March 12, 2019Which Cloud Data Warehouse Should I Migrate To: Redshift vs. Snowflake vs. BigQuery
Microsoft’s recent announcements at Microsoft Ignite should be resonating with global enterprise companies whether they are pure Microsoft client or not. Microsoft is taking a very bold swipe at Google (GCP/GBQ) and Amazon (AWS/Redshift) during thier Synapse (next gen Azure SQL/DW) demo. Microsoft went so far as to stress test Synapse on stage against GBQ and Redshift. Results were impressive, but interestingly, Microsoft did not run against Snowflake.
In speaking with AtScale’s customer base, we hear that they like Azure’s integration points between the products offered. Specifically, we’ve heard that AWS has a better portfolio, but nothing worked seamlessly together. Yesterday at Ignite, this was a core message. Synapse is meant to be the one-stop-shop for collection of data, data prep, data munging and query. Of course also boasting deep integration with PowerBI.
During the Synapse demo, there was a lot of language that will make the enterprise think twice. For example, Microsoft used terminology like:
- “Peta-byte scale analytics”
- “PowerBI Semantic Model”
- “Query relational and non-relational data”
- “Intelligent multi-level caching”
- “In-memory hierarchies”
- “Query schedules are workload aware”
Sounds familiar, right? First of all, let’s congratulate Microsoft. Well done! Synapse is decoupling compute and storage, much like Snowflake. Microsoft recognizes that scaled analytics actually does require a new architecture, and can’t be stand-alone servers (SSAS and AAS). The business requires easy access to all their data at scale. We’re in agreement here.
So what’s the problem? The problem is what Synapse does not solve for. Our customers have told us the following is important to them, and AtScale can help solve these problems and definitely addresses this paradigm.
Hybrid Cloud Initiatives
There are a lot of private companies and government agencies that are making the conscious decision not to bet on one cloud provider. In fact, Microsoft recognizes this at Ignite and announced Azure Arc. Arc intends to be an Azure console where you manage on-prem, Azure, GCP, AWS, etc. But this is from an infrastructure perspective only. It does not offer cross platform query virtualization the likes that AtScale offers.
- Leave Data Where It Is
Microsoft proclaimed at Ignite that ‘Synapse can access data wherever it is’. What they mean is anywhere in Azure (i.e., Azure Datalake, SQL/DW, etc.). AtScale’s Intelligent Data Virtualization supports the access of data regardless of cloud provider, on-prem, vendor, etc. Our Autonomous Data Engineering reads and writes for optimal query experience at scale and adheres to data regulations.
- BI Ecosystem
Microsoft is going ‘full hog’ on PowerBI. But most enterprise organizations also need support of PowerBI along with tools like Tableau, Excel, etc. The BI market is one of the largest and growing industries. End users will not abandon Tableau anytime soon. Adopting Synapse is great, but at the risk of disrupting the business from actually making insights? AtScale offers the business the use of their BI tool of choice, while maintaining one consistent view of the data.
- Manual Data Engineering Is Brutal
Again Microsoft is referring to cache, aggregates, in-memory hierarchies. How much of this is automated? How much of this is generated off user behavior from BI tool of choice? With ease of data access, there is a real risk of the proliferation of purpose built data marts, which then have a life of their own and have to be maintained forever. AtScale’s Autonomous Data Engineering relieves this. AtScale will generate acceleration structures on-demand, optimize (structurally, locality, partitioning), manage (lifecycle, updates, versioning) and intelligently engage the right acceleration structure(s) for the query (query planning, node graph).
- Target Audience
These conferences tend to focus on the pure stack clients. They also tend to focus on a developer community that cherish the ease of interopability. And while these are important, AtScale benefits the pure stack, hybrid stack (best of breed) and business consumer. AtScale makes advanced analytics possible no matter where the data is, how complex the underlying systems/business logic is, or who you are.
In short, the event and announcements at Ignite takes aim at clients who are looking to put their hand on the Microsoft bible and go all in. What the trends and market tell us in the past 20 years (and especially in the past 24 months) is that competition is healthy and desired. Our customers need technologies like AtScale that provide governance, security and performance – all while insulating them from the constant evolution and platform warring. This enables our customers to provide the business an easy, innovative and business approach to analytics while hedging their bets on the vendor gamble.