Podcast #130: The Latest from Pure Storage – ’24 //Accelerate Highlights

Brian invited Peter Skovrup, VP of Product Management for Pure Storage, to continue their conversation from the Pure //Accelerate Las Vegas event in June. Brian and Jordan spent a few days with Pure, learning the latest for what’s new with Pure Storage technology – like their 150TB flash module!

This is another live event where the audience gets to ask Brian’s guests questions as they arise throughout the conversation. In case you missed it, our first two live events had a few hiccups, but we seem to have ironed those out now. This jam-packed session covers hardware and software, the all-important NPS scores, flash technology, and how it all fits together with Pure Storage solutions.

This is an exciting conversation that also tends to be a learning experience. If you have the time, it is worth listening to this entire podcast. However, if you are short of time or need to find details on a particular topic, we have created an easy way to jump around the podcast to get to the issues that are most interesting to you.

The podcast timestamps have been broken down into five-minute increments to make finding what you are looking for easier.

00:00  Introduction

• Brian introduces Peter and talks a bit about Pure //Accelerate
• Advancements in Pure Storage, flash modules, AI integrations, and Evergreen enhancements
• The importance of NPS

06:00  Core Products

• Where Peter is focused
  • FlashArray for Scale-up
  • FlashBlade for Scale-out
  • Operate as a unified storage system
• System Architecture
• DirectFlash Module (DFM)

10:00  Why DFM is important to Pure

• Personal anecdotes about football
• What is it
• What does it do
• Global management for flash storage
• Online Community

15:00  Energy Efficiency

• DFM benefits
• Reducing flash power consumption at DFM level two
• Simplifying design by limiting DRAM
• Centralizing processing
• DRAM becomes part of the controllers
• Increasing density and efficiency in storage systems

21:00  Questions from the Live Audience

• What are the speed trade-offs?
• Challenges managing increased data storage
• Performance impacts
• Pure’s approach to non-disruptive updates
• Business aspects for non-disruptive operations

26:00  Pure Storage Evergreen//One

• Clarification on how Evergreen programs work for customers
• Discussion on the financial and operational benefits of continuous upgrades
• Explanation of subscription models and their flexibility
• Benefits of Evergreen One and subscription-based storage solutions
• Focus on delivering performance SLAs and scalability

30:00  Back to the importance of NPS

• Traditional hardware investment model vs. Evergreen program
• Emphasis on the simplicity and cost-effectiveness of Pure Storage’s upgrade process
• Financial planning benefits of Evergreen programs
• Importance of the clarity and visibility of costs
• Non-disruptive upgrades and flexibility for customers

36:00  Introduction of Pure One

• Introduction of Pure One
• Data visibility features
• Simplifying storage management
• Management-as-a-service capabilities
• Copilot
• Workload planner, anomaly detection, and security assessment
• Benefits of Pure One for operating and optimizing storage environments at scale

42:00  Finally, mentioning AI

• AI workloads and the importance of data support
• Nvidia’s presence at the event
• Collaboration with Nvidia and the certification of Superpod

45:00  SuperPod at scale

• Importance of 400 gig Ethernet for AI workloads
• Flashblade’s scalability and performance for AI and HPC environments
• Hardware requirements for AI workloads and the advantages of Flashblade
• Flashblade’s architecture and its benefits for AI and unstructured data
• Flashblade versus other parallel file systems
• Importance of metadata operations in high-performance file storage

51:00  Data Management

• Challenges of managing data for AI
• Importance of a comprehensive storage solution
• SLA-based solutions for AI workloads
• Long-term investments in AI
• Guiding customers
• Pure Storage’s technology for large AI deployments

58:00  Managing data for AI

• Pure Storage scalability
• Small deployments
• Pure Fusion
• Edge and Distributed deployments
• Non-disruptive operations
• Optimization and Utilization
• Evergreen programs

62:00  Wrapping up

• Brian and Peter begin wrapping up the discussion
• Let’s talk real football—The EUROs