In AI computing, two AI development platforms, Microsoft Azure and Amazon’s AWS, stand out for their performance and market dominance. But, another relatively new kid on the block, Google Cloud Platform (8.9 % market share), doesn’t usually get as much press as Azure and AWS, especially in everyday corporate IT.
However, if you switch from office-bound networking to cloud solutions optimized for remote employees, Google Cloud is as capable as any of its rivals. This GCP review will break down its features, performance, and other parameters to see whether GCP lives up to its hype.
Google Cloud AI Platform At A Glance
| Details | |
|---|---|
| What it is | A managed AI and cloud computing service from Google |
| Pricing model | Pay as you go, with some free options |
| Free trial | $300 in credits for new users |
| Global reach | 35 regions, 105 zones and 174 network edges |
| Uptime guarantee | 99.99% minimum for Compute Engine, with service credits if it falls short |
| Most popular with | Large enterprises |
What Is Google Cloud AI Platform?
GCP is a managed service offering customers access to AI computing resources stored in Google data centers worldwide for free or on a pay-as-you-go basis.
GCP is ranked 6th among the top AI development platforms. Most users give it a rating of 7.7 out of 10, and it’s most popular among the large enterprise segment.
The users prefer Google Cloud AI due to the following benefits:
Easy setup
Once upon a time, you needed to be a computer whiz kid to use Google Cloud, but nowadays, the setup has shifted from the complicated command line to a new graphical user interface (GUI). Thanks to this GUI, you can use a browser to access screens and menus, and once you complete the server configuration, you can also use its GUI.
The cherry on the cake is that if you are a first-time GCP user, you are automatically credited with $300 worth of trial credits to test the platform and see if the platform and features work for you.
High performance
GCP AI has scaled up its footprint globally, now covering 35 regions, 105 zones, and 174 network edges, serving customers in over 200 countries. With Google announcing even more cloud regions at its Next ‘23 conference, it’s safe to say that there is a data center wherever you are.
In a previous statement, Google said that each data center is continuously being upgraded and retrofitted with the latest Google Axion Processors. GCP’s internal network now achieves speeds of 16 Gbps, providing a secure connection between regions and offering capabilities that are developed enough to compete with IBM Cloud, Azure, and AWS.
AI Platform And Vertex AI
Google has since brought its AI Platform tools together under the Vertex AI name, so newer Google documentation and pricing pages refer to Vertex AI.
Pricing
Pricing is best assessed on a case-by-case basis, and Google provides a straightforward approach to this with a pricing calculator and accessible price sheets. Just like Azure and AWS, all prices are in U.S. dollars – which makes comparisons easy. However, it exposes non-US users to fluctuations in exchange rates.
As for contracts, Google’s Compute Engine’s SLA guarantees a minimum uptime of 99.99%. If you experience a monthly uptime percentage ranging from 99.95 to 99.99, you get a 10 % credit refund; 95 to 99% attracts 25%, and below 95% gets a 50% refund.
Pros
- GCP AI is a powerful AI computing platform that features excellent support for Linux and Windows
- Its broad reach of global private networks offers unparalleled speed and seamless connectivity
- Features a suite of data analysis and AI apps
Cons
- It is significantly more expensive than AWS
- You might experience difficulty integrating it into non-Google products
Who Google Cloud AI Platform Suits
It suits enterprises that need global infrastructure and AI tools in one place, and teams that already work with Google products. Companies built around Microsoft products may find Azure a better fit, as the conclusion below explains.
Google Cloud AI Platform FAQ
Is Google Cloud AI Platform free?
New users get $300 in trial credits, and some services have free options, but most production use is billed on a pay-as-you-go basis.
How is Google Cloud AI Platform priced?
Pricing depends on the services and resources you use. Google provides a pricing calculator and price sheets, and all prices are listed in U.S. dollars.
Is Google Cloud more expensive than AWS?
In our assessment it can be significantly more expensive than AWS, so it is worth running your workload through both pricing calculators before you commit.
What happened to Google Cloud AI Platform?
Google now offers its AI Platform tools as part of Vertex AI.
Conclusion
Although Google is still struggling with the bad reputation of killing off services, GCP Cloud AI remains a secure and reliable foundation for the company’s infrastructure, so it’s less likely to be placed on the chopping block any time soon. But if your company’s network and server infrastructure revolve around Microsoft products, go for Azure.
More AI tool reviews on HighFlyers: Genspark AI review and Magic Studio review.
