January 31, 2025

289: DORA The Explorer… Of EU Regulations

51 minutes

Welcome to episode 289 of The Cloud Pod – where the forecast is always cloudy! Justin, Ryan, and Matt are here this week to bring you a riveting podcast on EU regulations! Are you asleep yet? No? Ok great. We promise it will be a good show – despite the title.

Titles we almost went with this week:

Stargate: We’re not saying its Aliens, but its $500 Billion

️AWS: Now with extra sessions

EC2 Flex: Bigger, Badder and Probably still expensive

SNS FIFO: So fast, it’ll give you whiplash

‍⚖️Azure: Now with added Legalese (Thanks, EU)

OpenAI’s Stargate: From Chatbots to Interdimensional Travel (maybe)

☢️GCP’s Biochar Initiative: Turning Waste into… Well, Less Waste (hopefully)

️AWS Console Multiple Sessions: So you can prove you dropped those databases from multiple accounts

☁️Amazon still adds new features to SNS and the cloud pod is impressed

☠️AWS tries to kill chrome profiles

A big thanks to this week’s sponsor:

We’re sponsorless! Want to get your brand, company, or service in front of a very enthusiastic group of cloud news seekers? You’ve come to the right place! Send us an email or hit us up on our slack channel for more info.

AI IS Going Great – Or How ML Makes All Its Money

01:47 Announcing The Stargate Project

Open AI announced a joint investment of $500 billion dollars over the next four years building new AI infrastructure for OpenAI in the US, with the intent to deploy $100B immediately.

This infrastructure will secure American leadership in AI, create hundreds of thousands of American jobs, and generate massive economic benefits for the entire world.

The initial equity funders in stargate are SoftBank, OpenAI, Oracle and MGX.

Softbank and OpenAI are the lead partners for Stargate, with Softbank having financial responsibility, and OpenAI having operational responsibility.

Arm, Microsoft, Nvidia, Oracle and OpenAI are the key initial technology partners.

The buildout is currently underway starting in Texas, and they are evaluating potential sites across the country for more campuses as they finalize definitive agreements.

As part of Stargate, Oracle, Nvidia and OpenAI will closely collaborate to build and operate this computing system. This builds on a deep collaboration between OpenAI and NVIDIA going back to 2016, and a newer partnership between OpenAI and Oracle.

This also builds on the existing OpenAI partnership with Microsoft. OpenAI will continue to increase its consumption of Azure as OpenAI continues its work with Microsoft with this additional computer to train leading models and deliver great products and services.

“All of us look forward to continuing to build and develop AI—and in particular AGI—for the benefit of all of humanity.” This quote TOTALLY didn’t terrify us…

Our conversations back in December about OpenAI trying to figure out their ownership model makes a lot more sense now.

07:22 Justin – “…it’s interesting that SoftBank is investing so much money into it considering, you know, the trade issues with China and SoftBank, you know, being mostly Chinese owned and invested in. Yeah. It’s one of the things about SoftBank that’s interesting as well as I didn’t think their funds had done that well after crypto kind of blew up on them in pretty spectacular ways, although it’s back apparently. So yeah, it’s interesting, you know, post inauguration day.”

AWS

06:11 The AWS Management Console now supports simultaneous sign-in for multiple AWS accounts

AWS is announcing multi-session support, which enables AWS customers to access multiple AWS accounts simultaneously in the AWS Console.

You can sign in to up to 5 sessions in a single browser

This can be a combination of root, IAM or federated roles in different accounts or the same account.

WOW! This is a huge improvement and we’re REALLY excited about this.

Tab junkies, rejoice!

07:05 Ryan – “This is the biggest thing they’ve ever announced ever.”

15:27 Introducing new larger sizes on Amazon EC2 Flex instances

AWS announced the general availability of two new larger size (12xlarge and 16xlarge) EC2 Flex instances (c7i and m7i variants)

This new size expands the EC2 flex portfolio, providing additional compute options to scale up existing workloads or run larger-sized applications that need extra memory.

These instances are powered by custom 4th gen Intel Xeon scalable processors, only available on AWS, and offer up to 15% better performance over comparable x86-based intel processors.

17:08 Ryan – “It’s the way that you had to provision memory and CPU and the relationship between the chosen Amazon was the instance type. And now I think if you select this instance type, you can tune those specifically and do a little bit of shaping.”

17:59 AWS CodeBuild now supports test splitting and parallelism

You can now split and run your tests across multiple parallel-running compute environments.

Based on your sharding strategy, CodeBuild will divide and run your tests across the specified number of parallel environments.

