INFICON Holding AG · SIX:IFCN
INFICON — Technical Presentation (September 2024)
44 slides from INFICON's technical presentation, September 2024
Slide 1: Maximize Your Manufacturing Efficiency: Three Fundamental Drivers of Fab Cycle Time
Maximize Your Manufacturing Efficiency: Gain Insight into the Three Fundamental Drivers of Fab Cycle Time Jennifer Robinson Cycle Time Evangelist for INFICON INFICON/FabTime Star Webinar for FOA 1
Slide 2: Why is cycle time improvement worthwhile?
Why is cycle time improvement worthwhile? Cost/revenue drivers include: Reducing Time to Market Increasing Fab Utilization Cutting Costs Cutting cycle time gets Cutting cycle time allows you Cutting cycle time may customer qual lots out the to run closer to full capacity. reduce scrap and rework.
Slide 3: Other reasons to improve cycle time include…
Other reasons to improve cycle time include… Happier customers • Shorter cycle times are usually also more predictable cycle times, resulting in better on-time delivery performance. Happier managers • A leaner fab with less WIP is more flexible and ready for market changes. • A fab carrying less WIP
Slide 4: Why are CTs high? Short answer: cost, complexity, and change
Why are CTs high? Short answer: cost, complexity, and change • Wafer fab equipment is expensive and tends to be highly loaded. • Wafer fab process flows are long and complex. • Product and technology mix change rapidly (so we don’t qualify new recipes on tools, and the tools can be unreliable). • An
Slide 5: Other challenges keeping fabs from getting great CT include…
Other challenges keeping fabs from getting great CT include… • Downtime (scheduled and unscheduled) • Operator unavailability • Product mix • Scheduling/dispatching • Bottlenecks/utilization • Engineering use of tools • One-of-a-kind tools/single path operations • Running development lots • Holds in
Slide 6: How to improve fab cycle time: a three-step process
How to improve fab cycle time: a three-step process Fundamentals Data Action • Help your team to • Collect and analyze • Act using understand where data to identify fundamentals and to look for current problems actual data from the problems and why. in the fab, fab. highlighting the • Use this webin
Slide 7: The fundamental drivers of CT can help you improve profitability by…
The 3 fundamental drivers of CT can help you improve profitability by… Reducing forced idle time on key tools (allowing We’ll use operating curves to illustrate impacts on CT increased loading of the fab). • Driver 1: Utilization Reducing variability, especially from downtime, and thus allowing fabs
Slide 8: Cycle Time and Utilization
Cycle Time and Utilization At the tool group level, utilization is the biggest driver of cycle time. INFICON/FabTime Star Webinar for FOA 8
Slide 9: We'll use the SEMI E10 tool state definition of utilization
We’ll use the SEMI E10 tool state definition of utilization Utilization = (Productive Time) / (Manufacturing Time) where Manufacturing Time = Productive + Standby (time the tool is available manufacturing) and Productive Time = time spent working on products Utilization = 15/20 = 75% INFICON/FabTi
Slide 10: When a tool group has no variability, CT is low regardless of utilization
When a tool group has no variability, CT is low regardless of utilization • Operation CT = Queue Time + Process Time (+ Hold, Travel, etc.) • Suppose we have a tool where lots arrive at a regular interval (gray bars), with no variability, and process time (blue bars) is always one hour, again with n
Slide 11: It's variability that causes lots to wait, and drives up cycle time
It’s variability that causes lots to wait, and drives up cycle time • Suppose now average process time is still an hour, but process times and times between arrivals both have moderate variability. • Now, we see some queueing even at low utilization (50%). • Lots 10 + 11 had longer process times,
Slide 12: Once there's variability, cycle time increases as utilization increases
