Google Cloud Next London 2017 04 May 2017
I’m writing this a bit after the event but think it’s worthwhile before it entirely slips my mind.
Google Cloud Next London was a two day event held at the Excel Centre in London. I’d just flown back to London from New Zealand the day before, after 8 months away, but felt relatively charged for this event.
The main comparison for me is the AWS Summit that I’ve attended at the same venue the last couple of years. Amazon host this for a single day and there’s always interesting talks to attend. Though longer at two days, Google’s event had plenty to keep you engaged. The itinerary, while managing to be quite varied, was dominated by Big Data and ML. These are the areas where Google probably has the biggest advantage over their Cloud rivals and the talks I was most interested in attending. I haven’t had the opportunity to really assess Google’s offerings for a while so it was eye opening to see what they had.
Their Big Data offerings were the most impressive to me. Big Query, Cloud Dataflow and Cloud Dataproc all tied together to make processing huge and disparate data sources relatively straight forward.
Cloud Dataprep especially made an impression, a service to visually view, clean and join data together.
The service offerings didn’t seem too different from what Azure, AWS and others offer. There’s a bit of an arm’s race between all these providers to offer services for voice, translation, image and video recognition and the like; they seem to be constantly jump frogging each other in terms of who offers the better service at any particular time.
I did like the open approach Google have taken with Tensor Flow and the ease at which you can run your own models in their Cloud infrastructure with Google Cloud Machine Learning Engine.
Yet another database offering, however a rather compelling one. A relational database that scales, Google Spanner has been used internally by Google for years and is now offered as a service.
The one talk where it felt that Google was extremely lacking in regards to AWS was about getting data into Google. The impression was that they have big plans in the pipeline that they couldn’t talk about them yet; what they did have on offer was very basic and miles begins what Amazon offer with Snowball.
Google Cloud Next London was a simulator event that did a good job of highlighting what Google has to offer. If working on a Big Data and/or Machine Learning centric solution then I think GCP would be the primary contender in my eyes.