MySQL HeatWave on OCI vs. Snowflake
Delivering 35X better price performance than Snowflake—according to a performance comparison with publicly available benchmarking code from Oracle, enabling anyone to run the benchmarks—HeatWave increases MySQL performance by orders of magnitude for analytics and mixed workloads. Oracle MySQL HeatWave is the only service that enables customers to run transactional and analytics workloads directly from their MySQL database.
MySQL HeatWave outperforms Snowflake in several areas:
- Speed: MySQL HeatWave is 6.8X faster than Snowflake.
- Cost reduction: MySQL HeatWave is one-fifth the cost of Snowflake.
- Single database for transactional and analytics workloads: MySQL HeatWave eliminates the complexity, risk, and cost of using two different databases.
- Automation powered by machine learning: MySQL AutoPilot uses advanced machine-learning techniques to automate HeatWave, making it easier to use and further improving performance and scalability.
- No changes required to existing applications to benefit from real-time analytics: Transactional and real-time analytics workloads run simultaneously in a single MySQL database, without any modification to existing applications.
- In-database machine learning: Customers can build, train, deploy, and explain machine learning models within MySQL HeatWave. No need to move data to a separate machine learning service.
1. MySQL HeatWave is 6.8X faster than Snowflake
HeatWave is an in-memory query accelerator architected for scalability and performance. MySQL HeatWave is faster than comparable cloud database services, including Snowflake—at a fraction of the cost. This has been demonstrated by multiple standard industry benchmarks such as TPC-H, TPC-DS, CH-benCHmark—and on real-world customer workloads. For a 10 TB TPC-H analytics workload, MySQL HeatWave is 6.8X faster than Snowflake.
See the performance details and learn more about the benchmark configuration
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Can customers achieve best performance with minimum cost?HeatWave implements state-of the-art algorithms for distributed in-memory analytic processing, which deliver very high query performance. With 10TB TPC-H data, MySQL HeatWave is 6.8X faster and one-fifth the cost of a Snowflake X-Large cluster.Snowflake users must increase the cluster size to improve performance but will incur additional costs, which escalate as they scale up due to very large cluster sizes. |
yes |
no |
"It's hard to look at what Oracle is doing with HeatWave and not be impressed. In many ways, HeatWave is the very essence of what the cloud is about – commodity infrastructure architected in a scale-out fashion, enabling the best performance and price-performance for a given workload."
—Matt Kimball, Senior Analyst, Data Center Compute, Moor Insights & Strategy
2. MySQL HeatWave is one-fifth the cost of Snowflake
MySQL HeatWave is one-fifth the cost of Snowflake and provides 35X better price/performance based on a 10 TB TPC-H data set.
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Does the database service enable customers to expand or downsize their cluster to any number of nodes nondisruptively?Customers can expand or downsize their HeatWave cluster to any number of nodes nondisruptively. During the resizing operation, HeatWave remains online for queries, DMLs, or load operations—there is no downtime. As part of the resizing, data is automatically downloaded from object storage, automatically repartitioned among all available cluster nodes, and becomes immediately available for queries.Snowflake lacks the equivalent real-time elasticity capability. Snowflake provides compute resources only in building blocks of 1, 2, 4, 8, 16, 32, 64, and 128 nodes. Customers have no option but to overprovision their deployment by choosing a much larger size than needed, spending more money than necessary. |
yes |
no |
"Not investing in MySQL HeatWave and instead spending twice the amount of money on other slower database cloud services such as Snowflake makes about as much sense as spending more to take a flight that gets you from London to New York in 6 days with 15 connections rather than a direct route that takes 8 hours."
Futurum Research
3. Single database for transactional and analytics workloads
Eliminate the risk, cost, and complexity of using two different databases. MySQL HeatWave is the only service that enables database administrators and application developers to run transactional and analytics workloads directly from their MySQL database.
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Can customers run transactional and analytics workloads using a single database service?MySQL HeatWave is the only service that enables customers to run both OLTP and OLAP workloads in MySQL without the need to extract, transform, and load (ETL) data to a different database for analytic processing. No changes to existing applications are required.Snowflake is designed for OLAP workloads only. Customers can't run OLTP workloads directly on Snowflake; they require a different database and an ETL process to load data into Snowflake for OLAP processing (Snowflake’s Unistore is only in preview). This increases complexity and costs. Security and compliance risks also increase as data moves between data stores. |
yes |
no |
Do customers benefit from real-time analytics?Changes completed by MySQL transactions are propagated in real-time to HeatWave and become immediately available for analytics queries, enabling real-time analytics.Snowflake users load data from their transactional applications into Snowflake through a complex ETL process. By the time data is available in Snowflake, it’s already stale, so customers don’t get real-time analytics. |
yes |
no |
Can customers eliminate the cost, complexity, and risk of ETL?MySQL HeatWave delivers a unified MySQL platform for analytics and transactional workloads. This eliminates the need for the complex, time-consuming, expensive ETL process and integration required by a different analytics database.Snowflake doesn't provide a single, unified service for OLTP and OLAP workloads (Snowflake’s Unistore is only in preview). Snowflake processes only OLAP workloads. Customers are required to move all data from MySQL/OLTP data sources through an ETL process to Snowflake, which increases complexity, risks, and costs. |
yes |
no |
"Open source developers who have not yet moved to MySQL Database Service with HeatWave are running out of reasons not to give it a try. Not only has Oracle simplified their lives with a unified OLTP and OLAP MySQL service, it has eliminated the need for a separate analytical database or data warehouse and ETLs between them. Plus, now it has delivered unparalleled performance and cost/performance."