Once there’s variability, cycle time increases as utilization increases At 80% utilization (same avg. process time, less time between arrivals), WIP bubbles are likely to be larger. • Long process time for Lot 10 is followed by a series of short times between arrivals. Just as CT starts to dip afte
Slide 13: Looking at a range of utilizations, over a longer period
Looking at a range of utilizations, over a longer period (230 arrivals) … Random numbers are simulated using an Excel spreadsheet tool. Contact Jennifer Robinson for a copy of the spreadsheet. INFICON/FabTime Star Webinar for FOA 13
Slide 14: We build an operating curve
… We build an operating curve Operating curve is a graph showing cycle time (or X-Factor, which is CT/Theoretical CT) vs. utilization for a fab or a tool group. • Usually generated from simulation or mathematical models. • Can generate from actual data and fit a curve to the data. • Useful for exp
Slide 15: The relationship between CT and utilization creates a Core Conflict for fabs
The relationship between CT and utilization creates a Core Conflict for fabs Utilization is the largest driver of CT at the tool level. To reduce costs, fabs need to increase tool utilization. To improve CT, fabs need to decrease tool utilization. Some standby time is needed on the tools, to keep
Slide 16: There are several ways that fabs can resolve the Core Conflict
There are several ways that fabs can resolve the Core Conflict 1. Improve equipment uptime, thus extracting standby time from current unavailable time. 2. Eliminate forced idle time (converting wasted time, recorded as standby-WIP- waiting time, into true standby time). 3. Reduce process times, so t
Slide 17: Food for Thought: Biggest source of forced idle time
Food for Thought: Which is the biggest source of forced idle time in your fab? Operator Holding Tools Something Dispatching Lot Transport Unavailability for Hot Lots else… INFICON/FabTime Star Webinar for FOA 17
Slide 18: How can fabs reduce forced idle time to improve productivity?
How can fabs reduce forced idle time to improve productivity? 1. Minimize the number of distinct tools for which each operator is Ideas to keep tools from the steep responsible, and stagger break schedules. part of the operating curve • This is to reduce time that tools spend with WIP, waiting for o
Slide 19: Cycle Time and Variability
Cycle Time and Variability Variability (in arrivals or process times) shifts the operating curve up and to the left. Variability makes it impossible to achieve theoretical cycle time in any real fab INFICON/FabTime Star Webinar for FOA 19
Slide 20: CT increases with variability in both process times and time between arrivals
CT increases with variability in both process times and time between arrivals Contributors to process time variability: Contributors to interarrival time variability: • Different recipes on the same machine / • All the above different process times. • Transfer batching/automated material handlin
Slide 21: Food for Thought: Biggest source of variability
Food for Thought: Which of these is the biggest source of variability in your fab? Extended Batch Time Links Something Hot Lots Downtimes Processing and Rework else... INFICON/FabTime Star Webinar for FOA 21
Slide 22: We can expand the queueing formula to account for variability
We can expand the queueing formula to account for variability Recall the single-tool X-Factor estimate: • X-Factor ~= 1 / (1 – Utilization) • Assumes medium variability in arrivals and process times More generally (again under certain assumptions) • X-Factor ~= 1 + [Utilization/(1-Utilization)] *
Slide 23: Changing the CV changes the shape of the curve
Changing the CV of arrival OR process time variability changes the shape of the curve INFICON/FabTime Star Webinar for FOA 23
Slide 24: So which type of variability has more of an impact?
So which type of variability has more of an impact? Variability Factor = (CVa2 + CVp2)/2 With equal amounts of arrival and process time variability, the impact will be equal. In practice: • Process time variability will be high on some tools (like implanters) and low on others (like furnaces). • Ar
Slide 25: Food for Thought: Should you group your PMs?
