As IT professionals know, figuring out the true cost of cloud computing is difficult at best, maddening at worst. There are so many variables in CPUs, memory, storage, and network bandwidth that figuring out what your proposed workload will cost in the cloud can be an exercise in futility.
And, to compound the problem, many costs may not be apparent at the start but cause sleepless nights once manifested.
It's an open secret in cloud computing circles that getting data into a given cloud is cheap—in fact, it is usually free. But getting data out is a completely different story. Traditional cloud vendors charge hefty premiums on data flowing out—or what they call data egress.
So how can we take the mystery out of what a given cloud implementation will cost when it's actually used by the business? How about a cloud calculator that assesses the price of resources you need to meet your requirements before actual deployment?
Meet the Oracle Cloud Workload Cost Estimator, which Oracle launched this week to help customers assess Oracle Cloud Infrastructure and Amazon Web Services with an apples-to-apples comparison (or as close to that as you can get). We think you will be surprised by the results, which show Oracle is a more cost-efficient (cheaper) option for many high-performance applications.
The calculator prices out not only cost of computing and storage, but also IOPs (data input/output per second), and data transmission out of the cloud. That last factor, known as data egress, is often a wild card. That's because traditional cloud companies charge a markup after a given amount of data flows out. So once you hit that magic monthly number—1 GB for AWS—data egress charges kick in. At Oracle, the meter doesn't start running until after 10,000 times more data egress – or 10 TB—per month. Furthermore, the unit price for additional data is much lower than that charged by our friends in the Pacific Northwest.
These networking charges are not chump change. And if your company is streaming video (a very data-intensive application) out to thousands of users every day, your video transmission gets pretty pricey pretty fast.
Take 8x8, the video streaming company, as an example. When it ramped up its capabilities to meet soaring demand when the COVID-19 pandemic led to remote working mandates, data egress usage quickly surpassed 1.5 PBs of outbound data per day.
Using the handy Oracle Cloud Workload Estimator, 1.5 PB (or 1,500 TB) of "data out" would cost just over $80,000 using AWS ($80,691 for the nitpickers among us) compared to roughly $13,000 ($12,969) on Oracle Cloud Infrastructure.
See what I mean?
Or, if you have a database or other application that requires very fast data flow from CPU into and out of storage, the workload estimator includes an IOPS module that lets you explore how Oracle Cloud Infrastructure stacks up against AWS in this dimension as well.
So, if you are dealing with 10 TB of block storage, AWS logs in at $47,872 per month compared to $609 per month for Oracle Cloud Infrastructure. That means Oracle charges 97 percent less for IOPS than the supposed cost leader.
Use the calculator to design your dream implementation (don't forget to include important workloads) from the ground up and then run your own comparisons. You will find that Oracle Cloud offers the best price performance for many of the most common cloud computing scenarios. We think even AWS stalwarts will be surprised at how much more they will get for their buck with Oracle Cloud.
