Case Analysis of Iluvatar CoreX General-Purpose GPU Applications in Finance and Healthcare

As a leading enterprise in China’s general-purpose GPU sector, Iluvatar CoreX has achieved large-scale deployment of its products in multiple key industries such as financial services and healthcare. Based on public information [1][2][3], the following elaborates on its specific application cases in the finance and healthcare sectors in detail.

Taiping FinTech is a technology innovation platform under China Taiping Insurance Group, facing growing demands for intelligent identification and risk management in the financial risk control field. Traditional CPU computing architectures have obvious bottlenecks in processing large-scale concurrent requests, making it difficult to meet the dual requirements of real-time risk control for response speed and accuracy [1].

Iluvatar CoreX provided Taiping FinTech with core computing power support based on the TianGai series general-purpose GPUs, and built a dedicated AI inference infrastructure. This solution is mainly applied in the following two core scenarios:

• OCR Optical Character Recognition
  : Used for intelligent recognition and information extraction of financial documents, supporting automated processing of business processes such as insurance claims and account opening [1][2]
• Face Recognition Authentication
  : Applied in scenarios such as remote identity verification and online business processing, providing high-precision, low-latency face matching services [1][2]

This solution successfully helped Taiping FinTech complete the localized deployment of the DeepSeek R1 large model, achieving full-stack computing power coverage from the cloud to the edge. The relevant application case was successfully selected into the list of typical application cases released by the Ministry of Industry and Information Technology, fully verifying the reliability and maturity of Iluvatar CoreX products in key financial scenarios [1][2][3].

Financial quantitative trading has extremely stringent requirements for the latency and accuracy of computing systems. High-frequency trading strategies need to complete market data processing, strategy calculation, and order execution at the millisecond or even microsecond level, while also ensuring high-precision floating-point operations to guarantee the accuracy of strategy execution [3].

Iluvatar CoreX’s general-purpose GPUs have been adopted by multiple leading securities firms to accelerate deep reinforcement learning and time series prediction models:

• DQN (Deep Q-Network) Model Acceleration
  : Supports reinforcement learning-driven adaptive trading strategies, enabling dynamic responses to changes in market conditions [3]
• LSTM (Long Short-Term Memory) Model Acceleration
  : Used for time series prediction in financial markets, analyzing price trends and fluctuation patterns [3]

According to technical verification data disclosed in the prospectus, this solution effectively shortens the model development cycle and significantly improves model training and inference efficiency. Measured data shows that Iluvatar CoreX GPUs exhibit excellent stability when processing high-concurrency, high-precision floating-point operation tasks, and can meet the stringent requirements for system reliability in high-frequency trading scenarios [3].

As a regional medical center, Quzhou People’s Hospital generates massive amounts of medical record data and medical image materials in daily diagnosis and treatment. Traditional manual medical record writing and disease analysis methods are inefficient, and it is difficult to achieve comprehensive intelligent analysis of multi-source data [1][2].

Based on the ZhiKai 100 inference GPU, Iluvatar CoreX collaborated with the hospital to deploy the DeepSeek Medical All-in-One Machine. This solution has achieved deep integration with the hospital’s core information systems:

• System Integration
  : Connects to the Hospital Information System (HIS) and Picture Archiving and Communication System (PACS), enabling unified access to clinical data and image data [1][2]
• Localized Deployment
  : The 70B-parameter large model is deployed locally within the hospital, ensuring the security and privacy protection of medical data [1]
• Intelligent Computing Power Support
  : The ZhiKai 100 provides high-throughput inference computing power to support the real-time response requirements of large language models [1]

This system has been efficiently applied in the following scenarios:

• Automated Medical Record Generation
  : Automatically generates standardized medical record documents based on patient diagnosis and treatment information, greatly reducing the paperwork burden on doctors [1][2]
• Intelligent Disease Analysis
  : Assists doctors in clinical decision-making, providing disease analysis and diagnosis and treatment recommendations based on medical knowledge bases [1][2]

Medical image analysis (such as PET/MR image correction) and AI-assisted diagnosis need to process massive amounts of unstructured data, which puts extremely high requirements on the parallel processing capability of computing systems. Traditional CPU architectures are difficult to meet the requirements of real-time image analysis and high-precision diagnosis [3].

Iluvatar CoreX provides a complete integrated software and hardware solution for medical image scenarios:

• General-Purpose GPU Hardware Platform
  : Provides powerful parallel computing capability based on the TianGai/ZhiKai series GPUs [3]
• Medical Scenario-Optimized Software Library
  : Optimizes algorithms for the characteristics of medical images, improving image processing efficiency and diagnostic accuracy [3]

This solution has helped hospitals realize full-process intelligence from image analysis to electronic medical record generation, significantly improving the efficiency and quality of medical diagnosis. Through GPU-accelerated deep learning models, complex image data processing and lesion recognition can be completed in a short time, assisting doctors in making accurate diagnoses [3].

As of June 30, 2025, Iluvatar CoreX has cumulatively delivered over 52,000 general-purpose GPU products to more than 290 customers, completing over 900 actual deployments in key industries such as financial services, healthcare, and transportation [1][2]. The company’s revenue increased from RMB 189.4 million in 2022 to RMB 539.5 million in 2024, with a compound annual growth rate of 68.8%, and achieved revenue of RMB 324.3 million in the first half of 2025, a year-on-year increase of 64.2% [2].

The widespread application of Iluvatar CoreX products in the finance and healthcare sectors stems from its following technical advantages:

Iluvatar CoreX’s general-purpose GPU products have formed a mature solution system in the finance and healthcare sectors. In the finance sector, its collaboration case with Taiping FinTech was successfully selected as a typical application by the Ministry of Industry and Information Technology, verifying its technical strength in intelligent risk control and AI large model deployment; in the healthcare sector, the DeepSeek Medical All-in-One Machine project at Quzhou People’s Hospital has realized the intelligent upgrade of medical record generation and disease analysis. These application cases fully demonstrate that domestic general-purpose GPUs have the capability to replace imported products in core scenarios of key industries, providing strong support for China’s computing power independent controllability strategy [1][2][3].

[1] Xinhua Daily Network - Iluvatar CoreX: Leading the Domestic General-Purpose GPU Track, Full Independent R&D Technology Deployed Across All Scenarios (https://www.xhby.net/content/s694248d3e4b096256d6babf4.html)

[2] Sina Finance - Iluvatar CoreX: Decoding the Breakthrough and Rise of China’s “Low-Key King” in General-Purpose GPUs (http://k.sina.com.cn/article_5953190046_162d6789e06702j4t4.html)

[3] OFweek AI Network - Iluvatar CoreX: Being a “Long-Termist” in China’s Computing Power (https://mp.ofweek.com/ai/a056714260617)