5 Drivers Behind the Growth of the GPU Cloud Computing Market – Visual Capitalist

5 Drivers Behind the Growth of the GPU Cloud Computing Market

Graphics Processing Units (GPUs) are specialized computer chips originally designed for use in gaming, but have recently found a second life on the cloud because of their capabilities in high performance computing.

GPUs use parallel processing to efficiently and effectively solve complex problems across a wide range of domains, much faster than the serial processing approach of CPUs.

In this visualization for HIVE Digital, we look at five drivers behind the growth of the GPU cloud computing market.

Large Language Models, like OpenAIs ChatGPT and Googles Bard, have surpassed human performance on a variety of tasks in part because of increasingly large training datasets.

In 2008, a NEC Laboratories model used a training package with 6.33 x 108 (633 million) datapoints. By 2023, an Abu-Dhabi-based lab used a training dataset with 2.63 x 1012 (2.63 trillion) datapoints, 10,000 times larger than 15 years earlier.

Autonomous vehicles collect huge amounts of data from a host of onboard sensorspotentially between 380 and 5,100 terabytes every yearand finding an effective and efficient way of processing that information has been a key focus for carmakers.

One approach, which combines onboard processing power with the resources of the cloud, prevented collisions in simulations of vehicles moving up to 45 mph, even when network lag was as high as 3 seconds. That improved to 50 mph when cloud response time was reduced to half a second.

Modern drug development typically begins by running computer simulations on millions of candidate molecules to look for potential matches to advance to expensive, lab-based testing.

Normally these operations run on scarce high-performance computing resources that can take weeks or months to complete. A recent study that used the cloud was able to complete a typical dataset in around two days, comparable to a Tier 2 supercomputer.

Streaming games on the cloud is big money, with Microsoft recently doling out an eye-watering $68.7 billion to acquire Call of Duty publisher Activision Blizzard and cement their already significant market share.

According to papers filed with the U.K. Competition and Markets Authority, Microsofts xCloud held 50-60% of the global cloud gaming market in 2022, twice that of second-place Nvidia. Googles Stadia, on the other hand, had less than 5% of the market when they wound down operations in January 2023.

When latency is critical for an application, the shorter the distance between the cloud and the user, the better.

Thats why edge computingso named because processing happens at the edge of the networkhas become critical in industries as varied as health care and autonomous vehicles, where every fraction of a second counts.

So its no surprise that the global GPU cloud market is projected to increase from $3.2 billion in 2023 to $25.5 billion by 2030, at an impressive CAGR of 34.8%.

HIVE Digitalexports renewably sourced computing power to customers all over the world, helping to meet the demands of emerging technologies like AI.

Read the original post:

5 Drivers Behind the Growth of the GPU Cloud Computing Market - Visual Capitalist