TheCUBE’s Rob Strechay kicks off this KubeCon + CloudNativeCon North America preview with Backblaze CEO and co-founder Gleb Budman, discussing why storage is the cornerstone to proper AI deployment. Budman outlines how Backblaze B2 – an S3-compatible, affordable, high-performance storage platform – is used across the AI pipeline: data collection, labeling, training, inference and logging. He shares how organizations are minimizing the idle time for graphics processing units time by staging data for fast access. Budman highlights B2 Overdrive, described as “the fastest throughput-per-dollar storage,” delivering up to a terabit per second to hyperscalers, neocloud GPU providers or on-premises clusters. Unpredictable AI costs come up in conversation, with Budman emphasizing fixed per-TB pricing and free egress so teams can move data where they need it without surprise charges.
Budman goes on to detail Universal Data Migration, a service that moves data from AWS, Google, Microsoft, on-prem and more into Backblaze while helping to subsidize source egress fees, enabling “land once, use many” strategies across clouds and Kubernetes environments. He also previews Backblaze’s first-ever KubeCon presence at Booth 1742 near Startup Village – complete with a limited-edition Overdrive hot sauce – illustrating how Backblaze is bringing the spice to AI at scale.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
KubeCon + CloudNativeCon NA 2025. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for KubeCon + CloudNativeCon NA 2025
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for KubeCon + CloudNativeCon NA 2025.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
KubeCon + CloudNativeCon NA 2025. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to KubeCon + CloudNativeCon NA 2025
Please sign in with LinkedIn to continue to KubeCon + CloudNativeCon NA 2025. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Gleb Budman, Backblaze
TheCUBE’s Rob Strechay kicks off this KubeCon + CloudNativeCon North America preview with Backblaze CEO and co-founder Gleb Budman, discussing why storage is the cornerstone to proper AI deployment. Budman outlines how Backblaze B2 – an S3-compatible, affordable, high-performance storage platform – is used across the AI pipeline: data collection, labeling, training, inference and logging. He shares how organizations are minimizing the idle time for graphics processing units time by staging data for fast access. Budman highlights B2 Overdrive, described as “the fastest throughput-per-dollar storage,” delivering up to a terabit per second to hyperscalers, neocloud GPU providers or on-premises clusters. Unpredictable AI costs come up in conversation, with Budman emphasizing fixed per-TB pricing and free egress so teams can move data where they need it without surprise charges.
Budman goes on to detail Universal Data Migration, a service that moves data from AWS, Google, Microsoft, on-prem and more into Backblaze while helping to subsidize source egress fees, enabling “land once, use many” strategies across clouds and Kubernetes environments. He also previews Backblaze’s first-ever KubeCon presence at Booth 1742 near Startup Village – complete with a limited-edition Overdrive hot sauce – illustrating how Backblaze is bringing the spice to AI at scale.
TheCUBE’s Rob Strechay kicks off this KubeCon + CloudNativeCon North America preview with Backblaze CEO and co-founder Gleb Budman, discussing why storage is the cornerstone to proper AI deployment. Budman outlines how Backblaze B2 – an S3-compatible, affordable, high-performance storage platform – is used across the AI pipeline: data collection, labeling, training, inference and logging. He shares how organizations are minimizing the idle time for graphics processing units time by staging data for fast access. Budman highlights B2 Overdrive, described as “th...Read more
exploreKeep Exploring
What is the focus and background of Backblaze as a cloud storage company?add
What is B2 Overdrive and how does it address customer needs in data accumulation and GPU utilization?add
What features does Backblaze offer for enterprise-level cloud storage solutions?add
What cloud storage features and philosophies does Backblaze offer for enterprise users?add
>> Hello, and welcome to this KubeCon CloudNativeCon North America 2025 preview. We'll be live in Atlanta in a short time, but today we'll be diving into one of the most critical areas of supporting organizations' journeys to AI, and that's storage. I'm really excited to be joined by Gleb Budman, who's the CEO and co-founder of Backblaze. Welcome on board, Gleb.
