If you recall about 6-7 years ago there was a huge hype of cloud. Everybody was excited by it and not many people knew what it was …same as big data today!! Over the period, some of the corporate applications like CRM and other lightweight applications did move to cloud but nothing more. Cloud was looked at something that cannot deliver the heavy lifting and performance of large scale applications like ERP, Data Warehousing, BI, Analytics etc.
Organizations found a smarter way of managing the H/W infrastructure and they started outsourcing their data center operations to specialized players. All this is about to change with the emergence of cloud for all type of large applications like Data warehousing, ERP, Analytics, BI etc. If you check out Amazon Web Services marketplace, you can purchase SAP on a per/hour basis!!! Similarly you can buy storage and DW platform on per TB basis and pay only for actual usage.
Never before such commoditization of software and hardware has happened in the past. It has huge implications (mostly negative) for the vendors who supply these products. Recently both SAP and IBM earnings came under huge pressure due to fast adoption of cloud.
The key question is that why cloud computing is gaining importance and even feasible now. Before getting into that, here are latest piece of information on cloud and the major players who provide cloud for BI and Analytics and the recent news of new investments-
- Amazon Web Services (AWS) clearly has the largest cloud. It is said that that their cloud infrastructure is five times the combined infrastructure of all other players. They are now planning to setup a public cloud in India by next year.
- Microsoft Azure has a good infrastructure that can be leverages for BI and analytics
- IBM to invest $1bn in cloud infrastructure including one in India. They also bought Softlayer who was a prominent player in this space.
- Google has setup a large cloud infrastructure
- Cisco is investing $1bn in cloud infrastructure
I believe the emergence of cloud has happened due to technical advancements at multiple level of networking, data processing, data storage, security systems etc.
- Obviously the internet bandwidth is increasing on daily basis. The bandwidth cost is declining at about 20% per annum and the total bandwidth available if growing at about 32% per year. Having a good internet bandwidth is a very important factor in getting performance from any cloud infrastructure.
- The computing power of systems is still growing as per the Moore’s law!! Higher computing power makes cloud perform as per the requirements.
- Reduction in memory prices has brought in-memory analytics to the fore. In-memory analytics has huge impact analytical performance as the query response is no more constrained by the database performance – which has been stronghold of players like Teradata.
- Growth of data now is huge. This gives the cloud players economies of scale to invest. Since setting up of cloud infrastructure requires massive financial outlay, the usage has to be high to give the required ROI.
There is no doubt that cloud will encompass all areas of IT requirements and all type of applications, even the large ones like Analytics, data warehousing will move to cloud. An Australian bank has moved their entire Data warehousing and analytics on Amazon Web Services (AWS) and reduced their datacentre by 75%.
The adoption of cloud for analytics is growing at a fast pace and it does make sense for any organization to explore the possibility of moving to cloud. For SMEs, cloud has brought level playing field with large players in the areas like analytics that require high investments.
Try deep learning using MATLAB