Backblaze

02/06/2026

One vendor. One console. One contract.

Sounds simple, right?

Until your workloads change, AI tooling evolves, costs climb, and that convenient platform starts feeling more like a constraint.

Our latest blog looks at why modular infrastructure is winning again—and why best-of-breed architectures give teams more flexibility, better cost control, and the freedom to put each workload where it performs best.

Because the companies that win won’t be the ones locked into every layer of a single stack. They’ll be the ones that can build, adapt, and scale without constraint.

Read the blog: https://hubs.ly/Q04jgssb0

29/05/2026

Ransomware doesn’t usually succeed because of one catastrophic failure. It succeeds because systems were designed for scale—not for adversaries.

In this piece, Kari Wilson breaks down what it means to create resilient systems aren’t built on assumptions of safety. For IT and security leaders, this isn’t theory. It’s design strategy.

Read the full article 🔥➡️
https://hubs.ly/Q04jbjT-0

27/05/2026

The hidden killer of GenAI margins isn't the GPU bill. It’s data egress.

Production changes the game. Teams launch models and face massive, punitive penalties just for moving data across clouds. B2 Overdrive eliminates the infrastructure bottleneck. https://hubs.ly/Q04j4g1W0

📹: Senior Director of Solutions Engineering Troy Liljedahl speaking at AI & Big Data Expo in San Jose last week.

19/05/2026

GPUs may be the engine of AI infrastructure, but even the fastest engine needs fuel.

For AI clouds, that “fuel” is data—and if the storage layer can’t move it quickly and consistently, expensive GPUs can sit idle.

Our latest blog breaks down why AI platforms need more than GPU availability to win. They need a data supply architecture built for sustained throughput, predictable performance, and production-scale workloads.

That’s where Backblaze B2 Neo comes in.

Read more: https://hubs.ly/Q04hdhKT0

19/05/2026

Day 1 of AI & Big Data Expo is a wrap! Our team is having a blast in San Jose talking to attendees about how you can hoard all your data and (actually use it) with your existing compute in the cloud or on prem. Come see us at Booth #434.

18/05/2026

Backblaze Drive Stats helped put us on the map. Now the conversation is shifting toward AI infrastructure, performance, and economics. Appreciate the thoughtful analysis from Alastair Cooke on where the market is headed.

https://hubs.ly/Q04h0JXw0

28/04/2026

New data alert: AI workloads don’t look like traditional internet traffic.

How? They move in high-magnitude bursts between a small set of endpoints—and they’re reshaping how networks get built.

Where? Check out our geographic heatmaps to see the trends.

The Q1 2026 Network Stats report is live 🔥➡️
https://hubs.ly/Q04dMxhm0

23/04/2026

AI workloads aren’t just bigger—they’re fundamentally different.

Sustained, high-volume “elephant flows” are moving petabytes between storage and compute for hours (or days) at a time. That’s a very different demand profile than traditional cloud traffic—and most infrastructure wasn’t built for it.

So what does AI-grade infrastructure actually require? We break down what we’re seeing across real-world network telemetry—and what it means for how teams should be designing infrastructure now. 🔥➡️
https://hubs.ly/Q04dg3Kp0

11/04/2026

How does S3 compatible object storage compete with VAST, WEKA, and DDN?

The short answer: It doesn't—they're complementary.

A scalable AI data pipeline needs both: high-throughput object storage as your always-ready data layer, and the right compute-adjacent storage to feed GPUs at speed.

https://hubs.ly/Q04bdcR80

Missed GTC? Watch out full recap session here: https://hubs.ly/Q04bdBHB0

The Backblaze team is heading to Nvidia GTC to get the latest updates on GenAI, agentic AI workflows, and emerging AI capabilities. Learn how increasing reli...

09/04/2026

OpenClaw hit 250K ⭐ in 60 days.

But the bigger story? The data problem no one’s talking about. Autonomous agents generate nonstop outputs—reports, memory, audit logs.

Without durable storage, it’s all dark data.

If you’re building with NemoClaw:
→ Decouple storage from runtime
→ Automate backups
→ Track lineage from day one

That’s how agent systems actually scale. Read the full article 🔥➡️
https://hubs.ly/Q04bb_hx0