All-Flash Data Centers: Turning Your Data Green


    Aug 30, 2021

    The all-flash data centre (DC) is the next leap forward in technical innovation that will accelerate digital transformation (DX). “Great,” you might say to yourself, “But what does that mean for me?” In this series of blog posts, I’ll explore the all-flash DC, helping you make an informed decision about whether it’s right for your enterprise.

    Spread across six posts, I’ll cover things like

    • Digital transformation
    • What data means for your enterprise
    • The advantages of the all-flash DC
    • Challenges you might face when upgrading your IT infrastructure
    • Best practices for building an all-flash DC.

    To kick off the series, this post will focus mainly on the “all-flash” part of the all-flash DC and how it can turn your data green.

    So, what is an all-flash DC? A DC is considered “all-flash” if it provides 90% or more of its storage capacity via high-density, high-reliability, low-latency, and energy-efficient solid state disks (SSDs).

    SSDs — aka flash media — far surpass hard disk drives (HDDs) in terms of performance, offering ~100-times faster data access, 100-times higher throughput, and more than 1,000-times higher per-disk IOPS than HDDs — aka magnetic media.

    Find out more about SSDs: The Future of SSD (Podcast)

    In the past, enterprises have often been concerned about the service life of SSDs — after all, ensuring data reliability is crucial. Unlike HDDs, an SSD’s service life is determined by the number of write and delete operations of the NAND flash storage. But thanks to advances in write algorithms, the service life of SSDs has been extended significantly at the software level, so it won’t become a bottleneck in terms of the usage and replacement cycle of SSDs.

    Another major benefit of SSDs is their lack of mechanical moving parts, eliminating the potential for mechanical failures. Compared with HDDs, enterprise-level SSDs are more resistant to bumps and collisions, meaning that all-flash DCs have higher stability and are less likely to encounter accidental data loss.

    Currently, SSD price is the key factor slowing down wide-scale adoption of all-flash DCs. But as the price gap between SSDs and HDDs continues to narrow, all-flash DCs will become the norm. The price per GB of SSDs decreased 25% annually on average from 2015 to 2020, and this trend is expected to continue over the next five years. By 2025, the price of enterprise-class SSDs will be lower than that of 2.5-inch 10K HDDs. As SSD prices continue to drop, all-flash DCs will not only be more energy efficient, but also give you more bang for your buck.

    Ready to Swap to SSDs?

    Given the benefits of SSDs and the falling price per GB, is it enough simply to throw out your old HDDs and replace them with SSDs?

    You could do, but that wouldn’t tap into the full potential of all-flash DCs. Building an all-flash DC involves more than just upgrading the media; it involves integrating data, networking, and compute resources to build an innovative IT infrastructure architecture.

    In terms of data, your ability to unlock its value will determine how far you can digitally transform your enterprise. DX is the integration of digital technologies into all aspects of your enterprise, fundamentally changing how you operate and deliver value to your customers. IDC expects global spending related to enterprise DX to reach US$2.4 trillion by 2024, so it’s not something to be sneezed at. Data is a key driving force behind sustainable digital development and a fundamental strategic resource for your enterprises.

    In terms of networking, storage IP networks have become more vital as service volumes increase. To maximise the performance potential of SSDs, we need to adopt the faster non-volatile memory express (NVMe) protocol instead of the conventional small computer system interface (SCSI) protocol. And as we adopt faster media and protocols, we need faster networks. This is where NVMe over Fabrics (NoF) comes in. Storage networks running NoF provide higher throughput and lower latency, allowing us to move data to where it’s needed much faster.

    To fully realise the potential of SSDs in all-flash DCs, you should consider innovations in storage networks besides the existing fibre channel (FC) networks. NoF is the trend and comes in two mainstream forms: FC-NVMe and NVMe over RoCE. The former is typically adopted because FC delivers certain advantages in terms of stability. However, the latter delivers advantages in terms of a more open Ethernet and rapid improvement of bandwidth speeds. NVMe over RoCE is considered a more promising option that can improve performance, reduce costs, and simplify management.

    In terms of compute resources, they’re essential to making sense of the huge volumes of data we hold. Compute resources refer to things such as CPUs and storage media and function as the brains and processing power needed to carry out computational tasks. By offloading your data processing and storage requirements to DCs, you can scale flexibly as your requirements change. And the enhancement of an all-flash DC helps you reduce costs and increase efficiency while achieving the ultimate goals of being green and saving energy.

    So an all-flash DC is more than just a DC with SSD-based storage. It encompasses data, networking, and compute resources — it is an innovative IT infrastructure architecture.

    Ready to Go Green?

    Although DCs support the development of the digital economy, they also consume vast amounts of energy, especially as data volumes and applications continue to surge. Their energy consumption hinders sustainable digital development, increases maintenance costs, and counters efforts of green development worldwide.

    To reduce greenhouse gas emissions, numerous countries have signed up to the Paris Agreement and many have adopted low-carbon DC policies. For example, the US has proposed the Data Center Optimization Initiative (DCOI) policy, while the European Commission has encouraged the construction of climate-friendly, energy-efficient, and sustainable DCs. Reducing the energy consumption of DCs has become a hot topic.

    Driven by various policies and the demand for higher efficiency at lower costs, it’s likely that your enterprise will implement energy-saving practices throughout its production and operation processes. By adopting all-flash DCs, you can significantly reduce your carbon footprint and energy expenses, enabling not only sustainable digital development, but also higher resource utilisation and cost efficiency.

    All-flash DCs offer a smart choice to adapt to the digital future. Whether you want to take advantage of higher performance and capacity, simplify your infrastructure to reduce energy and space needs, or streamline your management operations, you need to carefully plan how you’ll build your future storage infrastructure.

    If you’re looking to update your DC, it’s a good idea to also consolidate workloads (i.e., leveraging the increased infrastructure density of newer systems). The fewer platforms you buy, use, and manage, the more cost-effective it will be. By adopting an all-flash DC, bringing all your storage under one roof, you’ll be able to handle more intensive workloads with higher per-unit performance and reliability.

    In my next post, I’ll be diving a little deeper into DX and how the all-flash DC can help speed up your DX journey. Check back soon for part 2.

    This post is adapted from the Huawei-sponsored IDC White Paper Moving Towards an All-Flash Data Center Era to Accelerate Digital Transformation.

    Read More:

    Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.


      Leave a Comment

      Posted in


      Posted in