UploadLanguage (EN)Support
BusinessMobileSocial MediaMarketingTechnologyArt & PhotosCareerDesignEducationPresentations & Public SpeakingGovernment & NonprofitHealthcareInternetLawLeadership & ManagementAutomotiveEngineeringSoftwareRecruiting & HRRetailSalesServicesScienceSmall Business & EntrepreneurshipFoodEnvironmentEconomy & FinanceData & AnalyticsInvestor RelationsSportsSpiritualNews & PoliticsTravelSelf ImprovementReal EstateEntertainment & HumorHealth & MedicineDevices & HardwareLifestyle
Upload
Sign in
Uploaded byjoechee602
PPTX, PDF958 views

Safe implementation in intel

The document summarizes Intel's implementation of the Scaled Agile Framework (SAFe) at its MDO organization. It describes how Intel transitioned from a vertical to horizontal structure with 177 scrum teams organized into 12 Agile Release Trains. While the intent was to follow SAFe principles and standard rules across the organization, variations in agile maturity and other challenges emerged. Key lessons included that SAFe is a set of guidelines rather than a prescriptive template, and successfully implementing even simple rules at scale is difficult due to factors like team sizing, co-location constraints, and pacing teams across geographies.

Presentations & Public Speaking◦
Downloaded 14 times
1 / 28
2 / 28
3 / 28
4 / 28
5 / 28
6 / 28
7 / 28
8 / 28
9 / 28
10 / 28
11 / 28
12 / 28
13 / 28
14 / 28
15 / 28
16 / 28
17 / 28
18 / 28
19 / 28
20 / 28
SAFe at Intel MDO Intel Agile and Lean Development Conference - 2014 Sean Goh & Joseph Chee MDO Process & Efficiency Team Penang Sept 2014
Agenda 1. What’s on the Market 2. MDO Organization 3. SAFe Implementation at MDO 4. TIPS for success 5. Challenges
Process and Efficiency Team 3
What’s in the Market? LeSS – Craig Larman & Bas Vodde DaD – Scott Ambler SAFe – Dean Leffingwell
SAFe
SAFe Principles - Intent • Output Quality – “You can’t scale crappy code.” • Predictable Execution – “Agile Release Trains – self organizing teams of agile teams – reliably and frequently deliver value on demand” • Alignment & Synergy – “More value is created by overall alignment than with local excellence” • Up/Down Transparency – “Transparency builds confidence, alignment and trust”
MDO Value Stream All Reporting to Single Organization Pre-MDO reporting structure: 7
See the change BO and CE are Project Managers MDO reporting structure: 8 5 Horizontal Org 3 Horizontal Org
Vision 9 Horizontals S/C TP DCD Analog PHI MPV Client BO S/C TPI, fuse Arrays, reset, BF, scan IO, PLL, power/thermal Yield, BS PPV, CMV SOC BO Server BO Chipset BO Reporting structure. Responsible for the “how” Productstructure–responsible forthe“what”
ART Release Planning Plan Committed!!
Release Planning Artifacts Sprint Planning Board Risk Board Dependency Board
Release Planning Artifacts Sprint Planning Board Risk Board Dependency Board
TIPS for Success 1. Leadership Support & Investments 2. Team Engagement & Strategy 3. Art of Standardization. 4. Can’t scale bad teams 5. Inspect and Adapt
Servant Leadership
Team Engagement & Strategy
Standardization?
Can’t Scale Bad Teams
Inspect and Adapt
Reflection – After Action Review (AAR) • INTENT vs. ACTUAL – What did we intend to happen? – What actually happened? Why? • KEY LEARNINGS AND ACTION PLAN – What can we learn from Previous Iteration? – What can we Sustain? What do we need to Improve? – What actions can be taken during the next Iteration?
Intent vs. Actual Intent • Building on our good scrum & release planning practices, launch ARTs for all the products at once in WW04 • Use SAFe as a template to remove variation in the organization’s execution • Follow a simple rules across organization to guarantee efficiency and adherence to agile principles when constructing the agile teams and trains Actual • Took us a quarter to pull this off for business reasons – All trains finish the release on time, but not all start on time • 1200 engrs and tech in 177 scrums connected to 12 trains • Variation in our agile maturity is showing up at the train level – Can’t scale bad/immature agile • SAFe is not a prescriptive template, rather a set of guidelines about what you might need to pay attention to. – Trick is to figure out where to implement as a rule and where to experiment with an idea • Though the rules are simple, following them is not.
Intent vs. Actual (Rules) • 100/177 Scrums are too small/big • 74/177 Scrums are X- Site • 15/177 Scrums work on multiple trains • 6/12 trains are run across 3 geos Scrum Level Rules 1. Size – 5-9 Team Members 2. Commitment – Permanent members 100% dedicated 3. Co-Location – Same site, sitting together. 4. X-Func – Can execute value stream, end to end, w/o handoffs Train Level Rules 1. Size - Train can have 5-15 scrums 2. Commitment - 1 Scrum works on only 1 train at a time 3. Co-Location – max 2 geos per train with 3+ overlap hours Accounting errors - Plans to fix – Waivers – Rule Changes Some teams cannot follow some of the scrum rules . .
Just a Framework
Challenges 1. Conversation Over Indicator? 2. Pacing the Teams 3. Co-location Complexity
Conversation over Indicators? CAN WE COMPROMISE?
Pacing the Teams CAN WE COMPROMISE?
Co-location Complexity
27
Backup Intel Software Professionals Conference - Intel Confidential 28