Thursday, October 6, 2022

Google Compute Engine Review

Google’s Compute Engine, sometimes called Google Cloud Compute Engine, delivers on-demand virtual machines (VMs) to provide flexible computing needs within the Google Cloud.

Google Compute Engine deploys VMs as needed and enables organizations to obtain computers on an hourly or even a second-by-second basis.

See below to learn all about Google Compute Engine and where it stands in the compute sector:

Google Compute Engine and the Compute Market

Google Compute Engine competes in the cloud computing market. Allied Market Research forecasted this market to grow with a compound annual growth rate (CAGR) of 15.8% from $325 billion in 2019 and reach $1.6 trillion by 2030.

In 2021, Markets and Markets estimated a base market revenue of $445.3 billion market with a CAGR of 16.3% that will reach $947 billion by 2026. Grand View Research projected a smaller market size of only $369 billion and a slower CAGR of 15.7%.

The major competitors in the cloud computing market include: Alibaba; Amazon Web Services (AWS); Hewlett Packard Enterprise (HPE); IBM; Microsoft; and Oracle.

Google Compute Engine Key Features

  • Virtual Machines
    • VM manager
    • Confidential VMs
    • Live migration for VMs
    • Sole-tenant nodes
    • Support for Docker containers with Google Kubernetes Engine
  • Operational certainty and flexibility
    • Custom machine types
    • Predefined machine types
    • Affordable spot VM
    • Instance groups
    • Persistent disks
    • Local SSD (solid-state drive) block storage
    • Linux & Windows support
    • Per-second billing
    • Reserved VM instances
  • Optimized Performance
    • Graphical processing unit (GPU) accelerators to boost workload performance
    • Global load balancing
    • Automated OS patching
    • ML-recommended machine type suggestions
    • Placement policies
    • Suspended VMs (see pricing)
  • Customers can buy software licenses (license-included instances) or transfer them from existing deployments (bring-your-own-license, or BYOL, instances)

Google Compute Engine Key Benefits

Organizations that use Google Compute Engine or other cloud computing seek the following key benefits:

Cloud Cost Savings

Moving to the cloud eliminates the operational burdens and expenses of physical data centers, such as cooling systems, physical security systems, and electrical power. Cloud computing also outsources the IT resources otherwise required to install, update, maintain, troubleshoot, and retire the bare-metal IT infrastructure.

High-Performance Computing Resources

Once a computer is purchased for a physical data center, it will be cutting edge only until the next evolution of computer components (processors, graphic cards, etc.) ships. Additionally, a physical data center will be limited to the computing power on-hand and those resources must be divided among all existing jobs and tasks for the organization.

Moving to a cloud environment like Google Compute Engine allows data researchers to select the latest, most powerful equipment and devote as many resources as needed. Short-burst, high-volume computing needs can analyze genomics or run simulations quickly and cost-effectively on the cloud.

Operational Efficiency

Cloud computing not only allows data centers to grow to be as powerful and large as needed, they can also shrink when the need passes to avoid unnecessary ongoing expenses. Organizations can optimize for performance, costs, or power based upon the specific needs of a specific job by changing the parameters of the VM resources.

Google Compute Engine offers price reductions for commitments to resources, interruptible resources, and sustained operation. There is even an option to hibernate or suspend the operation of a VM and store its status on the cloud at an even lower rate.

Google Compute Engine Use Cases

Airbus

As a unit of Airbus Defense and Space, Airbus Intelligence provides high-resolution satellite imagery to customers on their OneAtlas platform. Airbus Intelligence needed a solution that could process as much as 2PB of imagery data annually, incorporate artificial intelligence algorithms, and improve efficiency over the legacy solution.

Selecting Google Compute Engine permitted Airbus to reduce its cloud-detection error rate from 11% to 3% and lower the time needed to stream images from hours and days to less than a half-second.

“Google Cloud Platform gives us access to much more compute power than we had in the past, so we can do so much more than we used to,” said Laurent Gabet, optical R&D manager at Airbus. “Back then, we worked on projects. Now, the whole world is our playground.”

