Google I/O Thoughts

I am on my way back home from another Google I/O conference and I thought I will briefly write about my thoughts on what I heard and saw at the event. This is not a full fledged analysis and hence posting here instead of Rishidot Research blog. Feel free to jump in and add your thoughts.

  • Every year, developer excitement for Google platform and services is increasing. The announcements regarding Android tools, Google Compute Engine, Google Glass Mirror API and SDK announcement, etc. only added to their excitement. 
  • Google Cloud Platform announcement is very important. First, they are unifying their cloud portfolio (PaaS and IaaS) into one and, more importantly, the public availability of Google Compute Engine is an important step to bring the much needed competition to AWS in the public cloud side. As I wrote in October of last year, the performance you get with GCE is amazing. Especially, the network performance and the fact that fiber connection between their datacenters will allow applications designed for failure across multiple regions to offer seamless experience. Imagine having better performance with seamless HA all built on the applications. As Adrian Cockroft of Netflix mentioned in one of the conversations, Google has built the cloud for their internal use first before opening it up to others. So, it is not surprising that their cloud offers much better performance over competing public cloud providers.
  • Google Compute Engine also offers much higher value to developers than competing public cloud providers (including AWS) because Google has tons of data and APIs around their data. Developers can easily take advantage of these data without the data gravity problems.
  • The release of Android Studio and Google Compute Engine means that Android only developers will have additional value by hosting their apps on Google Compute Engine or Google App Engine.
  • Google Data Store is truly an important part of their announcement.
  • Google Cloud is gonna be a much superior option for MapReduce kinda jobs compared to other service providers. I had a chance to talk to some MapR folks and some of their users. They were raving about the performance they get with GCE.
  • But enterprise…… I didn’t get much hint on whether they understood the enterprise game. Unless they sort this out, it is gonna be a tough sell.

These are few of my thoughts. Feel free to add yours here.

Disclosure: Google gave away Chromebook Pixel to all attendees.

Comments

  1. Good summary Krish. As mentioned in my post (http://wp.me/p1xtvm-4R), I believe the enterprise doesn’t seem to be an immediate priority in terms of a product target compared to gaining developer buy-in to wider Google platforms. It seems Google’s betting on the developer economy.

Leave a Reply