From energy-hungry data centers to math-heavy AI models, the current state of artificial intelligence faces critical roadblocks—especially when it comes to scalability, sustainability, and efficiency. At ISCA 2025, Brain-CA Technologies is introducing a bold new direction: AI built from first principles, not layers and parameters.

We’re excited to share that three Brain-CA papers have been accepted at this year’s conference, and we’ll be presenting them across three specialized workshops:

  • Workshop on Machine Learning for Computer Architecture (MLArchSys)

  • Workshop on Designing for Energy-Efficient Computing (DCEE)

  • Workshop on Cognitive Architectures (CogArch)

Each paper explores a different facet of our architecture, but they’re united by one mission: proving that intelligence doesn’t have to be complex to be powerful.

🔍 What We’re Presenting

1. A Uniform, Tessellated Architecture for Energy-Efficient Learning and Inference
📍 MLArchSys – Saturday, June 21 | 5:20–6:00 PM JST
This paper introduces a novel AI architecture where intelligence emerges from wavefront collisions and spatial geometry. There’s no backpropagation, no centralized memory—just distributed cells working together in real time. It’s a new way to build AI that mirrors the simplicity and power of biological systems.

2. Brain-CA: A Cellular Automata-Inspired Architecture for Energy-Efficient CPU Design
📍 DCEE – Saturday, June 21 | 1:25–1:45 PM JST
This presentation focuses on how Brain-CA reimagines the CPU itself. Instead of simulating learning through costly math, Brain-CA embeds learning in the hardware through event-driven, localized interactions—delivering massive energy savings at the architecture level.

3. Brain-CA: A Dual-Speed, Low-Energy Architecture for Cognitive Learning via Cellular Automata
📍 CogArch – Sunday, June 22 | 1:45–2:10 PM JST
This paper introduces a dual-speed system inspired by biological cognition. The discovery phase operates fast, using wave interactions to uncover correlations. When a strong relationship is found, Brain-CA forms accelerated prediction paths—enabling real-time responses with minimal energy. It’s a dynamic, adaptive system built to scale across devices and environments.

✈️ No Jet Lag Required

We’re thrilled to be heading to Tokyo to present at ISCA! For everyone who can’t join us in person, we’re bringing the breakthroughs to you.

📡 Visit brain-ca.com/isca to explore related content, white papers, and resources that reflect the innovations we’re sharing in Tokyo.

Whether you’re building the future of AI or just exploring what comes next, you’ll find that Brain-CA is more than a platform—it’s a shift in how we think about intelligence.