Etsy 

Etsy’s global marketplace offers more than 66 million unique products from 2.8 million sellers and needs to scale on demand without sacrificing teamwork or sustainability. Etsy selected Google Compute Engine because Google engineers worked with Etsy team members as partners, committed to carbon-neutral computing, and could provide flexible delivery of scale.

“The first priority was to clear away the wasted resources created by our use of a lift-and-shift approach,” said Dany Daya, senior program manager for cloud migration at Etsy. “Some of this was low-hanging fruit, such as minimizing or resizing solid-state drives (SSDs) to match their load, and resizing virtual machines for utilization and performance. We compressed the data and used the different storage categories provided by Google Cloud.”

Etsy also used Google’s Compute Engine committed-use discounts (CUDs) to lower its compute cost by 42%.

Tokopedia 

A pioneer in e-commerce in Indonesia, Tokopedia launched a shopping festival that delivered 20x the usual traffic over 332 million visits from 73 million visitors — more than Tokopedia enjoyed during their first five years in business. Tokopedia wanted to build on that success by transitioning to the cloud, but their internal teams did not have sufficient experience.

By selecting Google Compute Engine, Tokopedia’s engineers could leverage the experience of the Google team to make up for their own inexperience.

“We benefited from the knowledge of Google Cloud engineers who have the experience of running large-scale events,” explained Tahir Hashmi, vice president of engineering at Tokopedia. “If we had to roll out the project with our limited resources, we would have to read a lot of documentation, run many experiments, and perhaps still end up in blind alleys.”

Google Compute Engine Differentiators

When choosing Google Compute Engine, customers often do so because of the following key differentiators:

Confidential VMs

Google’s breakthrough VM technology permits encrypted processing of data that does not affect performance. Confidential or sensitive information can be manipulated without exposing it by maintaining encryption throughout all processing procedures.

Custom Machine Types

Although Google offers a huge range of preset server types, Google also allows customers to create their own custom VMs tailored for their specific needs.

Google Brand Power

While No. 3 in the cloud computing market, Google remains a force to be reckoned with. Customers will feel comfortable hosting data in a Google-backed cloud knowing that the Alphabet parent-company finances provide security and stability for the long term.

Google Ecosystem

Google offers many tools and utilities directly, but also works with many partners to create a Google Marketplace with additional tools and resources that may be licensed for use. Applications and services offered in the marketplace provide solutions for analytics, big data, developers, and many more needs.

Google Compute Engine User Reviews

Review site Rating
TrustRadius 8.2 out of 10
G2 4.3 out of 5
Capterra 4.0 out of 5

Google Compute Engine Pricing

Google Compute Engine pricing is billed by the second per GB with a minimum usage of one minute. Google provides a pricing calculator to help potential customers budget for their needs.

New customers receive $300 in free credits during their first 90 days and all customers receive a general purpose e2-micro instance per month for free in addition to their purchased VMs. Disk size, machine type memory, and network usage all play a significant role in the price of the resource.

Software licenses (SQL, etc.) will be needed for the VMs and may be purchased with the VM instance, or existing licenses may be transferred from elsewhere. Google offers discounts for sustained use, committed use, spot VM instances, or preemptible VM usage; however, only one discount may apply at any time.

Customers may request resources in regions hosted around the world, but VMs are not guaranteed to be available in every zone. Customers can reserve predefined machines in a specific zone.

Google offers VMs for:

  • General purpose
  • Compute-optimized
  • Memory-optimized
  • Accelerator-optimized
  • Shared core machines
  • Higher bandwidth

VMs can be suspended with memory and device state preserved in storage at a reduced rate. This suspended VM can easily be relaunched if needed and in the meantime saves money for the customer.

Conclusions

Many growing organizations find themselves out-pacing the capabilities of internally managed data centers and on-site computing. Moving to the cloud can provide cost and scale advantages that deliver growth at a rapid pace without rapidly accelerating long-term costs.

Organizations considering a move to the cloud should consider Google as a serious contender for their business. Data teams should definitely include Google Compute Engine for evaluation testing under all scenarios to determine if Google’s solution might be the best fit for their needs.

Similar articles

Latest Articles