Google Cloud Next is over, long live the new way to Cloud
It was a (very!) busy week to say the least at Google Cloud Next in San Francisco.
After the last three online editions, it was a long awaited return to an in-person conference. For having attended Next since 2017 it definitely had a different feeling to it this time. I’ll try to break down the experience in the following paragraphs.
General comments
I said this to multiple people during the week, Google Cloud Next now feels like a “big tech conference”, and I did not necessarily mean this in a bad way. Billboard on highway 101, digital billboards on taxis and cars, benches advertisement, Next was definitely all over the city.
The conference felt very active throughout the whole week, with lots and lots of people roaming around the large expo area, sessions in multiple locations at the same time, an almost fully independent Leader Circle program, a dedicated Partner Lounge within the conference center and a plethora of side events organized by members of the Google Cloud ecosystem.
The positive side is that it means the adoption, maturity and momentum for Google Cloud is there, driving all that attendance and bustling. The negative is that you do get this overwhelming feeling that any big tech conference brings with itself and you miss the good old days where you knew everybody :). My take on this observation is that it is ultimately a good thing for Google Cloud. It is the natural evolution for such a platform and it translates the growth, market position and maturity level it has reached.
For having been in the Google Cloud ecosystem for about 7 years, I was impressed how the ecosystem has grown. The focus on the Enterprise market from a couple years ago is definitely felt with a myriad of “traditional” enterprise infrastructure players entering the ecosystem. Still, Google always had a vested interest towards startups and founders, which still transpires, especially propelled with the wave around Artificial Intelligence. Google has been quite vocal about the market shares it has acquired with top (Generative) AI companies on the market, and they doubled down through the conference on the AI topic. The keynotes were centered around that technology applied to the portfolio of services and solutions available on the platform, and not the other way around with each product being presented as its own category (infrastructure, application development, databases, etc…).
(Main) Announcements
As you probably heard, except if you’ve been out of the tech world this year, there was a lot of …. AI, AI and more AI sprinkled across almost every product category.
The wrap up link is what you want to check out to see the exhaustive list of the 161 announcements from Next
I’ll try my best to highlight the main announcements :
- Duet AI — your personal copilot for Google Cloud or Workspace.
This was a very large announcement and full unveiling of the product as it applies across a lot of different services on the platform and leverages a lot of technologies in the backend - GA for Vertex AI Search and Conversation — Renaming from GenAI App Builder and now GA
- Reinforcement Learning with Human Feedback (RLHF) in public preview
- New 3rd party models available in Model Garden — Llama 2, Code Llama, Claude 2, Falcon, and others
- Vertex AI Extensions and data connectors
- Colab Enterprise
- BigQuery Studio — a collaborative analytics workspace for BigQuery
- Vertex AI foundation models accessible directly from BigQuery
- AlloyDB AI - a vector DB option with the benefits of AlloyDB
- Cloud Spanner Data Boost - an interesting feature to analyze transactional data separately from the transactional workload.
- New instance types and hardware : Nvidia H100 A3 VMs, TPU v5e, Arm-based C3A, AMD-based C3D
- TPU support in GKE — Very interesting capabality for people using GKE for large scale ML training or inferences, now getting access to TPUs for the node-pool configurations
- Titanium - an architecture for system offloads benefiting storage, compute and networking capabilities.
- L7 ILB Global Access — this one is a straightforward but very effective feature that has been awaited for some time.
- GKE entreprise — Includes a set of features very useful for large scale computational jobs on GKE among other things
- S3C - Secure DevOps solution with GitLab
- Environment APIs : solar, air quality and pollen
Conclusion
More than the technology and the frenetic pace of announcements, a conference like Next is first and foremost a great time to connect with people, get to catch up with colleagues, partners, and customers alike.
It was a great opportunity for the Zencore team to get together and enjoy some time as a team. We made the most out of it to create lasting bonds internally and externally with the people we work with every day. We hosted several events around the conference, a private reception as well as a large happy hour, both turned out great and we all left with new connections made, new learnings, new perspectives, new partnerships and with a continuous desire to make companies successful on Google Cloud.
Thank you again to everyone who made the trip to San Francisco and that I had the chance to connect with ! See you Next’ time (Literally :) …) !
