Innovating with ISV partners

By September 8, 2021

Power E1080 close-up

ISVs have always played a critical role in the value IBM Power delivers to customers. Through joint development and optimization with ISVs, IBM Power has continuously led the market in performance, scalability, and flexibility. As we launch Power10 we’re excited to highlight just some of the continued innovation with ISVs, going beyond performance optimization to how we are integrating AI, extending capabilities around Linux containers and Red Hat OpenShift, and teaming to further enhance security of one of the most secure platforms in the industry.

Jointly Innovating

The launch of the IBM Power E1080 system is only the beginning. We have a number of new innovations coming to Power10 on which we are partnering with ISVs such as SAP. Changbin Song, Vice President, SAP HANA & Analytics at SAP shared “SAP and IBM are working continuously on projects and technologies to benefit their joint customers. Together, SAP and IBM plan to look at new and improved capabilities brought by Power10 for its SAP HANA workloads.” 

In Core AI Inferencing and Machine Learning

The Power10 processor is designed to democratize AI across applications and operating environments, be it on Linux, AIX, IBM i, and across the entire Power10 family of servers. IBM Power users can now infuse AI directly into the core business applications and enterprise databases and run AI close to where the data resides. Our ISV partners can leverage this via four new Matrix Math Accelerator (MMA) units in each core. MMAs are alternatives to external accelerators for the execution of statistical machine learning and inferencing and are designed to reduce cost and provide a simplified solution stack for AI. IBM has optimized math libraries to enable AI tools to benefit from the acceleration provided by MMA units. The IBM Power E1080 has 5x faster AI inferencing per socket over the IBM Power E9801 and, by supporting MMAs, ISVs can use the capabilities to help customers to speed up their applications and run high performance AI with security and efficiency.

Accelerating business decisions with Red Hat

IBM and Red Hat have extended their synergy around Red Hat OpenShift and Power10 to accelerate business insights with the infusion of AI into applications and data stores with the new Power10 MMA engines. This is designed to eliminate specialized systems or attached accelerators, to lower the total cost of ownership. In addition, IBM has enabled greater efficiency in cloud-native workloads with the ability to have 4X more containerized throughput per core running Red Hat OpenShift applications than compared x862.

End-to-end cloud-native security

Securing your data is more important today than ever before with the high threat of cyber-attacks. Power10 is architected for security, leveraging accelerated encryption across the stack so that data is available only to people authorized via encryption keys. To help support our customers’ modernization effort, IBM has teamed with Aqua Security, a pure-play cloud-native security provider that builds on our strong, secured Power10 base. Amir Jerbi, Co-founder and CTO at Aqua Security, shared that “IBM customers are shifting to hybrid cloud environments to help address operating costs and increase automation, but in doing so they also demand security and compliance. Aqua plays a key role and provides a further layer of cloud native security to support this transformation. We are proud to be one of the first security providers to help IBM and their customers achieve the potential of cloud native applications.” For more details on our relationship, check out Aqua’s press release and video. Partnering with Aqua gives our customers the ability to securely deploy cloud-native Red Hat OpenShift solutions on IBM Power.

Getting Started with Power10

We understand that many of our customers may not upgrade their software and hardware concurrently. These customers can take advantage of the ability to run in Power9 compatibility mode, designed to enable them to modernize software environment at their own pace.

IBM Power has a long history of partnership and co-creation with ISVs, providing support such as complementary access for development and testing on the latest equipment and collaboration with technical professionals. Whenever we have a new server release, as with Power10, we encourage ISVs to work with their IBM contact to be nominated for early access to hardware through our Early Support Program. ISVs can learn more about the support and programs available through the ISV Resource Center.

Three things ISVs can do to get started today:

  1. Read the Power10 press release
  2. Contact IBM to learn about the nomination process for getting early access to hardware
  3. Visit the ISV Resource Center for information about support and programs and connect with IBM Power professionals
[1] 5x improvement in per socket inferencing throughput for large size 32b floating point inferencing models from Power9 E980 (12-core modules) to Power E1080 (15-core modules) Based on IBM testing using Pytorch, OpenBLAS on the same BERT Large with SqUAD v1.1 data set
[2] Based on IBM internal testing of Red Hat OpenShift Container Platform 4.8.2 worker nodes running 80 pods using the Daytrader7 workload (https://github.com/WASdev/sample.daytrader7/releases/tag/v1.4) accessing an AIX Db2 database. Comparison E1080 running OCP accessing AIX Db2 on a S922 versus OCP on Cascade Lake accessing AIX Db2 on a S922. Valid as of 8/25/2021 and conducted under laboratory conditions. Individual result can vary based on workload size, use of storage subsystems & other conditions. IBM Power E1080 (40 cores/3.8 GHz/2 TB memory) in maximum performance mode, 25 Gb two-port SRIOV adapter, 1 x 16gbps FCA, Websphere Liberty 21.0.0.6, Java™ SE Runtime Environment (build 8.0.6.36 – pxl6480sr6fp36-20210824_02(SR6 FP36)), CoreOS Linux 4.18.0-305.10.2.el8_4 with PowerVM. Configuration consists of 2 OCP worker lpars each with 10 cores running SMT8 with 256GB of memory. X86 system: Intel® Xeon® Gold 6248 CPU (Cascade Lake) in performance mode, 40 cores/3.9GHz/512GB memory), 25Gb two-port SRIOV adapter, 1 x 16gbps FCA , Websphere Liberty 21.0.0.6, Java(TM) SE Runtime Environment (build 8.0.6.36 – pxl6480sr6fp36-20210824_02(SR6 FP36)), CoreOS Linux 4.18.0-305.10.2.el8_4, RHEL 8.4 KVM. Configuration consists of 2 OCP worker KVM guests each with 20 cores running hyperthreading with 256GB of memory. Database system S922: Model 9009-22G with 22 cores and 1TB of memory. Configuration consists of 2 AIX lpars each with 8 cores running SMT8 with 131GB of memory, and a VIOS lpar with 2 cores and 16GB of memory.
Statements by IBM regarding its plans, directions, and intent are subject to change or withdrawal without notice at the sole discretion of IBM. Information regarding potential future products is intended to outline general product direction and should not be relied on in making a purchasing decision.
[autopilot_shortcode]