18:12 Justin – “Now I appreciate this, but, uh, I would like to run multiple tasks on the same environment versus spending more money on more parallel environments. Or I’d like you to handle all the automation of spinning up all those parallel environments so I don’t have to do that. So if CodeBuild could get on that part of it, I’d be much happier.”

20:17 AWS CodePipeline introduces new debugging experience in AWS Management Console

AWS Code pipeline now offers an enhanced debugging experience in the AWS Management Console, enabling you to identify and resolve pipeline failures more efficiently.

The new debugging interface gives you a dedicated debugging page, accessible via the action details button.

20:47 Justin – “…now I can curse out CodeBuild and Code Pipeline at the same time!”

24:20 Announcing high-throughput mode for Amazon SNS FIFO Topics

Amazon continues to push the high-throughput boundaries of SNS FIFO topics. Now with default throughput matching SNS standard topics across all regions. When you enable high-throughput mode, SNS fifo topics will maintain order within the message group, while reducing the deduplication scope to the message-group level.

WIth this change you can leverage up to 30k messages per second per account by default in US-East region, and 9k MPS per account in US west and Europe regions, and request quota increases for additional throughput in any region.

25:09 Justin – “It’s still cheaper than Kafka!”

GCP

26:37 GKE delivers breakthrough Horizontal Pod Autoscaler performance

Google Cloud is committed to providing the fastest and most reliable K8 platform with GKE.

Now, they are announcing an improved Horizontal Pod Autoscaler (HPA), the K8 features that automatically update workload resources to match demand.

With the new performance HPA profile you get

2x faster scaling

Improved metrics resolution

Linear scaling to up to 1000 HPA objects

This matters because customers are regularly asking for it, and they frequently over provision resources to account for delays in the autoscaling stack, resulting in lower efficiency and higher costs.

You’ll minimize waste

Improve application responsiveness

Increase operational efficiency

“With GKE’s Performance HPA profile, we’ve witnessed a remarkable boost in horizontal auto-scaling speed. In our tests with over 1000 HPA objects, workloads scaled up twice as fast. We’re excited to leverage this performance enhancement in our production environments.” – Sophy Cao, Senior Engineer, Spotify.

27:34 Ryan – “So I am dubious because every time I’ve. If my day job scaled up larger events while if you can create the containers great, but something else is going to fall down within the Kubernetes infrastructure. And so I was flabbergasted when joining a new team and I found out that they still have a huge process to warm their pods by pre-launching containers, because they found that they would crash the DNS server container or another sidecar that did proxying or something else in there. So I’m hoping that this profile will fix a lot of those issues.”

29:19 The EU’s DORA regulation has arrived. Google Cloud is ready to help

No, it’s not *that* DORA.

The Digital Operational Resilience Act (DORA) takes effect as of Jan 17th, and financial entities in the EU must rise to a new level of operational resilience in the face of ever evolving digital threats.

To help you tackle this new set of regulations, Google is sharing the DORA customers guides on Register of Information and Information and Communications Technology (ICT) Risk Management and their new Google Cloud Third-Party Risk Management Resource Center.

In addition, financial entities can request their DORA subcontractor listing today.

29:52 Matthew – “When I had to do some research on this for my day job, it looks like it mainly maps over to ISO. So if you are ISO 2700 and one, you’re mostly covered for this, which does make your life easier. I’m waiting for Azure to kind of come out with their same offering because it will make my life a little bit easier.”

31:19 C4A, the first Google Axion Processor, now GA with Titanium SSD

Google is making the new C4A virtual machines with Titaniums SSDS generally available.

The Titanium SSD is custom designed for Google Cloud workloads that require real-time data processing, with low-latency and high-throughput storage performance.

Titanium SSDs on C4A VMs deliver storage performance of up to 2.4M random read IOPS, up to 10.4GiB/s of read throughput, and up to 35% lower access latency than previous-generation SSDs.

C4A is a VM instance family, based on Google Axion Processors, that give you a 65% better price performance and up to 60% better energy efficiency than comparable current generation x86 based instances.

C4A with Titanium SSDs, offer up to 72 vCPU’s, 576 GB of memory and 6tb of local storage in two shapes – standard with 4gb of memory per vcpu and high-memory with 8gb of memory per vcpu.

They both support up to 50 gbps of standard bandwidth and up to 100 gbps with tier 1 networking for high traffic applications, as well as supporting the latest generations of Balanced And Extreme hyperdisk storage.

34:13 Justin – “This is the not the first axion processor. This is one of the second models they’ve released with it. This is the first one with the axion and the titanium SSD.”

35:08 Smaller Releases of note:

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.

Congratulations Google, you’re doing what everyone else is already doing.

36:56 Google is supporting new solar projects in Oklahoma

Google has entered into long-term agreements with Leeward Renewable Energy to support over 700 megawatts of solar projects in Oklahoma.

