# Navigating the Costs of Cloud Networks

Cloud networks have grown from what was seen as a passing trend by some experts, into full-fledged solutions that power some of the most important parts of various industries at this point. Large companies like Google and Microsoft have been investing steadily in the growth of their own solutions.

At the same time, more tightly specialized networks have seen significant evolution as well. MetaQuotes with its MQL5 Cloud Network is perhaps one of the most notable examples, having grown to over 34,000 cores since its original inception.

## MQL5 Cloud Network

Aimed specifically at the financial market, MetaQuotes’ MQL5 Cloud Network has gone through some significant developments in the last years. The actual network capacity of the number of agents and the composition of processors involved in the cloud can be seen in real-time on the project website. The network offers a variety of useful features not found in competing solutions, such as the ability to share CPU power through Strategy Tester and earn money in return.

The network is also heavily used for optimization tasks. It’s not rare for users in this field to work with problems that have to be optimized around a specific set of inputs, and the MQL5 Cloud Network has proven very reliable for that. Major users like PWE Capital have reported significant growth as a result of integrating MetaQuotes’ solutions into their workflow. PWE Capital has seen a return rate of between 8% – 60% during times when the S&P 500 has closed at -5% and even lower.

## Alibaba Cloud

Alibaba Cloud was founded in 2009, but it took some time for the network to gain global popularity. After several major investments from the Alibaba Group, resulting from the promising initial potential of the network, it’s managed to expand to the global presence it now enjoys. Alibaba Cloud has partnered with various major organizations, including Vodafone’s German branch and even the Olympics.

The network offers two main billing models – pay-as-you-go and subscription plans. The former is suitable for smaller projects that can fit into a smaller pool of resources, while subscription billing is aimed at users who want easy expansion and access to significant computing power. Various different plans are available, including analytics and media services for more tightly specialized applications.

## Oracle

Despite Oracle itself has a long-standing presence on the market, Oracle Cloud was only made available to the general public in 2016, several years after some of the major players had moved in. Still, Oracle has invested lots of resources into the network over the years and has managed to push it to a state where it enjoys a lot of attention. Oracle Cloud is divided into multiple major service groups, including Edge Services, Business Analytics, and Content and Experience. Content and Experience in particular is a component that’s gained a lot of attention due to its simplified approach to content delivery and workflow management.

Oracle has several pricing models. The standard one, called General Purpose VM, starts at $0.0980 per hour and offers two cores and 32 GB of RAM. While this is lower than what some competitors offer by default, it’s still cheaper and more performant than some of the biggest companies on the market. Compared to AWS, for example, Oracle’s offer is not only half as expensive but also offers about 33% more performance. ### Which Networks Cost the Most to Build? The MQL5 Cloud Network seems to be the most budget-friendly for its users and at the same time a powerful solution for optimizing trading hypotheses. The network uses the computing power of users, who participate in the network to earn money when the entire infrastructure is maintained by MetaQuotes on its side. Amazon’s AWS is probably one of the most expensive cloud networks in terms of investments by its founding company. Amazon has been steadily ramping up its investment in the network. In 2011, the company was already spending around$100 – $200 million per quarter on the product. That number rose to over$20 billion per quarter over the next 10 years.

Microsoft’s Azure is on a similar level and has been quite an investment for the company. In 2015, Microsoft was already spending $10 billion every year on new data centers alone. It might be surprising in this context to find out that Google has been spending a lot less on its own data centers in comparison – it invested$9.5 billion in new data centers in 2022. While still a lot, this investment is easily overshadowed by the amounts Amazon and Microsoft have been pouring into their own solutions.

And that doesn’t even account for ongoing maintenance and support. While there is not much public data on that, some heavy users have reported estimates of around \$80 thousand per year just for AWS support alone. Scaling this up with the number of users on the platform makes it obvious that Amazon’s own costs on this front are quite significant, and similar estimates can be made for the other major players on the market.