The growth and demand to process unprecedented volumes of data is forcing organisations from all industries to reevaluate their computing capabilities, which is why we are seeing a bigger emphasis placed on the need for HPC environments.
Data centre facilities are ideally placed to benefit from this as they have the necessary infrastructure in place to support denser power and cooling requirements, but it is increasingly becoming the case that simply offering HPC capabilities is not enough to stand you apart from your competitors – the emphasis now is very much on how best to deliver it.
The typical adopters of HPC have changed so much in recent years and with this so have the requirements. No longer are users accepting a one-size-fits-all model – they want a bespoke offering which can be delivered fast, account for fluctuations in their data usage and also fit in line with efficiency targets. Because of this, data centre operators are very much under pressure to act – if they don’t their customers will simply move onto someone who can.
For those data centre operators there are a number of considerations which need to be addressed. Firstly, proofing your facility for HPC must not be at the expense of mid-range customers. Focusing an entire site on HPC will considerably limit your customer base and therefore it is imperative that any investment is still able to account for traditional mission critical applications, as well as increased customer demand for dense configurations – by doing so you will significantly enhance your ability to process fluctuating data volumes.
Another consideration focuses on cooling capabilities. Data centre providers are going to great lengths to minimise cooling costs for both them and their customers, but understandably HPC brings certain challenges. Some HPC providers have tried to ensure consistent cooling by deploying liquid cooling, larger fans or conductive cooling methods. Typically, however, the rate at which heat is produced is greater than the rate at which cooling accelerants or fans can dissipate this heat. It is therefore important for centres to be built with systems in place that far exceed the maximum cooling requirements, especially those embracing HPC.
Space is often seen as a major consideration, especially as the majority of facilities are located in or around major cities. Building on an existing site is a big capex expense and is often hindered by physical boundaries. Building an entirely new site with HPC capabilities can be a more viable option, but again this is a significant investment. In light of organisations carrying and processing unprecedented data volumes it’s imperative data centre providers are able to understand where current and future demand for HPC capabilities is likely to stem from – this will allow you to tailor you offering accordingly, therefore better servicing your customers.