Already feeling the vibe of this year’s festive season? Here’s some fresh cloud technology news to keep you on the right track.
This series brings you up to speed with the latest releases, acquisitions, research, and hidden gems in the world of cloud computing – the stuff actually worth reading.
What happened in the cloud world this November? Keep on reading to find out!
Story of the month: All the new things at AWS, re:Invent style
Our team is at AWS re:Invent right now and shared some blazingly hot news with us, including the big news from AWS CEO Adam Selipsky’s keynote.
Here are the top things that stood out to us:
- More speed for less money! AWS announced Graviton 3, the freshest generation of its Arm-based Graviton processors. The new chip will be 25% faster and use 60% less power, powering some new instance types AWS is rolling out as well (C7g).
- AWS released its open-source autoscaling tool Karpenter to production. We hope that it helps people save money on their cloud bills because back when we tried an earlier version a year ago it wasn’t all that different from a vanilla Kubernetes cluster autoscaler.
- Machine learning made easy with a new service called Amazon SageMaker Canvas made for non-specialized engineers or business users. The promise? Anyone can build machine learning prediction models using a point-and-click interface.
The Business of Cloud
The top cloud vendors have a hard time breaking our CAPEX on their IaaS and SaaS offerings, but they manage to make it overall. Take a look at the recent public cloud growth rates, and you’ll see a correlation between capital expenditure and revenue. The top cloud vendor, AWS grew 10% faster in the last quarter than it did a year ago. Scale matters a lot in the cloud and CAPEX helps to secure it.
Dell is setting up one of the largest corporate spin-offs ever by shedding its 81% stake in VMware to create a software company with a stock market value of almost $64 billion. This is the epilogue to Dell’s 8-year-long commitment to transforming his interest in the slightly forgotten PC maker into a data center hardware and software company worth $40 billion.
In an interesting turn of events, the UK’s spy agencies - MI5, MI6, and GCHQ - are going to use AWS for cloud storage and computing.
Source: Financial Times
Nvidia’s acquisition of Arm saga just released another chapter. The deal is now under scrutiny by the FTC, and the EU won’t hesitate to shut it either. The global chip shortage is probably one of the reasons why.
The cloud data service vendor partly owned by an Indonesian tycoon - DCI Indonesia - just became the world’s top IPO of 2021, noting a 10,852% increase since its share float in January. This makes the company the second-biggest contributor to this year’s 12% gain of the Jakarta Composite Index.
Food for thought
Single and private clouds are out of the door. IBM released a study where it shows how this trend dropped from 29% in 2019 to only 2% in 2021. Who won the race instead? Hybrid cloud and multi cloud! Yet another piece of research showing us that multi cloud is set to become the dominant architecture for cloud service delivery.
So, Google Cloud went down in the middle of the month and took down services like Home Depot, Snap, and Spotify. What caused the outage? A glitch in a network configuration. It’s yet another scenario demonstrating the risk of betting on a single cloud provider to handle all your applications.
Source: Google Cloud Status Dashboard
Meanwhile in CAST AI
Here are some new product features hot off the press:
- If you have ever worked with manual cost management tools, you know how time-consuming that can be. We built a feature that brings you from zero to an optimized state in a few minutes - Instant Rebalancing. Learn more about it here.
- Also, we are officially SOC 2 Type II compliant!
- We also added the Spot Fallback feature that guarantees capacity by temporarily moving your workloads to other nodes until there are new spot instances available.
- The Available savings report is now enhanced with a graph that displays point in time actual and optimal cluster costs as well as other dimensions (i.e. CPU, Memory, node count).
- Scrapable cluster metrics are now available, here’s a handy guide.
- For AWS/Kops clusters we previously deployed a Lambda function per cluster, it’s no longer the case. From now on a single Lambda function is deployed per account.
- Added support for Kops version 1.20.
Why not get the next portion of cloud technology news directly to your inbox?