Mikä on verkkosivusto?

Flow Computing

Etusivu
Suomi
Helsinki
Flow Computing

Verkkosivusto

Enabler of the CPU 2.0 - the new era of Supercharged CPUs Speed is the superpower of computing.

Our Parallel Processing Unit (PPU) gives CPUs a significant performance boost.

15/10/2026

Hot Chips 2026 mini series continues. 🔥
Part 5: Arm AGI CPU.

Arm presented AGI CPU, a chiplet-based server SoC designed for the agentic AI era at Hot Chips.
One thing stood out to us: scaling compute means scaling more than cores.

Arm is addressing that challenge across memory bandwidth, coherency, I/O and chiplet-based scaling.
Our takeaway: as workloads become more parallel and data-intensive, the architecture surrounding the CPU matters too.

We explore Arm’s approach, and how Flow PPU adds another complementary capability: scalable parallel acceleration alongside the CPU.

🔗 Read Part 5: https://eu1.hubs.ly/H0z1M9l0
Stay tuned for Part 6.

Thank you to Saurabh Pradhan, Deepak Goel, Arm and Hot Chips Symposium for the presentation and Q&A.

12/10/2026

Our CEO & Co-Founder, Timo Valtonen, has been in the US for the past two weeks and will be there until mid November, connecting with industry leaders and exploring opportunities for collaboration.

Timo is attending the 2026 Open Compute Project Foundation Global Summit in San Jose, October 12–15!

As AI infrastructure continues to evolve and demand for AI inference grows, CPU performance remains a critical part of the conversation.

Interested in discussing the future of CPU performance and how our Flow PPU technology can help enable the next generation of CPU performance?
Timo would be happy to connect! Reach out to arrange a meeting during the summit or while he's in the US.

08/10/2026

Hot Chips 2026 mini series continues. 🔥
Part 4: Fujitsu MONAKA.

Fujitsu presented MONAKA, its next-generation Arm-based CPU designed for AI and HPC at Hot Chips.
One message stood out to us: as AI and HPC evolve, CPU throughput still matters.

MONAKA approaches that challenge across the CPU core, memory architecture and chiplet design.

Our takeaway: CPU performance isn’t just about adding compute. It’s about designing the architecture around the workload.

We explore Fujitsu’s approach, and how Flow PPU adds a complementary layer of scalable parallel acceleration alongside the CPU.

Read Part 4: https://eu1.hubs.ly/H0yXlJR0
Stay tuned for Part 5.

Thank you to Ryohei Okazaki, Fujitsu and HOT CHIPS Symposium for the presentation and Q&A.

01/10/2026

Hot Chips 2026 mini series🔥
Part 3: NVIDIA Vera.

At Hot Chips, NVIDIA presented Vera, its 88-core CPU designed for agentic AI infrastructure.

1 thing stood out to us: the focus on performance per core, not simply core count.
Because more cores don’t automatically mean proportionally more performance.

Our takeaway: CPU performance is about how effectively compute resources work together, not just how many cores you have.

We explore NVIDIA’s approach, and how Flow PPU complements advances at the CPU-core level with scalable parallel acceleration alongside the CPU.

Read Part 3: https://eu1.hubs.ly/H0yL2HY0
Stay tuned for Part 4.

Thank you to Jonathon Evans, Polychronis Xekalakis, NVIDIA and Hot Chips Symposium for the presentation and Q&A.

29/09/2026

September was a journey for us: stages, deep tech conversations across Europe, and our next chapter in Silicon Valley.

At Nordic Chip Summit in Espoo, our CEO & Co-Founder Timo Valtonen took the stage to discuss AI factories, their efficiency and economics, and why keeping each compute engine focused on the work it handles best becomes increasingly important at scale.

At ScalPerf'26 in Bertinoro, our Co-Founder, CTO & Chief Architect Dr. Martti Forsell presented the Flow Parallel Processing Unit® architecture and selected parallel algorithms, covering methodological alternatives, optimization possibilities, and early performance and migration results.
Before heading home, Martti also made a stop in Torino for an in-person working session with Flow CPU Modeling Engineers Andrea Coluccio and Emanuele Lovino.

Back in Espoo, Timo joined Nordeep - 5th Anniversary Edition, connecting with founders, investors and industry leaders from across the Nordic deep tech ecosystem.

And now, the journey continues in the US. 🇺🇸
We joined the Nordic Innovation House - Silicon Valley strengthening our connections across the Nordic and US innovation ecosystems. Timo will spend October - mid-November in the US, meeting with partners, investors and others across the industry.

As a busy September comes to a close, we’re looking ahead to an equally exciting autumn.

Want to follow along? Stay ahead of what’s happening in computing with Flow’s Tech Insider newsletter, featuring CPU technology insights, Flow PPU updates, industry perspectives and more.
Subscribe: https://eu1.hubs.ly/H0yCbDG0

24/09/2026

Hot Chips 2026 mini series 🔥
Part 2: Intel Wildcat Lake

Different workloads need different compute.
At Hot Chips 2026, Intel’s Wildcat Lake showed how CPU, GPU and NPU can work together, each taking on the work it’s best suited for.

That same idea is central to how we think about CPU evolution at Flow: not replacing the CPU, but adding complementary capabilities alongside it.

Read the article here: https://eu1.hubs.ly/H0yxqBs0
Stay tuned for Part 3.

Thank you to Lance Hacking, Intel and Hot Chips Symposium for the presentation and Q&A.

22/09/2026

What made Flow stand out?

In Part 2 of our Parallel Perspectives: Investor Series, Jon-Kåre Hansen of Sarsia shares what differentiated Flow as he took a closer look at the technology and the problem we’re working to solve.

With a background in software development and experience working close to hardware, Jon-Kåre brings a technical perspective to what caught his attention, and why Flow stood out.

17/09/2026

Hot Chips 2026 mini series starts now. 🔥
Part 1: IBM.

At Hot Chips, IBM Distinguished Engineer Dr. Christian Zoellin presented a processor supporting both IBM Z and Arm architectures.

So what does that tell us about where CPU architecture is heading?
Our takeaway: CPU + the right architecture for the workload.

We explore IBM’s approach, and where Flow PPU fits into this increasingly heterogeneous future.
🔗 Read Part 1: https://eu1.hubs.ly/H0ym3S00
Stay tuned for Part 2.

Thank you to Dr. Christian Zoellin, IBM and HOT CHIPS Symposium for the presentation.

15/09/2026

The results are in.

Our performance page is live.

Access benchmark results, speedup data, and how Flow PPU approaches scalable parallel performance alongside the CPU.

Review the data: https://lnkd.in/eYBQPWh2

Osoite

Helsinki

Verkkosivusto

http://flow-computing.com/

Ilmoitukset

Saat tiedon ensimmäisenä: lähetämme sinulle sähköpostia, kun Flow Computing julkaisee uutisia ja tarjouksia. Sähköpostiosoitettasi ei käytetä muihin tarkoituksiin, ja voit perua tilauksen milloin tahansa.

Haluatko, että Yritys on listan kärjessä kategoriassa insinööritoimisto (Helsinki)?

Jaa

Kategoria

Tiede, teknologia ja suunnittelu