Teradata Cloud Offers Choices for Customers
by Ron Powell
Originally published March 7, 2019
Brian, Teradata recently launched Teradata Vantage. What does that mean for the cloud?
Brian Wood: It’s pretty exciting. Teradata Vantage is our new platform for bringing together the Teradata Database, the SQL engine for which we’ve been known for so many years, and the machine-learning and graph functionality of Aster. Part of Teradata’s strategy is to have the same software available everywhere.
In the cloud, that means Teradata Vantage is available on the AWS infrastructure, on the Microsoft Azure infrastructure and on what we call Teradata Cloud – which is our infrastructure in Teradata data centers. And, of course, Teradata Vantage can be deployed in customer data centers on the Teradata infrastructure. So it is the same software everywhere. And just as we’ve done in the past with the Teradata Database, we’re doing the same thing with Vantage – making it available across all these deployment options. That is exactly what customers want. They want the choice to be able to deploy the same software wherever they want and then move things around as they need.
What are some of the trends you’re seeing for the cloud?
Brian Wood: The cloud is very dynamic. It’s evolving rapidly. Particularly for our large enterprise customers, the cloud is becoming more and more central to the discussions we have with them. Some companies have what we call a cloud mandate. Sometimes that means augmenting what they have on premises with cloud infrastructure – in other words, they don’t want to expand their on-premises footprint with more data center infrastructure. Anything new needs to go in the cloud. That’s one approach.
Fewer customers have a cloud-only approach. They’re getting out of the data centers and moving everything to the cloud. That’s not typical, but we do have customers that are doing that.
And there’s also a trend of different types of users – for example, the data scientists – wanting access to Teradata and analytics. These are users that maybe were not part of the IT organization in the past with the traditional Teradata software-type access, but they’re wanting to apply new types of analytics using the data that’s in Teradata Database – now called Teradata Vantage. They are wanting cloud deployment because it’s easier and faster for them to spin up in that environment than to go through IT.
We’re having more and more discussions and questions about what Teradata offers in the cloud. And it goes back to the reasons I mentioned earlier:
So we’re seeing larger customers, more complex use cases, and a much higher interest level in cloud deployments.
Could you talk about cloud as a service?
Brian Wood: Great question. We at Teradata have two categories of cloud consumption: do it yourself, and as a service. We call the as-a-service offering Teradata IntellliCloud. That is our as-a-service offering for analytics at scale. So same software – I’ve said that a few times because that is a recurring theme. We provision the infrastructure, whether it’s AWS, Azure or Teradata infrastructure. We help the customer onboard in terms of moving their data into this environment. We manage that whole environment so the patching, the updating, the backing-up of data daily and the security monitoring. We’ve invested in compliance audits for things like GDPR, PCI, HIPAA and others. We maintain that entire environment so customers can focus on data and analytics and getting answers. We take care of the heavy lifting of the infrastructure so customers can focus on getting answers.
Could you comment on how your customers are benefiting from their cloud deployments?
Brian Wood: Absolutely. One of the key benefits and reasons why customers go to the cloud is much faster time to provisioning. Instead of literally taking months to get physical equipment that goes into the data centers, in the cloud environment they can spin up the same type of capability they have on premises in hours or even tens of minutes. That’s one of the advantages – the much faster provisioning.
Another benefit is the shift from capital expense to operational expense. A lot of companies want to account for their consumption of these resources using operational expense (OPEX). It’s a different pot of money than the capital expense, and a lot of times it’s a much easier approval process to pay for things out of the OPEX budget than to go through the much more formal and less frequent capital expense budgeting.
There’s the benefit of reducing the risk because they’re only paying for what they use and they don’t have to pre-purchase things in advance. They can just consume as they need it. They’re not leaving money on the table. They’re not over-buying and having under-utilized assets. They’re just paying for what they need as they need it, and they can start small and grow as needed.
Those are the main benefits: faster time to provisioning, the different way of accounting for it and reducing the risk. The fact that it’s the same software everywhere means that they don’t feel trapped into any one type of deployment. They can start in any type of deployment. They can start in the cloud and move on premises or the other direction.
Those are the main benefits. I will point out that some people think that moving to cloud will save them money. That’s not always the case. If they’re only using those resources a little bit and only paying for what they use, then they probably will save money. But, if the company is paying rental prices for something that’s always on, then they probably will not save money. When deciding what should move to the cloud, we always recommend customers make that decision based on the types of workloads and their consumption patterns.
Have you seen an acceleration of customers moving to the cloud or has that plateaued?
Brian Wood: We do see an acceleration. As you know, our customers are very large, and they tend to be somewhat conservative. They’re not just going to go in a different direction willy-nilly, but it’s becoming a bigger part of discussions. We always recommend start with a proof of concept, and then choose workloads that are not mission-critical. For example, that might be test/dev, an additional environment, data labs, discovery analytics or even disaster recovery. You want to go through your learning curve on a non-mission-critical application versus your production workload in case there happens to be a hiccup or failure. More and more, we’re seeing customers adding test/dev or discovery analytics to augment what they have on premises, and learning there. Then as they get comfortable they may decide to move more workloads into the cloud or not. Either way, they have the choice.
Brian, sounds like wise advice. As always, it’s been a pleasure to talk with you about what’s happening in the cloud.
Recent articles by Ron Powell
Copyright 2004 — 2020. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC