Moving your data or services into the cloud is a major decision that should be researched thoroughly.
At first glance, these large cloud providers all look the same, but understanding their differences will help you make the best decision to suit your goals. Let’s dive in and compare Google Cloud vs AWS.
Google Cloud Overview
The Google Cloud platform is one of the largest private interconnected networks and has availability in over 200 countries. While most think of Google Cloud as just a place for storage, Google’s Cloud platform offers a host of over 90 services ranging from DNS hosting to AI and machine learning tools.
Here’s a brief overview of the types of services Google Cloud offers:
Computing & Hosting – Allows you to build your own cloud-based infrastructure for complete serverless computing through Cloud Functions. Using Kubernetes you can host a variety of applications either as single nodes or in a cluster. Build entire services with virtual machines to host any image either privately or publicly.
Machine Learning – Uses pretrained models and algorithms to provide AI and machine learning services through the use of APIs. This includes services such as video intelligence, text to speech, language analysis, and conversational interfaces.
Storage – Stores instances of servers, applications, or files through five different types of storage options that offer their own unique availability and pricing options.
Databases – Offers cloud-based SQL and NoSQL databases services. Google Cloud Spanner gives additional flexibility and coverage from multinational companies needing global coverage.
Networking – Routes traffic and offers features such as load balancing, firewall rules, cloud DNS, and Virtual Private Cloud services.
Big Data – Allows larger organizations to process and query massive amounts of data through Google’s cloud computing. This varies from its database services as this requires a separate data analysis service called BigQuery. Massive datasets can be analyzed to help identify cures to disease, run statistical analysis, or even simulate complex astronomical events.
While Google Cloud used to sit firmly at the top of the food chain, Amazon Web Services has quickly given the platform a run for its money for the last few years. Let’s see how Amazon’s cloud services stack up.
Amazon Web Services Overview
Amazon Web Services (AWS) is a branch of Amazon.com and provides a multitude of cloud-based services in both the B2C and B2B space. Like Google Cloud, AWS offers services that cater to many workloads such as data processing, data archiving, game development, and web hosting.
Currently, AWS offers more than 175 cloud-based services. Let’s take a quick look at some of their core offerings.
Computing & Hosting – Through 11 different products, AWS offers flexible networking tools to cater to a variety of needs. Virtual Private Servers (VPS) can be hosted and launched in the cloud while Amazon EC2 lets you scale those services in a predictable way. AWS also has a feature called Wavelength, which can deliver ultra-low latency applications for 5G devices, and is great for IoT.
Machine Learning – Offers AI and ML services without the need to train an algorithm. Tools like DeepLens can analyze and identify events through a video feed while Amazon Comprehend can ingest massive amounts of text and provide insights and relationships based on patterns and anomalies. There are a few unique tools such as DeepRacer which allows you to program scale model vehicles and DeepComposer which enables machine learning to use a musical keyboard.
Storage – Like Google Cloud, AWS offers a series of flexible storage ranging from high-performance file systems to low-cost archive storage. AWS also offers CloudEndure and AWS backup which act as disaster recovery and AWS service synchronization.
Databases – Features products that can manage and host SQL and NoSQL databases as well as provide in-memory caching through Amazon ElastiCache. Amazon RDS allows for additional flexibility for other databases such as PostgreSQL, Oracle, and MariaDB.
Networking – Allows for the creations of isolated cloud resources with Amazon VPC and granular control of your applications with AWS Web Application Firewall. Other features include a Content Delivery Network (CDN), load balancing, application hosting, and cloud DNS.
Big Data – Amazon Braket offers a fully managed quantum computing service geared towards massive datasets used for research and discovery. A combination of classic and quantum algorithms can be run against massive datasets with those results being available in an Amazon S3 bucket.
At a glance, there are many overlapping services both platforms offer, but AWS has more specialized cloud-based tools available. Below we’ll take a more granular look at both Google Cloud and AWS’s service offerings.
Storage Capabilities
Disk storage can make all the difference when determining where you want to host your applications or store your data. The speed and availability a disk has will directly impact the performance of your services and data. There are two main types of storage both platforms use: block storage and object storage.