Gleb Budman
>> Thank you, Rob. Good to be here and looking forward to the chat, and also looking forward to KubeCon, or KubeCon, depending on how we all pronounce it.
Rob Strechay
>> Again, I look at it, and I think as long as you show up, they really don't care. I think it should be CloudNativeCon because, really, Kubernetes now being almost 11 years old has really had its time. But one of the things that maybe folks don't know Backblaze as much as KubeCon or KubeCon, depending on how you want to pronounce it. Why don't you give us an idea on what you're focused on and how it was founded and came to be about?
Gleb Budman
>> Backblaze is a cloud storage company, and we are basically creating the data backbone for the AIH. We provide really affordable, really fast, easy-to-use cloud storage. It's S3 compatible and it's ideal for all kinds of applications, AI, media streaming, surveillance, and the like. We got founded because we started the company actually building a backup application, and we needed a place to put all of that data. We originally were planning on putting the data on Amazon S3. We realized we were going to lose money on every single customer, and so we ended up starting from first principles, figuring out how to take hard drives and attach them to the internet. We wrote our own file system, we wrote our own infrastructure as a service layer, we built our own platform, and we ended up designing this large scale, very efficient cloud storage system. And then customers came to us and said, "Hey, I want access to that." And so for the last decade or so we've been offering Backblaze B2, which is our cloud storage offering for all of these application and AI use cases.
Rob Strechay
>> Yeah, I think you just hit the nail on the head. Again, cost and affordability is one of the key things when we start talking to organizations. Let's dive in there, because you can't have AI without data, and data has to live somewhere. What are you seeing from organizations as they dive down the AI path and really embark on this journey?
Gleb Budman
>> It's fascinating. We're seeing the entire workflow of what it looks like to build and use AI in the processing pipelines. First, companies need collect all the data that they want to use, then they want to label it, they want to build their models, they want to then do inferencing off of those models, and then they want to log and track all of the information. And so we're seeing Backblaze be used in all of those different parts of the pipeline. One of the things that we're seeing, for example, is customers are spending on an average 100 days at times training their model. It's very expensive to use the GPUs for all of that training, and so they want to make sure that they have all the data ready and then able to get to those GPUs as fast as possible when they're ready to go. AI is being used across just so many different applications. We have customers using it in their surveillance workflows, they're using it to build and design buildings better, using it to do drug discovery. The number of places that AI is being used is amazing, and they're needing to create large data sets to train the models, and then the gen AI part of it is also creating data sets itself, and then the inferencing workloads are all part of where I think all of this is headed. It's just going to become part of the fabric of the internet.
Rob Strechay
>> Yeah. And you're seeing some pretty significant momentum, I would expect, in this space as well. I had heard 70% increase in AI customers and over 40x growth in that data stored, but one of the things that was also important was that you had a new offering within B2, and it's called B2 Overdrive. Why don't you help people understand why that came about?
Gleb Budman
>> B2 Overdrive came about because we were actually talking with customers and prospects and hearing this problem from them, which is that, when they needed to accumulate all of this data and when they were ready, they wanted to get it to their GPUs as fast as possible, and it was very expensive to have those GPUs sit idle. And so B2 Overdrive, we believe, is the fastest throughput- per-dollar storage on the market. And so B2 Overdrive can give you up to a terabit per second of throughput at an affordable rate. And so customers are accumulating, accumulating, accumulating data, and then they use B2 Overdrive to get it to either one of the hyperscalers, one of the Neocloud GPU providers, or even their own GPU clusters, and get it there as fast as possible.
Rob Strechay
>> Yeah, yeah. You hit the nail on the head again, because I like how you talked about not only the speed of it, but the cost of it. Because I think one of the biggest issues we hear from organizations is the cost aspect of AI and how it can be so unpredictable, to put it mildly. What really sets Backblaze apart from this perspective in cost?
