What's new in Cloud FinOps? – Details, episodes & analysis

Send us a text

January 2025 FinOps news.

Instances/Compute

• https://cloud.google.com/blog/products/compute/announcing-smaller-machine-types-for-a3-high-vms - The A3 Ultra machine type is available in the following region and zone: St. Ghislain, Belgium, Europe - europe-west1-b
• https://cloud.google.com/blog/products/compute/first-google-axion-processor-c4a-now-ga-with-titanium-ssd - Google Axion Processor-based C4A VMs with Titanium SSD are now generally available. Part of our general-purpose machine family, these instances come with up to 6 TiB of Titanium SSD disks. Titanium SSD is our latest generation of Local SSD. It uses Titanium I/O offload processing and offers enhanced SSD security, performance, and management.
• Compute Engine - Generally available: Managed instance groups (MIGs) let you create pools of suspended and stopped virtual machine (VM) instances. You can manually suspend and stop VMs in a MIG to save on costs, or use suspended and stopped pools to speed up scale out operations of your MIG. For more information, see About suspending and stopping VMs in a MIG.

Data/DBs/AI

• Azure OpenAI provisioned reservations (June - but not sure we mentioned it)

With the introduction of 1-month and 1-year Provisioned Reservations, businesses that commit to using Azure OpenAI on a scale can enjoy substantial discounts compared to the standard pay-as-you-go pricing. You can manage these reservations in Cost analysis in the same way as any other reservation.

For further details on how to purchase and manage these reservations, please refer to the article below: 

Save costs with Microsoft Azure OpenAI Service Provisioned Reservations  

Storage

• https://aws.amazon.com/about-aws/whats-new/2025/01/amazon-s3-metadata-generally-available 
• Generally available: Azure NetApp Files now supports minimum volume size of 50 GiB.
• Generally Available: Azure Files provisioned v2 billing model for HDD (Standard) - (Added to Roadmap 8th Jan 2025 - but released in October 24) The provisioned v2 model for Azure Files HDD (standard) pairs predictability of total cost of ownership with flexibility, allowing you to create a file share that meets your exact storage and performance requirements. Provisioned v2 shares enable independent provisioning of storage, IOPS, and throughput. In addition to predictable pricing and flexible provisioning, provisioned v2 also enables increased scale and performance, up to 256 TiB, 50,000 IOPS, and 5 GiB/sec of throughput; and per share monitoring. Provisioned v2 is generally available in a subset of regions. For the current list of available regions, see provisioned v2 availability.

Visibility (Billing conductor - Tags - cost categories)

• New fields for cost allocation (Enterprise Agreement customers)
  
  

You can use different le

Send us a text

In this chat, Frank, Stephen, and Anthony share their thoughts on how cloud computing and AI have evolved from 2024 to 2025. 

They discuss the growth in the GPU market and how views on cloud versus on-premises solutions have changed. They also talk about the quality and performance of AI models and consider the environmental impact of AI. The group highlights the need for specialized AI models and the efficiency of the hardware used in AI applications. They wrap up with some insights into future trends in cloud instances and pricing.

In their chat, the speakers look at how AMD and Intel processors perform. They also discuss the future of ARM and Graviton technologies. Finally, they touch on the shifting world of cloud computing. They examine the competition among major cloud providers, such as AWS and Microsoft. They also highlight how important it is to be transparent in cloud operations, especially when comparing European and American providers. Here are the key takeaways.

• The GPU market saw a significant increase in providers in 2024.
• Many organisations are reconsidering their cloud strategies due to pricing and privacy concerns over AI.
• AI adoption is still in its early stages for many companies.
• Quality of AI models has plateaued, with a few major players dominating the market.
• Training models in the cloud may not be the best solution for everyone.
• Specialised AI models are becoming increasingly important.
• The environmental impact of AI and cloud computing is a growing concern.
• Efficiency in hardware usage is crucial for AI applications.
• Transparency in cloud provider operations varies significantly between vendors.
• The cloud market is stabilizing, with slower growth rates observed.
• Energy consumption and cost are critical factors.
• Benchmarking is essential for understanding cloud performance.
• The competitive landscape among cloud providers is shifting, with pricing diverging.

Send us a text

In this episode, Frank and Steve discuss the July 2024 FinOps announcements from AWS, Azure, and Google Cloud. They cover topics such as committed use discounts, fully licensed commitments, hibernation for VMs, cost management features, and service discontinuations. The conversation is filled with insights and analysis of the latest developments in the cloud industry.

