Baha Majid WCA4Z IBM Z Customer Council Boston June 2024.pdf

BahaMajid 205 views 21 slides Jun 17, 2024
Slide 1
Slide 1 of 21
Slide 1
1
Slide 2
2
Slide 3
3
Slide 4
4
Slide 5
5
Slide 6
6
Slide 7
7
Slide 8
8
Slide 9
9
Slide 10
10
Slide 11
11
Slide 12
12
Slide 13
13
Slide 14
14
Slide 15
15
Slide 16
16
Slide 17
17
Slide 18
18
Slide 19
19
Slide 20
20
Slide 21
21

About This Presentation

IBM watsonx Code Assistant for Z, our latest Generative AI-assisted mainframe application modernization solution. Mainframe (IBM Z) application modernization is a topic that every mainframe client is addressing to various degrees today, driven largely from digital transformation. With generative AI ...


Slide Content

AI-assisted mainframe
application modernization
with IBM watsonxCode Assistant for Z
Baha Majid –Z Client Architect
[email protected]
June6, 2024
IBM Z Customer Council –Boston, MA

Our principal challenge facing organizations
today is accelerating
mainframe application modernization
lines of COBOL are estimated to be
actively running by enterprises
1
230+ billion
1. Source: Reuters Survey (2020)
2

IBM’s mainframe application modernization strategy
3
IBM Z integrated in a Hybrid Cloud Platform With IBM Z in the Hybrid Cloud:
Enterprise standardization
Platform integration
Accelerate time to market for cloud native
services with a consistent DevOps experience
Reduce the talent gap with common tools and operating
models across platforms
Easily access IBM Z data without moving
off-platform
Optimize costs with a cloud consumption
model that extends to IBM Z
SaaS
Private
cloud
Public
cloud
Cloud Paks
watsonx
Watson Code
Assistant
© 2024 IBM Corporation

IBM is accelerating application modernization
with generative AI
Open-source Tools
PatternsApplication Discovery / co-creation
Build the right foundation
Optimize the core
Manage the cost and performance of running
current applications.
Enhance and modernize applications
Augment, refactor, and create applications leveraging
cloud-native principles.
Integrate across hybrid cloud
Leverage open APIs and event-driven architecture
to integrate hybrid applications.
Simplify information access and sharing
Optimize and secure data access and information
sharing across the enterprise.
Increase business agility & productivity
Get more agile with enterprise DevOps
Leverage enterprise DevOps with an integrated CI/CD
pipeline and full application observability.
Make AI-driven decisions at scale
Achieve AI-driven insight at scale to help make
decisions in real time.
Automate and standardize IT
Create standardized IT automation for mainframe
infrastructure and tasks using a common approach.
Accelerate your journey
IBM watsonx Code
Assistant for Z
© 2024 IBM Corporation
redbooks.ibm.com/redpapers/pdfs/redp5705.pdf
4

Objectives:

Address skills and
productivitychallenges
with automation and AI

Ensure IBM Z qualities
of service with mixed
language interoperability

Align with
industry standard
DevOps approaches
Accelerated
application lifecycle
IBM watsonx Code Assistant for Z modernization experience
Initial Scope
5.Recommend
and deploy
best-fit architecture
2.Refactor
business services
in COBOL
3.Transform
code with optimized
design and architecture
4.Validate
outcome with
auto-generated tests
6.Observe
across hybrid
application landscape
1.Understand
application landscape
with auto-discovery
and code explanation
© 2024 IBM Corporation
5

© 2024 IBM Corporation
6

Visualizing the Applications
© 2024 IBM Corporation
7

Hills: Who, what, wow Hill 1: Reduce the learning time
An application developercan
quickly learn what the COBOL
program is doing, with minimal
reliance on the application
subject matter expert (SME).
Hill 2: Documentation
An application developercan
generate code explanations
that can be inserted as
comments or downloaded for
documentation, with the
confidence that the
explanation is reliable and up-
to-date.
Hill 3: Refactoring
A COBOL developercan
generate code explanations
that canhelp them
understandand refactor the
COBOL code into a new COBOL
service,without relying on the
application SME.
IBM INTERNAL and CONFIDENTIAL
8
Download Insert as comment Copy

