WHAT’S HOT the analyst
microdatacenters, optimized for location and work, while others will be hyperscale or general-purpose. The network will become the computer.
9. Open Compute, standardization and buyer control.
Hyperscale datacentre managers are increasingly designing or commissioning their own computing hardware, buying from ODM (original design manufacturer) companies in developing economies, and embracing alternative datacentre architectures suited to their needs. In this way, they do not have to pay for the full branded capabilities and services that leading brand suppliers offer; some are doing the same for facilities equipment. In effect, the buyers have become powerful enough to dictate to the vendors.
The emergence of the Open Compute Project, with its open architectures and expertise shared by customers, will spread this practice out further. At the same time, more operators are using open source
software tools (mostly enterprise versions of open source tools supported by leading vendors). By buying and managing these systems and software in an open and unified way, some datacentre operators and service providers will take out a further layer of costs.
10. Wild cards will produce step changes. While physical datacentres have not developed or changed as dramatically, changes in IT (virtualization being one example) have produced some dramatic changes in the performance of datacentres. It is never easy to predict the impact of such technologies, but the pattern of innovation in the past several decades suggests that some dramatic improvements are likely in the future. Some of these are discussed in part one of this two-part report.
These include silicon photonics, which could enable some dramatic changes in datacentre density, cooling and latency; all-flash memory datacentres; and memristors, which could revolutionize storage. The introduction of silicon alternatives, such as graphene, into
commercial electronics could also enable Moore’s Law to continue to advance on its trajectory.
The 10 trends discussed here are not the only ones to affect datacentre design, operation or economics, nor are they set in stone. Although all have been carefully considered and are supported by research, they are presented here as discussion points. Some may argue for different or additional developments – and the degree of adoption and the relative impact of a number of these changes are unclear and open to debate.
The wider issue is not about the specifics of any one trend, but, as we stated in the first part of this report, any organization investing in or operating a datacentre, or using datacentre services, should now be continually assessing the evolution of emerging technologies and the competitiveness of the infrastructure it is using or is dependent on. The delta between the underperforming datacentre and the best in class is widening.
The emergence of the Open Compute Project, with its open architectures and expertise shared by customers, will spread this practice out further. At the same time, more operators are using open source software tools (mostly enterprise versions of open source tools supported by leading vendors)
10
www.dcseurope.info I Summer 2014
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56