The market for cloud computing services is growing.
Between 2017 to 2020, its revenue is set to nearly double to more than US$300 billion. This is despite the fact that cloud computing itself is a relatively new concept—the term was first thought to have been used by Google CEO Eric Schmidt in 2006. So, what is driving this rapid growth?
There is already some adoption of cloud computing: many of us already use cloud computing (mainly for file storage), and enterprises and governments are also using the service. But AI is expected to be the biggest growth driver for cloud computing in the next few years.
AI relies on crunching massive quantities of data, which consumes a lot of computing power. So, the race is on among tech firms to see which of them can best meet this need.
The current frontrunners
Amazon Web Services (AWS) is currently the most dominant force in the cloud computing market, holding a 33 percent market share. The company recently signed new deals with existing customers SAP and Symantec, reportedly worth a combined US$1 billion.
Competition for AWS is heating up, though. Microsoft is its closest competitor, and recent reports indicate that the company is aiming to secure top-level federal security authorizations by 2019.
Huawei is also intent on entering the race and seems to have the means at its disposal to make some significant early headway. The Chinese company announced that for the first time, it will be selling servers powered by its own chips, which it claims are twice as powerful as the nearest competitor. This may indicate that Huawei is serious about securing its position in the cloud computing market.
However, for AI developers, renting computing power is pricey. And while there’s a huge demand, supply isn’t increasing in the same way. This means prices are still going to get higher.
Building their own hardware
While the current big players are using their own existing server infrastructures to service their big clients, others in the race are thinking differently in an attempt to get ahead.
Alibaba is already a significant player in the Chinese cloud computing market. However, the company has now launched its Alibaba Cloud quantum computer, capable of processing 11 quantum bits (qubits). This follows the launch of IBM’s 20 qubit quantum computer at the end of 2017.
Startups are also coming into the mix. Rigetti is a US-based company that has received funding from Andreessen Horowitz and Vy Capital. The startup is aiming to build “the world’s most powerful computers,” launching a quantum-first integrated cloud platform.
While the full benefits of quantum computing are still subject to research, the power of using it could be a considerable differentiation compared to using existing high-power processors.
But building bigger and better machines is an expensive undertaking, and it may not be the only way to meet the need for cloud computing services.
Using untapped computing power
There is a vast amount of untapped computing power that already exists in the world today. Game servers, mining farms, and even the spare capacity of big cloud computing providers like AWS could all be put to use.
Distributed computing is at the core of blockchain technology, which means blockchain has the potential to harness unused computing power.
Currently, there are many mining farms around the world that were built for a single purpose: to mine cryptocurrencies. While this can be profitable, it’s essentially wasting energy, as it’s all used to secure a single network. Instead, this same energy (while still remaining profitable) could be used for AI development.
Tatau, for instance, is a blockchain startup aiming to create a distributed computing platform focused on providing AI developers with the compute power they need to bring their products to market.
However, there are two major problems with distributed computing:
- UI/UX: Blockchain is in its early stages, and I think it’ll take time for blockchain startups to provide the same high-level solutions like big tech firms.
- Market skepticism: People are also familiar with big tech companies like AWS and Alibaba and have learned to trust them. Why would they now switch to an unfamiliar startup?
The implications of this increasing competition in cloud computing are big not just for most of the AI industry, but also for IT companies who are moving their data to the cloud.
Currently, supply is much lower than demand, which causes price inflation. But increasing competition will lead to more cloud computing innovation, allowing more developers and companies to utilize cloud services by decreasing prices.
The price of computing power (either in renting cloud services or purchasing specific hardware) is perhaps the biggest reason why AI innovators are holding back.
Between the existing big players, those edging in with quantum computing, and the new entrants from the world of blockchain, the race will undoubtedly be an interesting one to watch.