To better understand the AI phenomenon and the trends driving change throughout the ecosystem, PNY Europe and 33INSIGHTS reviewed over 2,000 companies and startups.

The growth in the number of European AI start-ups has been exponential compared to many other ecosystems. In order to stay in the game and to fuel their growth, AI startups have adapted their operations to leverage rapid innovative breakthroughs, such as improvements in machine vision, which now rivals that of humans.

Unrivalled leadership

The peak of foundation of AI start-ups was reached in 2015 in the UK, the leading European country, one year ahead of continental Europe. Over the last decade, $5.5bn in venture capital (VC) investment has been secured for Europe’s AI industry. The UK is the continent’s unchallenged leader with almost 41% of the total investment ($1.9bn). According to 33INSIGHTS data, more than 260 startups raised significant funding in Europe. Staggeringly, seven start-ups raised more than $100m.

Of the $5.5bn invested over the last 18 years, the split across the value chain is $473m (8%) in hardware, $860m (16%) in foundation software, and $4,141m (76%) in applications. It is striking to see how much funding the hardware segment captured considering its relatively low number of startups (1% of the total). The UK enjoys a considerable lead in two segments – foundation software and applications. In the hardware segment, France emerges as a big player ahead of the UK, Sweden and Norway.

Disruptive force

The disruption of AI through GPU computing is in full force: the height of the creation and fundraising stages are probably behind us for both hardware and foundation software. For these segments, the next steps may depend on fast access to storage during GPU processing. This pace will likely continue in the coming years, due to continued improvements in GPU computing and the great potential in untapped industries: agricultural technology, real estate and manufacturing among others. According to NVIDIA, GPU computing capacity will have grown at a rate of 1.5 times a year by 2025; 1,000 times faster than CPU computing capacity used for parallel computing tasks.

With increasing processing speeds, the time it takes to access massive quantities of data has become the next bottleneck to overcome. Policymakers and key solutions providers like NVIDIA could foster open innovation frameworks and devise consortiums to better tackle AI challenges. For example, NVIDIA and PNY products and services have been critical to supporting innovation in industries such as health diagnosis, autonomous vehicles, robotics and many more. The progress expected of AI capabilities in the coming years will make fertile ground for new applications.