Key Points:
- Among the big players,
Meta
,
微软
,
谷歌
, and
亚马逊
as well as
甲骨文
and
Akamai Technologies
have been very active in the early months of 2024, with big plans for the months and years ahead.
-
Synergy Research Group
has forecast that over the next six years, the average capacity of hyperscale data centers will be more than double that of current operational hyperscale facilities.
-
Steven Dickens
, VP and Practice Leader, at
The Futurum Group
, sees AI as a defining characteristic for hyperscale data center operations in 2024 which typically includes large-scale deployment of? Graphics Processing Units (GPUs). He noted that power-hungry and heat-producing GPUs will place unique challenges on the already overburdened power and cooling systems of many data centers.
- In the story above,
Dell'Oro Group
analyst
Baron Fung
has identified four key reasons why hyperscale cloud and colocation service providers are in a race to build newer and more facilities globally.
You already know that every day at Data Center Knowledge brings advice, trends and strategies for data center professionals on how to design, build, and manage world-class data centers.
That means original reporting from our team of journalists and unique commentary you won’t see anywhere else! But in case you missed them, here are some of our other must-read favorites from this week:
Sustainability is in the Air
Key Points:
- For starters, wind is considerably more efficient than the other major source of renewable data center energy: solar. You'd need to install nearly 50,000 solar panels to produce the same amount of energy emitted by a single standard wind turbine.
- On top of this, because wind can produce electricity, it provides an all-purpose energy source that can power IT equipment as well as cooling systems. This makes wind advantageous compared to geothermal energy, which is typically only used by data centers for cooling purposes.
- Perhaps the most obvious drawback of wind energy is that wind is intermittent and unpredictable. That's one of the reasons why wind and other renewables (like solar, which is also not consistently available) are often not sufficient for powering data centers totally on their own.
- On balance, it's worth noting that smaller-scale data center operators don't tend to be major investors in wind power – perhaps because they don't face as much pressure to make public commitments to renewable energy, and perhaps also because they lack the financial resources of large tech companies, which can afford to sign agreements to source energy using wind farms even if solar or non-renewable energy sources make more financial sense.
The Great Debate: Defining "On-Premise"
Key Points:
- From television interviews to major conference stages, CIOs, CTOs, and CEOs have long been using ‘on-premise’ as an established term when discussing digital transformation, data workflows, and where data resides in hybrid cloud architectures.
- Indeed, the phrase has become so common that it’s easy to overlook the fact that ‘on-premise’ is a grammatically inaccurate truncation of the term ‘on-premises.’
- In a?cloud?and data center context, ‘on-premises’ refers to a group of servers that a business privately owns and controls. Simply put, the on-premises model refers to servers, data, or an entire data center physically located inside your corporate building.?
- It’s easy to see why ‘on-prem’ became a common synonym of this term – the handy truncation coming in at a whole two syllables (50%) shorter. Somewhere along the way, however, the phrase ‘on-premise’ joined the conversation. And this theoretically should have meant something else entirely.
Will AI Supercharge Data Center Infrastructure?
Key Points:
- Data center infrastructure management, or DCIM, is the process of monitoring and managing all the equipment within a data center. It applies to IT equipment, like servers, as well as the complementary systems, like HVAC infrastructure, that data centers depend on.
- DCIM is important because even a small data center could contain tens of thousands of individual equipment components, and data center operators want to know when one of them fails or behaves suboptimally.
- Some of the DCIM strategies that are sometimes described as AI – specifically, those that involve analyzing data to identify patterns and anomalies – aren't all that new and are only arguably a form of AI. You could also contend that they're just forms of descriptive and predictive analytics, and that not all analytics is AI.
- Meanwhile, flashier use cases for AI in DCIM – like using generative AI to create guidance or bespoke configurations for data center infrastructure – are fun to talk about, but implementing them remains unrealistic at the present moment.
Optimize Your Data Center Spend
Key Points:
- Data center financing is the process of obtaining funding, usually in the form of one or more loans, to support the purchase, construction, or expansion of a data center. In addition, data center financing deals often help pay for the assets and equipment that businesses intend to install inside a data center facility.
- In short, financing a data center project is challenging because there is not a lot of standardization or consistency in the way lenders approach data center funding requests, or in the way projects play out.
- Talking to multiple lenders is especially important for data center financing because you may find that some lenders approach the overall project in fundamentally different ways than others.
- Most lenders would rather modify a major loan than have a project fail because, for instance, construction costs proved higher than expected and the original financing is no longer sufficient.
Attend the conference made by and for data center professionals. Data Center World has what you need to propel your career, solve challenges, and spend time with your peers. Super Early Bird rates are available now.
??? Use my code: BRANDON300 To receive $300 off of All Access, Standard, and BUILD passes at Data Center World! Learn more at the link above!
This is just a taste of what’s going on. If you want the whole scoop, then register for one of our email newsletters,?but only if you’re going to read it.?We want to improve the sustainability of editorial operations, so we don’t want to send you newsletters that are just going to sit there unopened. If you're a subscriber already, please make sure Mimecast and other inbox bouncers know that we’re cool and they should let us through.
Our bi-weekly LinkedIn newsletters arrive on Saturdays, so keep your eyes peeled for the top stories you may have missed between now and then.
Data Center Operations
9 个月The aaaaaaaaaaawaaaaaa