Intel’s AVX-512 Implementation At PostgreSQL 17 Brings In a Whopping 48% Performance Uplift

Muhammad Zuhair
Intel Hints At Dedicated Cache Tile For CPUs But Not For Desktops 1

Intel's newest PostgreSQL 17 database server release has brought in phenomenal performance improvements for AVX-512 workloads, providing double-digit gains.

Intel Brings In Steady Performance Optimization For Xeon Users Leveraging The AVX-512 Architecture At PostgreSQL

Team Blue's efforts in optimizing its product portfolio have recently seen a new high, especially on platforms other than Windows, such as Linux. Not only is the firm active when it comes to bringing in next-gen support for architectures such as Lunar Lake and Arrow Lake, but the company has brought in decent improvements in its Arc GPU performance as well, showing that Team Blue is committed to providing extensive support for all consumer platforms. Now, according to Phoronix, Intel's new "pop count implementation" for PostgreSQL 17 has brought in massive performance gains.

Related Story Intel Executive Warns CPU Shortage Is Hitting “Everyone”, and Price Hikes Are Inbound but Nothing Like the Memory Crisis
On Dec. 14, 2023, Intel introduced its 5th Gen Intel Xeon Scalable processors, delivering increased performance per watt and lower total cost of ownership across critical workloads for artificial intelligence, analytics, networking, security, storage and high performance computing. (Credit: Intel Corporation)

The post is a bit on the technical side, but it really does show us the whole story when it comes to how actively Intel engineers are working to provide the optimal support for all consumer segments. For those unaware, PostgreSQL 17 is an open-source object-relational database system utilizing SQL to coordinate with the database. It is highly utilized by professional markets, mainly in enterprise applications, research projects, and data warehousing.

Intel says that with the presence of AVX-512 instruction set in the latest Xeon CPUs, they have managed to bring in phenomenal performance improvements, mainly by adjusting runtime data fetching and optimizing the overall process. Here is what Intel's engineer says about the new implementation:

Presently, pg_popcount() processes data in 32-bit or 64-bit chunks when possible. Newer hardware that supports AVX-512 instructions can use 512-bit chunks, which provides a nice speedup, especially for larger buffers.

This commit introduces the infrastructure required to detect compiler and CPU support for the required AVX-512 intrinsic functions, and it adds a new pg_popcount() implementation that uses these functions.

Intel's own testing of AVX-512 for PostgreSQL bit counting shows that the firm has managed to see 48% performance gains with an Intel Xeon Platinum 8480+ server CPU while stacked up against the traditional SSE implementation. This is a massive achievement for the company in the workstation segment, and it shows the capabilities brought in by the AVX-512 instruction set.

Image Credits: Phoronix

Team Blue has been scaling itself up with tiny steps, providing users with optimal support in order to gain a cumulative increase in adoption. The firm still has a long way to go, especially with platforms like Linux, but by the looks of it, the progress is indeed steady and positive.

Muhammad Zuhair Photo

About the author: Muhammad Zuhair is a hardware and technology reporter for Wccftech, specializing in the semiconductor industry and the complex interplay between technology, manufacturing, and geopolitics. His coverage focuses on the corporate strategies and technological roadmaps of industry giants like TSMC, NVIDIA, Samsung, and Intel. Zuhair's expertise lies in deconstructing complex topics such as fabrication nodes (e.g., 2nm process), the economic impact of policies like the CHIPS Act, and the strategic development of AI infrastructure from NVIDIA, AMD and Intel.

Follow Wccftech on Google to get more of our news coverage in your feeds.

Button