You can think of block storage as drive space for your applications and virtual machines. These drives need to be fast and have the ability to scale quickly alongside your services. You’ll want to pay special attention to a platform’s block storage if you plan on hosting services, databases, or anything that requires high availability in the cloud.
| Block Storage Features | AWS | Google Cloud |
| Volume Size | 1GB to 16TB | 1GB to 16TB |
| Replication | RAID-1 | Built-in redundancy |
| Encryption | 256-bit AES | 256-bit AES |
| Throughput per instance (SSD) | 4,750 MB/s | 240–1,200 MB/s |
| Max IOPS per volume | 20,000 (Provisioned) Max 75,000 per instance | 40,000 read 30,000 write |
Object storage, also known as distributed storage, is used for storing and accessing large numbers of independent objects such as documents, backups, images, and media. While block storage is dedicated to services and applications, object storage is unstructured.
While speed is still important, object-based storage doesn’t require the demanding low latency or high IOPS a server or app will need. If your cloud use heavily revolves around storage and not application hosting, it’s important to look at factors such as the size limit, object limit, and differences between hot and cold storage.
Hot storage is data that is ready to be accessed in a very short period of time, where cold storage is more for archival purposes and won’t need to be moved into a production environment in the immediate future.
| Storage Features | Amazon | Google Cloud |
| Object Limit | Unlimited | Unlimited |
| Size Limit | 5/TB per object | 5/TB per object |
| Cold Storage Platform | Glacier | GCS Coldline |
| Cool Storage Platform | S3 Infrequent Access | GCS Nearline |
| Hot Storage Platform | S3 Standard | Google Cloud Storage |
| Starting Storage Cost | $.03 per GB | $.02 per GB |
Both platforms allow the use of local storage devices for running virtual instances. In short, this gives a massive performance boost when compared to cloud-based block storage, with much less latency and a considerable increase in the input/output operations per second.
While this option isn’t for everyone, it does give additional flexibility and speed to organizations who want to manage a part of their own hardware. To put this into perspective, both Google Cloud and AWS get around 50,000 IOPS using cloud-based SSDs. Using a local SSD could increase that to well over 500,000 IOPS.
Security
Security in the cloud is a key concern, especially for organizations looking to run a majority of their services on a hosted platform. Overall both Google Cloud and AWS offer excellent security and identify access features, alongside tools you can use to self-audit and ensure you’re in compliance.
Security in Google Cloud
Google has deployed a robust Zero Trust model of security across its platforms for a number of years, and that system is applied to its cloud services. While there are dozens of different security features based on what tool you’re using, at its core Google Cloud is a very secure platform.
- Default encryption during transit. 256-bit AES encryption for persistent disks.
- Ability to be used in accordance with PCI and HIPAA standards.
- Audit logs with real-time access transparency.
- Data Loss Prevention.
- Identity and context aware access.
- Security key enforcement.
- Phishing protection.
- reCAPTCHA site/application protection.
- Incident response (Currently in alpha).
- Key storage/management.
Google Cloud has security as the backbone of its core offerings which can be seen from its infrastructure, all the way through its suite of security applications. There are a number of additional more granular security controls that administrators can use. For instance, VPC Service Controls can be used to specify a custom perimeter around your Google Cloud applications and services.
To protect virtual environments Google Cloud offers a Shielded VM, which uses measured hashes during the boot sequence to ensure none of your instances have been compromised on the boot or kernel-level. The integrity and baselines of your virtual environment can all be reviewed and monitored through integrity monitoring logs.
Security in Amazon Web Services
AWS also offers extremely thorough security features and best practices in its cloud network. Amazon has done a great job bundling these security services and giving them recognizable names, helping administrators keep their controls organized.
- Default encryption during transit. 256-bit AES encryption for persistent disks.
- Identify and access management.
- Managed threat detection.
- Configuration monitoring.
- DDoS protection.
- Network and application firewalls.
- Ability to be used in accordance with PCI and HIPAA standards.
- Incident response.
- Key storage/management.
AWS offers many in-depth security features deployed on top of the security that is built into the platform. A nice security feature by Amazon is its AWS Artifact service. This is a single hub where your organization can find and manage security compliance information for free in one place. This makes it easy to perform due-diligence and management agreements between you and AWS.
AWS Web Application Firewall (WAF) allows for customizable and agile protection of your hosted services and applications. This provides a nice balance between customizable policies and managed rules you can put to work right away. Similar to AWS hosted service pricing, you only pay for the rulesets you use with AWS WAF.
