Clarification on Data Streaming Accelerator (DSA) and In-memory Analytics Accelerator (IAA)

[Karl3]

Hello, 

We have an HPE DL 380 Gen11 with two Intel Xeon Platinum 8462Y+ Processors.

https://www.intel.com/content/www/us/en/products/sku/232383/intel-xeon-platinum-8462y-processor-60m-cache-2-80-ghz/specifications.html

These processors have Data Streaming Accelerators (DSA) and Intel® In-memory Analytics Accelerators (IAA).

We have a vendor-supplied database that has not been changed to use these accelerators.

Will a database application get the benefit of the Data Streaming Accelerator (DSA) without any change or does the application need to be recoded to make use of the Data Streaming Accelerator (DSA) calls?

This document seems to suggest an application needs to be changed. Is this correct?

https://www.intel.com/content/www/us/en/content-details/759709/intel-data-streaming-accelerator-user-guide.html

It is targeted for optimizing streaming data movement and transformation operations common with applications for high-performance storage, networking, persistent memory, and various data processing applications

This document’s intended audience includes system administrators who may need to configure Intel DSA devices and developers who want to enable Intel DSA support in applications and use libraries that provide interfaces to Intel DSA. It should be read in conjunction with the Intel® DSA Architecture Speci-fication and documentation for SW utilities and libraries that support Intel DSA, such as accel-
config/libaccel-config, Libfabric, and Intel® MPI.

Likewise will a database  application get the benefit of the In-memory Analytics Accelerators (IAA) without any change or does the application need to be recoded to make use of the In-memory Analytics Accelerators (IAA)?

Any advice would be greatly appreciated.

Regards, Karl

[IntelSupport]

Hello Karl3,

Thank you for posting intel community.

At the moment, we are actively working towards finding an appropriate resolution for your issue, and we will provide you with an update as soon as we can. We kindly request your cooperation during this time.

Best Regards,

Vijay N.

[IntelSupport]

Hello Karl3,

Greetings for the day!

• Intel IAA provides compression and decompression of very high throughput combined with analytic primitive functions commonly used for data filtering during analytic query processing. Intel QPL helps developers activate this accelerator’s advanced data analytical capabilities.
• Intel DSA is a high-performance data copy and transformation accelerator targeted for optimizing streaming data movement and transformation operations common with applications for high-performance storage, networking, persistent memory, and various data processing applications. Intel DML helps developers activate this accelerator’s advanced data movement capabilities.

To fully leverage the Data Streaming Accelerator (DSA) and In-memory Analytics Accelerator (IAA) in a database application, it is generally necessary to modify or recode the application to use DSA-specific and IAA-specific calls and libraries. While some benefits may be realized without changes, to achieve the maximum potential performance gains from these hardware accelerators, application-specific optimizations are often required.

Please find the below attached link for full details:

https://www.intel.com/content/www/us/en/developer/articles/technical/optimize-data-manipulation-with-qpl-and-dml.html#gs.76ajkm

Best Regards,

Vijay N

[Karl3]

Hi Vijay N,

Thanks for your last update. 

Is there any document that describes what the benefits of DSA and IAA are to an application without changing that application? Also are there any benchmarks for DSA and IAA for any application without changing that application?

Regards, Karl

[IntelSupport · Accepted solution]

Hello Karl3,

Greetings for the day!

Operation Offloading to IAA Devices:

Intel® Query Processing Library (Intel® QPL) The Intel® Query Processing Library (Intel® QPL) can improve the performance of databases, enterprise data, communications, and scientific/technical applications. Intel QPL provides interfaces for several commonly used algorithms. This library enables your application to tune automatically with generations of processors without unnecessary changes to the application. The Intel QPL provides high-performance implementations of data processing functions for an existing hardware accelerator and/or software path if no hardware accelerator is available. Code written with the library automatically takes advantage of available modern CPU capabilities. This can provide tremendous development and maintenance savings. The goal of the Intel QPL is to provide an application programming interface (API) with the following:

• C and C++-compatible interfaces and data structures to enhance usability and portability.

• Faster time to market.

• Scalability with Intel® In-Memory Analytics Accelerator (Intel® IAA) hardware.

More details on Intel QPL use are available in the https://intel.github.io/qpl/index.html (https://intel.github.io/qpl/index.html)

Please find the below attached links to get the more detailed information for DSA and IAA:

https://cdrdv2-public.intel.com/780887/354834_IAA_UserGuide_June23.pdf

https://cdrdv2-public.intel.com/759709/353216-data-streaming-accelerator-user-guide-2.pdf

Regards,

Vijay N.

[IntelSupport]

Hello Karl3,

We hope you're doing well.

We wanted to follow up on our previous post. If you have any questions or need more information, please don't hesitate to let us know.

Best regards,

Vijay N.

[IntelSupport]

Hello Karl3,

We hope you're doing well.

We have not received a reply from you , and as such, we will be closing your case.

If you want to continue support, please reply to this email and we will reopen your case or create a new one so that we can continue to support you.

Best regards,

Vijay N.