AI in
Fintech
Download the Solution Brief
Over 70 billion real-time payment transactions were processed globally in 2020, a 41 percent increase compared to the previous year.1
Applications of AI in Finance
Challenges
Why is Gaudi a good fit for Fintech Use Cases?
Deep neural network-based models for financial data include a large amount of processing that can be paralleled and thus accelerated. Finance use cases benefit specifically from accelerators that can handle data parallelism when the training dataset is huge and model parallelism when the models are large. There is an increasing need for frequent retraining and updating of models to reach generalizability across different location conditions.
The two primary considerations that come into play when employing AI processing are time- and cost-to-train. Habana’s Gaudi Training Processors are expressly designed—in both hardware and software—to deliver high-efficiency cost- and time-to-train, making AI training more accessible to more organizations and for more applications.
First-generation Gaudi delivers up to 40% better price-performance than comparable GPU-based solutions—for both the EC2 DL1 instance and on-premise systems. Training with Gaudi clusters is available both in the cloud with AWS EC2 DL1 instances consisting of 8 Gaudis and on-premises with the Supermicro X12 Gaudi Training Server also consisting of 8 Gaudis. In addition, Gaudi 2, which launched in May 2022, offers substantial performance advances that enable significantly faster training of models while preserving cost-efficiency. Gaudi 2 systems will be available from Supermicro in 2H 2022 for on-premises implementation.
Customer
Testimonials
In times of increased volatility and uncertainty, financial managers need to understand the impact of world events on market conditions in real-time. Unlike exchange-traded instruments, where values can be observed each time the instrument trades, derivative values need to be computed using complex financial models. AI solutions can provide real-time valuations and volatility analysis for financial instruments like derivatives.
Riskfuel, a FinTech startup, provides real-time valuations and risk sensitivities throughout the trading day. They deployed complex AI models with ResNet-inspired architecture and trained on synthetic datasets (derived from complex slow solvers) to provide fast, timely information on valuations throughout the trading day. Training time can be adversely impacted when models are encoding symmetries in neural network architecture while also being highly performant. In this case, our customer benefitted from Gaudi-based DL1 instances to overcome this problem, enabling them to build high-quality models with lower training costs.
[1] “Global Real-Time Payments Transactions Surge by 41 Percent”, ACI Worldwide Report, available at https://investor.aciworldwide.com/news-releases/news-release-details/global-real-time-payments-transactions-surge-41-percent-2020
[2] The impact of artificial intelligence in the banking sector & how AI is being used”, Business Insider, available at Business Insider