COBOL Code Explanation
© 2024 IBM Corporation
9

© 2024 IBM Corporation
10

Refactoring
© 2024 IBM Corporation
11

© 2024 IBM Corporation
12

© 2024 IBM Corporation
13
A sample of a COBOL CICS Transaction transformed to Java

© 2024 IBM Corporation
14

© 2024 IBM Corporation
15
Overview –Z Validation Assistant

Watsonx Code Assistant for Z use cases and outcomes
16
Use Cases:
Outcomes:
© 2024 IBM Corporation
Transform with JavaModernize your COBOL
Language
specific benefits
Developer skills
& productivity
COBOL
Refactoring
Understand your
application
Bring the tooling and
ecosystem benefits of
enterprise Java to mainframe
application development
Tap a broader pool of
developer talent and
enhance their efficiency
with generative AI
Gain agility by
decomposing your
monolith into modular
business services
Deep analysis to discover
and inventory data and
program relationships with
code explanation

17
WCA4Z
Overall
Architecture

DevOps and a modern pipeline is essential for IBM watsonxCode Assistant for Z
Integration/Functional/Performance Testing
Integrate and automate delivery of z/OS applications
Understand Refactor Transform ValidateContinuous
Integration
Edit (IDE), Debug, Build
Continuous Testing
On Demand Test EnvironmentsUnit Testing
Continuous
Delivery
*Automated Application
Deployment
Nexus

Continuous improvement with a DevOps pipeline is fundamental to iterative
application modernization using IBM watsonx Code Assistant for Z (WCA4Z)

20 to 50% productivity improvement with IBM Z DevOps modern development tools

50 to 90% time saving on test cycles with Automated testing.
Integration to Test and CI/CD pipeline © 2023 IBM 18

Watsonx Code
Assistant Service
Large Language Model
Transform
[IBM Cloud]
Cloud
3
WatsonxCode Assistant for Z
Service (SaaS)
WCA for Z Service
Distributed Infrastructure (x86)
LLM Inference
Engine
Cloud Pak for Data Services
New! Planned:
WatsonxCode Assistant for Z on
Prem LLM Service
On prem LLM Stack overview
Openshift& CP4D services
Are necessaryfor WCA4Z on-prem
Will come packaged in new on prem
part
GPUs required (1 GPU)
For Openshiftworker nodes
1 Nvidia A100 or H100
x86 hardware is required
Leveraging watsonx.ai stack
Investigating s390x
deployment (Linux on Z, but
that will be future option)
Today: Watsonx Code Assistant for Z initial GA leverages
a mixed deployment model

IBM watsonx Code Assistant for Z
2024 anticipated roadmap highlights
20
4Q ‘24 3Q ‘242Q '241Q ‘24
Outcome
Intended
Capability
Outcome
Intended
Capability
Outcome
Intended
Capability
Outcome
Intended
Capability
Ability for WCA for Z to
support PL/I applications
for application
modernization
PL/I
Support
Generative AI capability to
summarize and explain
JCL as written English to
enhance understanding,
aid skills transfer and
enable more efficient
hand-over between
System Programmers.
Code
Explanation
for JCL
Clients can deploy the
Transform phase on
their premises, which
removes SaaS
component
requirement, enabling a
full on prem experience
Transform
phase
available on
prem
Capability to compare
COBOL and translated
Java at the Unit level
Validate
phase
available
Tools for clients to
customize “their version”
of the LLM for COBOL to
Java translation with their
data and preferences
expected to improve
output quality
Model
Custom-
ization &
Tuning
IBM watsonx Code
Assistant for Z service
available to be deployed
in IBM Cloud’s Frankfurt
multi-zone regions
Frankfurt
Cloud
location
IBM WCA for Z
Refactoring Assistant UI
can now be experienced
in VS Code through Z
Open Editor Extension
Refactor
Phase
Available in
VS Code
New generative AI
capability to summarize
and explain COBOL
program source code as
written English to enhance
skills transfer and
application understanding
Code
Explanation
for COBOL
Improved subsystem
coverage across CICS,
IMS TM, IMS DB, MQ
and Db2 to enable
additional COBOL to
Java translation use
cases
z/OS
subsystem
support for
Transform
phase
Roadmap is subject to change
Continuous improvement to all available capabilities including GenAImodel(s)
© 2024 IBM Corporation
Tags