Google says it’s strategically located to support their data center operations, with one being less than one mile from their data center in Pryor, Oklahoma.

38:11 We’re announcing our first partnerships to scale biochar for CO2 removal

Google is announcing two long-term purchase agreements to help scale biochar as a carbon removal solution.

Partnering with Varaha and Charm to purchase 100,000 tons of biochar carbon removal from each company by 2030.

This will enable them to remove 200,000 tons of carbon, helping them achieve their net zero emissions goal, as well as help catalyze biochar production towards a scale that helps the planet mitigate climate change.

We think this one is cool, and hope companies continue with their green initiatives in this new administration.

Show copywriter note re: velociraptors with feathers. If you’ve ever seen a chicken stalk and eat a mouse or a lizard – you’ve seen velociraptors in action. They’re *terrifying*. Thank God chickens are small.

Azure

40:36 Innovating in line with the European Union’s AI Act

Microsoft during its recent AI tours, took a chance to meet with EU regulators and politicians to discuss AI and the new European Union AI Act, this is the first comprehensive legal framework for AI.

It aims to ensure that AI systems developed and used in the EU are safe, trustworthy and respect fundamental rights.

The Act classifies AI systems based on their risk level into: Unacceptable, High, LImited or Minimal or No Risk.

Microsoft approach to compliance against the act includes:

Proactive approach – they are preparing by conducting internal reviews, updating internal policies and contracts and engaging with policymakers directly

Focus on customer support by ensuring that documentation, tools and guidance all consider the act.

Shared responsibilities between AI providers and users, and emphasizing the need for compliance

Building compliant products

Plan to publish transparency reports and provide documentation for its AI systems to help customers understand their capabilities and limitations

Collaboration with policy makers, regulators and industry groups to shape the implementation of the act and ensure effective and interoperable practices

41:40 Ryan – “I both love and hate this. I feel like we have no idea what we’re doing yet and we’re trying to regulate it. And so it seems like that’s going to be a problem – because there’s so much in here where it’s like there’s plans to do a thing. They’re going to put the frameworks together. None of it exists. And we all know how fast compliance can grow and adapt to a changing technology ecosystem because our day jobs are super fun at times.”

44:25 Microsoft joins CISPE, the Euro cloud crew that tried to curb its licensing

In the old adage if you can’t beat them, join them – and Microsoft is the latest member of CISPE months after it negotiated a settlement with the trade association of European Cloud Providers over alleged anti-competitive software practices.

Of course not all members of the group are happy with the move.

AWS of course opposed Microsoft joining, but was outvoted by the board.

Google also attempted to join the board, but ended up joining the Open Cloud Coalition instead.

45:26 Ryan – “I don’t know if it’s a good model, right? It’s like lobbying and bribery just out in the open.”

47:08 Microsoft and OpenAI evolve partnership to drive the next phase of AI

Microsoft is thrilled to continue their strategic partnership with OpenAI and to partner on Stargate.

The announcement is complementary to what the two companies have been working on since they got together in 2019.

The key elements of the partnership remain in place for the duration of the contract through 2030, with their access to Open AI’s IP, revenue sharing arrangements and exclusivity on OpenAI API’s all continuing going forward.

Microsoft has rights to OpenAI IP (inclusive of model and infrastructure) for use within our products like Copilot. This means our customers have access to the best model for their needs.

The OpenAI API is exclusive to Azure, runs on Azure and is also available through the Azure OpenAI Service. This agreement means customers benefit from having access to leading models on Microsoft platforms and direct from OpenAI.

Microsoft and OpenAI have revenue sharing agreements that flow both ways, ensuring that both companies benefit from increased use of new and existing models.

Microsoft remains a major investor in OpenAI, providing funding and capacity to support their advancements and, in turn, benefiting from their growth in valuation.

Open AI recently made a new, large Azure commitment that will continue to support all OpenAI products as well as training. This new agreement also includes changes to the exclusivity on new capacity, moving to a model where MSFT has a right of first refusal. To further support OpenAI, Microsoft has approved OpenAI’s ability to build additional capacity, primarily for research and training of models.

49:18 Matthew – “It’s interesting also that Microsoft had to approve the opening ability to even be a part of that. At least that was the last sentence I read, you know, so I guess in the original agreement, maybe there was some like control that.”

Closing

And that is the week in the cloud! Visit our website, the home of the Cloud Pod where you can join our newsletter, slack team, send feedback or ask questions at theCloud Pod.net or tweet at us with hashtag #theCloudPod

...more

View all episodes

By Justin Brodley, Jonathan Baker, Ryan Lucas and Matt Kohn

4.9

3434 ratings