Axelera AI, a leader in AI hardware acceleration for generative AI and edge computing, has announced Titania™, a high-performance and energy-efficient AI inference chiplet. Titania builds on Axelera AI’s Digital In-Memory Computing (D-IMC) architecture, enabling near-linear scalability from the edge to the cloud. To support this development, Axelera AI has secured €61.6 million in funding from the EuroHPC Joint Undertaking (JU) as part of the Digital Autonomy with RISC-V for Europe (DARE) Project. This follows a $68 million Series B funding round, bringing the company’s total funding to over $200 million USD in just three years.
Why Titania™ Matters
- AI Market Growth & Rising Inference Demand: The AI industry is growing at 28%+ CAGR, driven largely by AI inference workloads.
- Challenges with Current Cloud-Based Solutions: Performance, cost, efficiency, and sustainability concerns are rising, especially with new reasoning models like OpenAI-o1 and DeepSeek R1, which require higher computing power.
- Targeting Next-Gen HPC & AI Applications: Titania is designed to deliver superior throughput and efficiency for data-intensive AI applications and future zetta-scale HPC centers by 2028.
Core Innovations of Titania™
Titania is engineered to overcome cloud-to-edge inefficiencies by integrating:
1. Digital In-Memory Computing (D-IMC) Technology
- Near-Linear Scalability: Reduces power consumption while maintaining high-performance AI inference.
- Optimized for Edge and Cloud: Unlike traditional architectures, D-IMC scales efficiently without excessive power and cooling demands.
2. RISC-V Architecture with Vector Extensions
- Custom AI Processing: Enhances adaptability and efficiency for diverse workloads.
- Rapid Innovation: Enables faster response to evolving AI industry needs.
3. Chiplet-Based System-in-Package (SiP) Design
- Modular Expansion: Multiple Titania chiplets can be packaged together, ensuring high efficiency for HPC, robotics, automotive, and data centers.
- Reduced Power Overhead: Achieves high computational density without excessive energy consumption.
The Role of EuroHPC JU & DARE Project
Axelera AI’s involvement in the DARE project aligns with EuroHPC JU’s mission to develop a world-class supercomputing ecosystem in Europe.
- Strengthening European Digital Sovereignty: Titania supports Europe’s ambition to build domestic high-performance computing (HPC) accelerators.
- Expanding AI & HPC Research Capabilities: Axelera AI will grow R&D teams in the Netherlands, Italy, and Belgium to drive innovation.
- Accelerating AI Infrastructure for Europe: Titania technology will contribute to next-generation data centers that balance scalability, efficiency, and cost-effectiveness.
Industry Impact & Future Developments
Market Applications for Titania™
- High-Performance Computing (HPC): Enables zettascale data processing with enhanced AI inference performance.
- Enterprise Data Centers: Reduces AI operational costs while increasing throughput efficiency.
- Autonomous Systems (Robotics & Automotive): Enhances real-time AI decision-making with low-latency processing.
- Edge AI & Industrial Automation: Delivers high-speed AI inference for on-device machine learning.
Vision for the Future
“Since Axelera AI was founded in July 2021, we have continuously delivered AI acceleration technologies to help customers overcome AI’s biggest challenges,” said Fabrizio Del Maffeo, Co-Founder and CEO of Axelera AI. “Titania expands our vision of cost-effective, energy-efficient AI inference for edge devices and HPC data centers.”
Axelera AI’s Titania™ AI inference chiplet represents a milestone in scalable, energy-efficient AI acceleration. By integrating D-IMC and RISC-V vector technology, Titania will redefine AI infrastructure for HPC, enterprise AI, and edge computing.
As the AI industry moves towards zettascale computing, innovations like Titania ensure Europe remains at the forefront of AI hardware development and digital sovereignty.