—Marc Staimer, Senior Analyst, Wikibon
4. Automation powered by machine learning
MySQL Autopilot automates key and typically challenging aspects of achieving high query performance at scale. MySQL Autopilot is available at no additional charge for MySQL HeatWave customers.
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Does the service include built-in machine learning automation for auto provisioning?MySQL Autopilot Auto Provisioning predicts the number of HeatWave nodes required for running a workload by adaptive sampling of table data on which analytics is required. As a result, customers no longer need to manually estimate the optimal size of their cluster.Snowflake does not have the equivalent capability for auto provisioning. Developers need to guess or manually test the optimal cluster size for their workload, increasing cost and development time. |
yes |
no |
Does the service include built-in machine learning automation for auto query plan improvement?MySQL Autopilot Auto Query Plan Improvement learns various statistics from the execution of queries and improves the execution plan of future queries. The performance of the system therefore improves as more queries are run.Snowflake does not have the equivalent capability of auto query plan improvement. Query performance does not improve automatically over time. |
yes |
no |
Does the service include built-in machine learning automation for auto scheduling?MySQL Autopilot Auto Scheduling determines which queries in the queue are short running and prioritizes them over long running ones in an intelligent way to reduce overall wait time.Snowflake lacks the equivalent machine learning-powered workload management capability. Queries are executed in a First In, First Out (FIFO) order, which can degrade performance of short queries. |
yes |
no |
Does the service include built-in machine learning automation for auto data placement?MySQL Autopilot Auto Data Placement predicts the column on which tables should be partitioned in-memory to achieve the best performance for queries. It also predicts the expected gain in query performance with the new column recommendation. This minimizes data movement across nodes due to suboptimal choices that can be made by operators when manually selecting the column.Snowflake lacks the equivalent capability of machine learning-powered data placement. |
yes |
no |
"For MySQL database developers, this is a game changer. Usually, developers scale MySQL through labor-intensive sharding schemes, often resulting in complex application SQL logic, and exposing the system to multiple levels of human error. With HeatWave, this is not an issue. MySQL developers also run into constant challenges configuring the database for maximum efficiency and coding the SQL for the best performance. With HeatWave, these are no longer concerns"
—Carl Olofson, Research Vice President, Data Management Software, IDC
5. Existing applications benefit from real-time analytics without changes
ISVs and enterprises running MySQL applications can immediately benefit from HeatWave for real-time analytics, without making any modification to their applications.
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Can customers run transactions and get real-time analytics in the same database without changing their existing applications?With HeatWave, existing MySQL applications benefit from real-time analytics without any changes. HeatWave is designed as a MySQL pluggable storage engine, which completely shields all the low-level implementation details from customers. Therefore, they can manage both HeatWave and the MySQL database with the same management tools including the OCI console, REST APIs, and command line interface.Snowflake is not designed for mixed workloads. Customers cannot run OLTP workloads directly on Snowflake and require an ETL process to load data from MySQL or another source into Snowflake for OLAP processing. | yes |
no |
"The MySQL HeatWave technology is by far the best in the market now...Wikibon strongly recommends that enterprise IT departments set a three-year plan to eliminate separate OLAP databases and ETL from MySQL transactional databases."
—David Floyer, CTO, Wikibon
6. MySQL HeatWave eliminates the need for a separate machine learning service
With HeatWave AutoML, customers can build, train, deploy, and explain machine learning models within MySQL HeatWave.
Capability and Evidence |
MySQL HeatWave |
Snowflake |
|---|---|---|
Does the database service provide in-database machine learning?HeatWave AutoML enables data analysts and developers to build, train, deploy, and explain machine learning models within MySQL HeatWave.Snowflake lacks the equivalent in-database machine learning capability, requiring customers to move their data to a separate, third-party machine learning service. | yes |
no |
Are all machine learning models explainable, so users can understand and explain what happens in the models from input to output?All models created by HeatWave AutoML are explainable, which improves their reliability, fairness, trust, and regulatory compliance.Snowflake requires developers and data analysts to move their data to a third-party machine learning service. Security and compliance risks increase as data moves between systems. | yes |
no |
Is the machine learning lifecycle automated?HeatWave AutoML fully automates the machine learning lifecycle, including algorithm selection, intelligent data sampling, feature selection, and hyperparameter tuning—saving customers significant time and effort. HeatWave AutoML enables developers and data analysts to build machine learning models using familiar SQL commands; they don’t have to learn new tools and languages.Snowflake does not provide in-database machine learning capabilities. Customers must use a third-party machine learning service, increasing complexity and cost. | yes |
no |
Try MySQL HeatWave for free