181 trillion gigabytes...

That’s the amount of data expected to be created, captured, copied and consumed globally in 2025 - nearly three times the data volume of 2020.

Data volumes have grown substantially with the rapid rise of generative AI, increasing demands on mobile data networks, and crypto mining.

In fact, by the end of 2025, the data industry is expected to consume 20% of the world’s entire energy supply, as emerging technologies continue to push the boundaries of what we need from computers.

The main performance bottleneck today is the central processing unit, the CPU, or the brain of a computer.

CPUs are responsible for processing data, and as data volumes increase, the demand for more powerful and efficient CPUs grows to handle larger datasets.

Current solutions, such as adding more ‘brains’ (cores) to the CPU or using graphics processing units (GPUs) to help out, only go so far and come with high costs and increased energy use.

So what if we could enhance CPU performance 100x?

Enter the era of next generation SuperCPUs, and our latest investment Flow Computing, who emerge from stealth with a world leading CPU design.

Flow’s innovation, called Thick-control-flow, enables CPUs to run up to 100 times faster. This is achieved by processing many tasks at the same time in an efficient way, without needing extra control systems or hardware.

By being built directly into the CPU chip, Flow’s parallel processing unit (PPU) allows CPUs to process many tasks simultaneously without requiring changes to existing software or hardware designs.

Essentially, it supercharges the CPU without the need for expensive and energy-intensive add-ons.

Key Benefits of Flow’s Technology

1. Incredible Speed: By enabling CPUs to work up to 100 times faster, Flow Computing meets the urgent need for speed in modern computing.

2. Seamless Compatibility: Flow’s technology is fully backward compatible, meaning it works with all existing software and applications without requiring any modifications.

3. Cost Efficiency: Flow’s solution reduces the need for additional cores or GPUs, cutting down on both costs and energy consumption.

If you want to learn more and go deeper on the technicalities of how this possible, you can read more from Flow here.

The below diagram may be a helpful visual. The CPU asks for data, the memory retrieves and sends the data, and the CPU uses this data to perform tasks, like opening an application. The arrows show the time taken for data to move between the CPU cores and memory; with Flow’s PPU, latency is hidden, and tasks are handled concurrently, eliminating delays and coherency issues.

It is hard to understate the impact of this technology.

Every industry in tech continues to suffer from merely incremental CPU improvements. Flow is the first major departure from this trend, offering  multiples of performance instead of some percentage points.

Despite the huge investments in AI, general purpose computing will constrain its progress, and increase its costs. Flow Computing is solving this problem by enabling the next generation of SuperCPUs to easily surpass current industry leaders such as the Apple M-series, Nvidia Grace or Google Axion.

The Strategic Fit for FOV Ventures

At FOV Ventures, we invest in the future of compute, focusing on cutting-edge fields such as spatial computing, 3D, AI, and computer vision. Recognising the essential demand for more processing power, we are excited to support Flow as they develop their next-generation Super CPUs.

Within this framework, there are also deep tech challenges that intersect the progress of this future. Our investments in Doublepoint (Touch Interface) and Kubos Semiconductors (LEDs) exemplify this. Flow and their CPU improvements address another key challenge across the whole Spatial technology stack where CPU performance has become a necessity in order to meet the increasing demand for spatial algorithms, handling real-time sensor data fusion, or AR input/outputs.

General purpose computing, unless bolstered by advancements in CPU technology, will also continue to dominate the costs and limit the capabilities of AI applications. Flow has made significant strides in this area, demonstrating that PPU-powered CPUs can greatly reduce the time spent on pre and post-processing data, crucial in advancing LLMs.

What’s more, Flow's PPU architecture will finally make locally-hosted AI a reality, which has huge implications for AI enabled VR and AI wearables (which we talk a little more about here).

This is a really exciting investment for the future of computing, with the era of Super CPUs dawning. We were joined on the €4M pre-seed round by Butterfly Ventures, Sarsia, Stephen Industries, Superhero Capital and Business Finland.

Keep up to date with Flow here, and keep an eye out for our next investment here!