Gleb Budman
>> We hear that a lot, and we've heard that from customers for years, that it's not just the cost, but it's the unpredictability, these surprise fees. You're running your business, you're running your application, and then, all of a sudden, you get this bill afterward that is just shocking in nature, and customers are then not sure what to do in those situations. And so what Backblaze does is it's a fixed price point per terabyte per month as long as you know how much you're storing, you know how much it's costing you. It's also we provide free egress. And this is one of those places where the traditional cloud companies often just get you with their fees, these egregious egress fees, because they're often multiple times bigger than even the cost of the storage itself. And if you want to take your data and you want to send it even from one hyperscaler to another hyperscaler because you want to use their other GPUs, you're paying these egress fees. If you want to send it from a hyperscaler to a Neocloud, if you want to send it from a hyperscaler and send it back on-prem, you're paying these egregious egress fees over and over and over again. And so Backblaze makes egress free. And we hear that over and over again from companies that that is a tremendous benefit for them because they know, "I'm storing this amount of data, I can get it where I want it and use it how I want it, and I can do that quickly, and I don't end up with some surprise fees on the back end."
Rob Strechay
>> Yeah, I agree. And I think one of the features I also like about it is the fact that the API calls costs are out the window as well, which again, at a company I was building a SaaS platform for, that caught me by surprise was all the API fees associated with that and the egress, to put it mildly. Just a great approach to this.
Gleb Budman
>> What I'll also mention is that some companies, not all, but some companies also charge these delete penalties. And the delete penalties are something that often surprise people, because you have your data, you're storing your data, then you delete it, and they actually charge you even after you've deleted the files because they require you to keep the data for a certain period of time. And especially in AI workloads, but in most applications, there are some files that you're going to keep forever, there are some files you're going to keep for a month, and there are some files you're going to keep for a minute, and you don't want to be paying these surprise delete fees because of that as well.
Rob Strechay
>> Yeah, I think the folks at KubeCon will really understand that, given all the ephemeral workloads that they're dealing with in a lot of cases that may or may not be AI-oriented, but as people build agentic applications, it's more than just the gen AI side of it. There's traditional AI as in ML, as well as other applications that are plugged in, which brings me to a little bit more is, what are some of the key features that you see that Backblaze is offering that make a compelling solution to organizations as they become more enterprise ready with this?
Gleb Budman
>> Backblaze provides this enterprise-ready cloud storage, and just at core, that's something that customers need. You're running your Kubernetes cluster and you need your data somewhere. You don't want to store petabytes of data inside of each Kubernetes instance, so you really want that data source outside of it and plugging in to Kubernetes from the outside. That's just the core cloud storage offering. Now, we provide a very durable, high-performance, affordable version, and so that's an important part, but we provide a lot of enterprise-ready features. We have an enterprise web console. We have single sign-on. We have multi-region choices. We have ability to have lifecycle rules for compliance reasons. There's a lot of enterprise-ready features and functionality in the Backblaze platform. And what I would also say is we believe in an open cloud environment, and what I mean by that is your data should not be locked into just a single hyperscaler. You should be able to use your data and where you want it. And so we partner with and work with a broad ecosystem of compute providers, CDN companies, AI companies, the whole variety of organizations out there. And so there's an ecosystem built up around that you are empowered to use because you can get your data to wherever you want it to go. And so Backblaze is enterprise-ready and has a great ecosystem to support companies.
Rob Strechay
>> Yeah, I love that, because, again, having built SaaS before, I remember, for one customer, they had a use case where we were landing it on cloud storage du jour for them, and they're like, "Oh, yeah, but we're using BigQuery over in Google, and now we need to move the data from where we landed in AWS over to Google. How do we do that cost-effectively?" And I'm like, "Well, you really don't." I'm like, "You're going to pay for it and you're going to pay for it to Amazon on the egress. You'll get it into BigQuery for free, but that's a whole different kettle of fish."
