Flow Computing

Winner of the High-Performance Computing track at DTM100. Top 7 out of 5,000+ startups. Our CEO & Co-Founder, Timo Valtonen represented Flow at Deep Tech Momentumin the in the DTM100 pitch competition. We advanced to the finals and were recognized as the High-Performance Computing track winner. Across sectors, these challenges surfaced repeatedly: - Compute infrastructure - CPU bottlenecks - Parallel performance Compute demands are growing. Efficient performance is critical. Flow PPU solves this challenge. Thank you to the Deep Tech Momentum team, judges, investors, founders, and everyone we met in Berlin. Learn more about our approach to scalable parallel performance: https://lnkd.in/d-zBa8AZ

Something new from Flow. Over the past few months, we’ve been validating new performance results for Flow PPU across parallel benchmarks and legacy code workloads. More soon.

Modern processor architectures often struggle with scalability, synchronization overhead, and inefficient resource utilization as parallel workloads become more demanding. Our architecture is designed differently. Flow PPU uses a scalable multi-fibered multicore architecture designed to: -Adjust parallelism according to the needs of the algorithm -Improve memory access scalability and bandwidth utilization -Minimize synchronization costs -Split common and data-parallel execution into architectural entities optimized for their designated use In Flow’s hybrid architecture, the CPU handles sequential execution and control flow, while Flow PPU handles execution of individual fibers independently of their number. Learn more about the science behind Flow PPU: https://lnkd.in/dfFNThUq

🚨 DTM100 is where Europe's frontier gets funded. The next 10 just confirmed the room. Doors are about to open. Berlin is the room. They'll pitch live to top-tier investors, corporates and policymakers at DTM26 in Berlin, 𝟮𝟬-𝟮𝟭 𝗠𝗮𝘆. 🛡 Defence → Bastionne — Charles von Büchler verifies identity, integrity and intent on-device, removing reliance on external trust layers ⚡ Energy → Nyxium — Thiri Shwesin Aung slashes energy infrastructure permitting timelines by 75% with agentic AI for siting, grid and regulatory analysis 🧪 Advanced Materials → Molecular Garden — Xinyi Dong cultivates plant cells in bioreactors to produce natural botanical ingredients without the land, water or seasonality of traditional agriculture 🦾 Manufacturing & Robotics → OxidOS Automotive — Daniel Rosner shields automotive ECUs with a Rust-based OS engineered for memory safety and certification 🧠 High-Performance Computing → Flow Computing — Timo Valtonen amplifies CPU performance up to 100x with a parallel processing unit architecture → Qambria — Elise J solves logistics, finance and materials problems with quantum algorithms tuned for NISQ-era hardware → nubrain — Priyanka Jain decodes brain signals into text, images and speech with a non-invasive brain-computer interface 🤖 Enterprise AI → Systellar — Sergi Company Aguilar iterates complex hardware design as fast as software with AI-powered systems engineering → Uthereal — João S. Carvalho stitches scattered enterprise knowledge into custom AI systems with secure no-code agentic RAG 🔬 Other → ENANTIOS — Carin Lightner measures chiral molecules with Raman optical activity, replacing slow chiral-HPLC methods DTM26 starts in 𝟮 𝗱𝗮𝘆𝘀. This is the deep tech sovereignty stack of the next decade. 🎟️ Don't miss DTM26 — get your ticket: https://lnkd.in/d6h4aB6s #DeepTech #DTM100 #DTM26

Earlier this month, team Flow gathered in Cossombrato, Italy for a focused face-to-face week filled with technical collaboration, planning sessions, long discussions, and plenty of espresso along the way. As a fast-growing deep-tech team, these in-person weeks are an important opportunity to align on priorities, exchange ideas, solve problems together, and build momentum for the months ahead. Alongside the work sessions, the team also explored Alba and Asti, shared meals together, and enjoyed conversations that are hard to replicate over video calls. The future of computing is built through collaboration ...and sometimes over pasta in Piedmont.

Our CEO & Co-Founder, Timo Valtonen, will be pitching at Deep Tech Momentum in Berlin next week, as part of DTM100. Looking forward to connecting with investors, partners, and industry leaders across the European deep tech ecosystem. Interested in meeting during #DTM26? Request a meeting with Timo here:https://lnkd.in/ddg9BZ9g

Traditional processor architectures often force a compromise between low-latency sequential execution and high-throughput parallel execution. Our approach separates these responsibilities into two tightly coupled architectural components: 1. The CPU for sequential execution and legacy code 2. Flow PPU for scalable parallel execution Both operate within a unified memory system designed to simplify co-processing and reduce synchronization overhead. This hybrid architecture makes it possible to scale parallel performance independently while remaining compatible with existing and future CPU architectures. Learn more about the science behind Flow PPU: https://lnkd.in/dzq_xc8H

Next stop: Berlin. We'll be at Deep Tech Momentum. May 20-21. Our CEO & Co-Founder, Timo Valtonen, will there, meeting with investors, partners, and industry leaders across the semiconductor and AI ecosystem. Interested in connecting during DTM26? Request a meeting with Timo here: https://lnkd.in/dYABUtHj

Flow was selected as a Hello Tomorrow Deep Tech Pioneer. Now our CEO & Co-Founder, Timo Valtonen will be attending Hello Tomorrow Investor Day & Summit. Meet him on June 10-12, 2026, Amsterdam, NL Request a meeting: https://lnkd.in/duGBpzBU

Compute architecture → digital infrastructure → growth That’s the full-stack conversation happening at Chips from the Very North in Oulu. Jussi Roivainen and Dr. Martti Forsell will be there, connecting with leaders across the semiconductor value chain. Increasing computing power today requires rethinking how CPUs execute parallel workloads at the architecture level, not just adding more cores. We're bringing scalable parallel processing CPUs. More on Flow: https://flow-computing.com BusinessOulu, FiCCC, Oulun kaupunki - City of Oulu