, IBM Telum enables applications to run efficiently where the data resides, helping to overcome traditional enterprise AI approaches that need significant memory and data movement capabilities to handle inferencing. With Telum, the accelerator near mission-critical data and applications assists enterprises to conduct high volume inferencing for real-time-sensitive transactions without invoking off-platform AI solutions and impacting performance.
Clients can also build and train AI models off-platform, deploy and infer on a Telum-enabled IBM system for analysis.
Full Stack Approach to Chip Design
Telum follows IBM’s long heritage of innovative design and engineering that includes hardware and software co-creation and integration that spans silicon, system, firmware, operating systems and leading software frameworks.
The chip contains 8 processor cores with a deep super-scalar out-of-order instruction pipeline, running with more than 5GHz clock frequency and optimised for the demands of heterogenous enterprise-class workloads. The completely redesigned cache and chip-interconnection infrastructure provides 32MB cache per core and can scale to 32 Telum chips. The dual-chip module design contains 22 billion transistors and 19 miles of wire on 17 metal layers.
Innovations across banking, finance, trading, insurance
Current businesses typically apply detection techniques to catch fraud after it has occurred, which is time-consuming and compute-intensive due to the technological limitations, particularly when fraud analysis and detection is conducted far away from mission-critical transactions and data. Due to latency requirements, complex fraud detection often cannot be completed in real-time. This means that a thief can successfully purchase goods with a stolen credit card before the retailer becomes aware of the fraud that has taken place.
Telum can help clients to shift from fraud detection to fraud prevention, thus evolving from catching many cases of fraud to preventing large-scale fraud without impacting service level agreements (SLAs), before the transaction is completed.
The new chip features an innovative centralised design, which allows clients to leverage the full power of the AI processor for AI-specific workloads, making it ideal for financial services workloads like fraud detection, loan processing, clearing and settlement of trades, anti-money laundering and risk analysis. With these innovations, clients can enhance existing fraud detection systems and accelerate credit approval processes, improve customer service and profitability, identify which trades or transactions may fail, and propose solutions to create a more efficient settlement process.
Telum is the first IBM chip with technology created by the IBM Research AI Hardware Center. Additionally, Samsung is IBM’s technology development partner for the Telum processor, developed in 7nm EUV technology node.
“The financial services sector deals with a huge volume of transactions and mission-critical workloads on a day-to-day basis. The newly launched IBM Telum, Z Series chip is designed to cater to the volume while enabling applications to run efficiently where their data resides. The chip will also allow businesses to capture insights and fight fraud in real-time,” said Ravi Jain, Director – Server Sales, India South Asia.