I think that, to me, as people build these out and look to alternatives to those hyperscalers, this becomes key to the thinking. Have you seen that with the customers is really that, "Hey, we have these interesting use cases where we want to land the data once and use it many kind of things?"
Gleb Budman
>> And we actually, to that exact point about migrating the data from AWS to BigQuery and GCP, one of the things that we offer is a service called Universal Data Migration. And what Universal Data Migration allows customers to do is we will actually migrate the data from wherever it is, whether it's in Amazon or Google or Microsoft or on-prem or wherever you have the data, we will migrate the data to Backblaze. We will help subsidize the cost of the fees they charge you, those egregious secret fees, we will help subsidize those fees so that you can bring the data into Backblaze and then be able to use it wherever you want. And so, if you want to use your data inside of AWS with some services, inside of Google with some services, like BigQuery, inside of a Neocloud provider, you can do that and keep your data inside of Backblaze and then use it freely wherever you want. With Universal Data Migration, it makes it much easier for companies to do that.
Rob Strechay
>> Wow, I love that. I think that helps to unlock a lot of the value of that data at a better cost point for those customers, because I know customers, lot of times, they're making choices of do they keep the data or throw it out, to put it mildly. I think that has to help them.
Gleb Budman
>> And actually you make a really interesting point about that. A long time ago, it used to be that people would say, "Okay, I'm going to keep this data. I'm going to put this data in an archival maybe on tape, and then, at some point, I'm going to throw it out because I'm not going to need it." One of the things that we're seeing a lot more of is, because of AI, companies are realizing that that data has value and they need to be able to access it, it can't be stored in something that's inaccessible, and they really don't want to get rid of it because it is often valuable later in ways they didn't even think about. Enabling you to keep that data and keep it on something that is active so that they can access it for either training or running inferencing on the old data is something we're seeing more companies want to do.
Rob Strechay
>> Yep. We're seeing that too. I think that's a great place. What's up next for Backblaze and what has you excited for KubeCon North America 2025?
Gleb Budman
>> I will say, this is my first KubeCon, so I'm just excited to come and see what it's all about. One of the reasons why we're coming to KubeCon is it's where cloud native lives, and over the last 10 years, as you said, Kubernetes has taken over in a lot of the cloud-native use cases, especially around AI. We help AI companies, we help cloud-native companies, with all of their data and all their storage and AI-processing workflows. We want to go there, talk to the big picture of AI, but also just specifically talk to all these developers who are building cloud-native applications. And we're going to be located in Booth 1742. It's near the Startup Village area. In part, we want it to be there because we're helping lots of startups with all of their new and innovative AI use cases. It's also near the meal break area, so hopefully, in between meal breaks, people can come over and see Backblaze. One thing I'll mention is we also are giving away hot sauce. The Backblaze spicy sauce has been quite popular, and we now have a special limited edition one that's the Overdrive version of that hot sauce. Hopefully, people will come, and I'll look forward to meeting some of them.
Rob Strechay
>> Yeah, I think you'll get a lot. I also think that, I love the Linux Foundation and CNCF and everything like that, but you can never have enough spice on the food that you get at any venue. I would say that you're going to be very popular with that. If for nothing else, go there and get some hot sauce to go with your hot storage that you got going on back there. Really appreciate you coming on board, Gleb.
Gleb Budman
>> Rob, really good to chat with you as well, and we'll look forward to seeing you and everybody else there.
Rob Strechay
>> Yep. See you soon in Atlanta.
Gleb Budman
>> Thank you.
Rob Strechay
>> And thank you for joining today on theCUBE, our preview of KubeCon CloudNativeCon 2025. We're going to be live in Atlanta really shortly and thanks for joining us on theCUBE, the leader in analysis and news. Take care.
Rob Strechay