In short, both platforms offer robust security services, but you may find managing security in AWS a bit easier due to its more compartmentalized services.
Reliability
Both platforms have massive networks and dozens of products. We’ll examine factors such as uptime, network coverage, service level agreements, and support for each platform.
Both platforms provide detailed Service Level Agreements (SLA), with Google Cloud having more detailed SLAs broken down per service, making it a bit more complicated. When comparing AWS Compute and Google Cloud Compute Engine, both offer credits in return if an SLA is breached. We’ll be using those two services as an example.
For AWS, an uptime less than 95% will apply a service credit of 100% towards future payments of the service. This tapers off to 30% and finally 10%.
| AWS Monthly Uptime Percentage | Service Credit Percentage |
| Less than 99.99% but equal to or greater than 99.0% | 10% |
| Less than 99.0% but equal to or greater than 95.0% | 30% |
| Less than 95.0% | 100% |
Google Cloud applies a similar “Financial Credit” towards your account for downtime associated with your services. Google Cloud has two sets of compensation, one based on the downtime revolving around a single instance, and one for instances across multiple zones.
| Google Cloud Monthly Uptime Percentage | Service Credit Percentage (Multiple Zones) |
| 99.00% – < 99.99% | 10% |
| 95.00% – < 99.00% | 25% |
| < 95.00% | 50% |
| Google Cloud Monthly Uptime Percentage | Service Credit Percentage (Single Instance) |
| 95.00% – < 99.50% | 10% |
| 90.00% – < 95.00% | 25% |
| < 90.00% | 50% |
HIstorically both Google Cloud and AWS have a solid track record of uptime. While this uptime can fluctuate both AWS and Google Cloud offer logs to track their current uptime along with historic server outages.
Both AWS and Google Cloud offer basic and premium tiers of support. Free support for both platforms includes access to a knowledge base, a service health dashboard, support forums, and read-only versions of repair notes. Any detailed assistance will require you to pay; both offer unlimited account support with no long term contracts.
Billing & Pricing
Both platforms bill primarily around data usage but do so differently. Comparing them can be difficult to calculate even if you roughly know how much data you plan to use. Both platforms have built their own pricing calculators you can use to estimate costs.
Both the AWS Simple Monthly Calculator and Google Cloud Platform Pricing Calculator can be used to plan out what features you want to host, and what services you may want to opt-out of based on a combination of usage fees and fixed monthly costs.
While it can seem complicated, both platforms are making strides to simplify the process and reward companies for using their cloud platforms for the long term. For example, Google Cloud offers discounts when you commit to using a certain amount of its Compute Engine services. Committed Use Discounts essentially give you the option to buy usage time before you use it at a deeply discounted rate.
AWS has a similar program called Reversed Instances. This allows you to save money and remain flexible by avoiding on-demand pricing by reserving cloud capacity. According to AWS, you can save up to 72% when comparing reversed instances to peak demand prices.
Strengths & Weaknesses
While both platforms are evenly matched, there are a few distinctive pros and cons for each platform.
AWS Pros
- Fast and flexible.
- Allows organizations to plan and scale their services.
- Robust security.
- Pay as you go. (can still be complex).
AWS Cons
- Migrations off of AWS can be complex.
- Ease of us can allow for over-reliance on the platform.
- Robust AWS services require skilled administration.
- Tech support fees can be based on a percentage of your AWS usage.
Google Cloud Pros
- Global fiber network.
- Allows for flexibility and scalability.
- Robust security.
- Competitive global network coverage.
Google Cloud Cons
- More complex SLA structure.
- Sometimes difficult to navigate user interface.
- A slower rate of innovation compared to AWS.
- Pay as you go. (can still be complex).
AWS vs Google Cloud – Who Wins?
It’s truly tough picking a clear winner, as both platforms are extremely competitive in many key areas. From a technical standpoint, both Google Cloud and AWS have impressive service offerings, global coverage, and competitive support plans.
While both are excellent, we’ve found that AWS tends to do a better job at providing detailed documentation and creating a smoother user experience on their platform.
From a user experience perspective navigating the many different services in AWS feels less complex than with Google Cloud. AWS also is leading the way by offering more productized cloud-based services for customers.
Do you use AWS or Google Cloud? What are your biggest pet peeves with the platform? Let us know in the comments below.