• There are new machine types available on GCP, such as the C3 bare metal machine types and the sixth generation Intel-based VMs.
• Serverless compute options, like Azure Databricks and OpenSearch serverless, provide efficient and cost-effective ways to run workloads without managing infrastructure.
• Storage enhancements, such as the Convert Azure Premium SSD v2 disk and Amazon FSX for Open ZFS, offer improved performance and cost optimization.
• Log management tools, like Azure Monitor Logs and AWS Cost Categories, help users collect, analyze, and act on telemetry data from various resources.
• Cost optimization features, like AWS Focus 1.0 and GCP backup commitment-based use discounts, provide ways to optimize cloud spending. Committed use discounts are available for backup for GKE, allowing users to commit to a consistent amount of usage in exchange for a discounted rate.
• AWS now offers fully licensed commitments and portable license commitments for purchasing VMware engine commitments, providing customers with more options for using VMware in the cloud.
• VM hibernation is now generally available in Azure and AWS, allowing users to save compute costs by persisting the VM state in memory and deallocating the VM.
• Azure has introduced a cost card in the portal, allowing engineers to estimate costs and adjust as needed before deploying VMs.
• Google Cloud's FinOps Hub now allows users to view their carbon footprint and estimated greenhouse gas emissions for their Google Cloud usage.
• AWS has announced the discontinuation of several services, including Cloud9, SimpleDB, Forecast, and CodeCommit, indicating a shift in focus and priorities.

Send us a text

SteveO and Frank sat down with Mel, a PwC engineer, to talk about what it's like to be an engineer and the hurdles they face. 

They dug into her daily routine - from stand-up meetings to ticket management and making time to focus. They also explored how random meetings can kill productivity and the importance of debugging. They touched on cool topics like pair programming, mob programming, and using tools like Stack Overflow and Git branches. 

This conversation is all about giving you a peek into an engineer's daily life and challenges. Mel shared her experience working with a bunch of tools and systems, including VDI, email, chat, Kubernetes, OpenLens, Terraform, Angular, React, Node, YAML, JSON, and Python. She explained how her team shares info about their system through tickets and updates in JIRA. 

Then, they talked about the struggles of saving money and how it's often overlooked in engineering roles. Mel and the hosts agreed that finance and engineering teams need to communicate better and work together. They also stressed the importance of prioritizing and getting feedback.

Send us a text

"An isomorphism is the fancy maths word for an analogy." -  Dr James Mitchell

Dr. James Mitchell, a commodities trading and renewable energy expert, joins Stephen and Frank to discuss FinOps financial language. He draws parallels between cloud computing and electricity economies, offering insights into future trends. The conversation explores market differences and the need for finance-engineering collaboration. 

James decodes cloud vendors' technical jargon, revealing financial roots. "Savings plans" and "reserved instances" are actually forward contracts. The discussion then shifts to unit economics' universal applicability across industries.

Optionality and flexibility emerge as key considerations in cloud pricing. The hosts emphasize managing commitments like a financial portfolio for optimal outcomes. This approach bridges the gap between technical and financial perspectives in cloud operations.

Key takeaways:

• Understanding financial language and communicating with finance professionals is crucial in FinOps.
• One can apply the concept of isomorphic economies to the cloud market to gain insights into its evolution and economics.
• Regulating the global cloud computing industry is challenging due to its global nature and lack of a competent authority.
• Better communication between finance and engineering teams is essential in FinOps. Building bridges between finance and engineering teams is crucial for effective communication and collaboration.
• A common language and understanding of financial terms is beneficial for both teams to work together effectively.
• Unit economics is a fundamental concept in finance and can be applied to various industries.
• Considering optionality and flexibility in pricing strategies is crucial for optimizing cost efficiency.
• Managing commitments as a portfolio can lead to better financial outcomes.

Send us a text

In this episode of What's New in Cloud FinOps, we (Frank and Stephen) discuss the latest news and updates in cloud computing impacting your wallet.

We cover topics such as right-sizing recommendations for Amazon RDS, data exports for cost optimization, and improved billing and cost management. We also highlight talks from FinOps X in San Diego and share their favourite sessions. 

Overall, the conversation provides valuable insights into the latest developments in cloud FinOps, including a rant from Stephen on the hardship that is finding good content for Azure. 

Send us a text

In this live episode, recorded at FinOps X 2024, we speak with TJ and Deana Solis about the social aspects of FinOps.

It is a real live episode with microphones not working, music being turned on, bar opening and more. I also had to use a 'beep' sound for the first time.

Send us a text

Izhak "I don't believe in tagging" Zimmermann discusses with Stephen and Frank about the importance of focusing on business value in the context of FinOps, emphasising the need to understand cloud costs within the context of business goals. He introduces a new approach to FinOps that leverages new technology and insights to support business goals and drive business value. The conversation covers the challenges of understanding cloud costs, the use of networking data to understand costs, and the transformation of cost data into business value data.

Send us a text

Summary
In this episode of What's New in Cloud FinOps, Frank and Stephen discuss a wide range of cloud-related news and updates. They cover topics such as Azure VM hibernation, Azure Compute Fleet, Google Cloud TPU, Amazon EC2 C7i Flex, DynamoDB, AWS Marketplace, Cloud Run, and more. The conversation also delves into the complexities of cloud pricing, energy progress, and the impact of cloud technology on businesses.

Send us a text

In this episode, Steve, Frank and Mike discuss the benefits of using the free tier in Microsoft Azure. They explore the various free offerings available, such as Azure Cosmos DB, Azure DevOps, and speech-to-text translation. They also discuss the importance of monitoring costs and implementing processes to prevent unexpected expenses. Mike shares insights on cost optimization for Cosmos DB and announces an upcoming video on the topic.