Food for Thought: Your maintenance team proposes grouping daily PMs into one long weekly PM for efficiency. Should you support the change? Yes! No! Maybe… Why or why not? More simply, for the same amount of scheduled downtime, is it better to have short/frequent PMs or long/infrequent PMs? Should yo
Slide 26: Impact of longer vs. shorter PM events on a one-of-a-kind tool
Consider the impact of longer vs. shorter PM events on a one-of-a-kind tool Suppose the tool is down for maintenance 20% of the time under moderate variability. For the same total amount of scheduled downtime (4.8 hours/day), consider taking the tool down for maintenance: • Once every 8-hour shift
Slide 27: Any type of variability shifts the operating curve up and to the left
Any type of variability shifts the operating curve up and to the left Anything that can be done to reduce variability in the fab will shift the operating curve downward and will improve things. Variability reduction gives the fab a choice: 1.Reduce cycle time, while maintaining the same throughpu
Slide 28: If we reduce variability, we can get a better CT at the same throughput rate
If we reduce variability, we can get a better CT at the same throughput rate INFICON/FabTime Star Webinar for FOA 28
Slide 29: Reducing CT can increase throughput if constrained by a CT target
If we are constrained by a CT target, then reducing CT can increase throughput INFICON/FabTime Star Webinar for FOA 29
Slide 30: Variability impairs performance for any manufacturing system
Variability impairs performance for any manufacturing system Variability is bad! Decreasing variability in arrivals or process times or from downtime will improve cycle times. Because you don’t have to reduce starts, or purchase additional tools, variability reduction is a relatively inexpensive wa
Slide 31: How can fabs reduce variability to improve profitability/performance?
How can fabs reduce variability to improve profitability/performance? Systematically monitor (using CVs and control charts) and reduce variability: • In process times. • In tool availability (or Green-to-Green time). • Separate maintenance events instead of grouping them. • Ensure sufficient MTs a
Slide 32: Cycle Time and Number of Qualified Tools
Cycle Time and Number of Qualified Tools The number of qualified tools in a tool group has a major impact on cycle time. This impact increases with higher levels of utilization and variability. INFICON/FabTime Star Webinar for FOA 32
Slide 33: Going from one to two qualified tools cuts CT per visit roughly in half
Going from one to two qualified tools cuts CT per visit roughly in half FabTimeOperatingCurveGeneratorCourse.xlsx INFICON/FabTime Star Webinar for FOA 33
Slide 34: As we increase the number of qualified tools, the benefit levels off
As we increase the number of qualified tools/recipe, the benefit levels off Tool group is 90% utilized, level of cross-qualification varies. INFICON/FabTime Star Webinar for FOA 34
Slide 35: The intuitive explanation is that we need a backup in case of variability
The intuitive explanation is that we need a backup in case of variability When there’s only one tool, a lot is subject to all the variability associated with that tool (long process times, downtimes, etc.). When there are two tools, the probability of both tools being hit by a long delay at the same
Slide 36: Keeping single-path operations visible in fab reports
We highly recommend keeping single-path operations visible in fab reports The FabTime chart on the left shows, for all route-step combinations that have WIP waiting, the number of tools qualified to run that step. Steps with more than two tools are not displayed. The FPS Dashboard chart on the right
Slide 37: Identify where soft dedication is occurring
It’s also useful to identify where “soft dedication” is occurring • Soft dedication is when operators have preferences for certain tools (layout, speed, etc.) • Graphing availability efficiency and moves by tool for a tool group can show where a tool has higher or lower moves than availability would
Slide 38: How should fabs manage tool qualification to improve performance?
How should fabs manage tool qualification to improve performance? Education: Teach team members (especially process engineers) the impact of tool qualification on cycle time curve. Systematically monitor and reduce: Active constraints (especially complex constraints!). Do all constraints have an e
Slide 39: Wrapping Up
Wrapping Up Utilization, Variability, and Tool Qualification issues all drive cycle time higher. Understanding the cycle time impact of these factors helps your organization identify and prioritize opportunities for cycle time reduction (= profitability improvement). INFICON/FabTime Star Webinar for
Slide 40: Recap: The fundamental drivers of fab cycle time
Recap: The 3 fundamental drivers of fab cycle time Utilization = Productive / (Productive + Standby) • For single tools, operating curve shaped like X-Factor = 1 / (1 – Utilization). When utilization approaches 100%, CT gets very high. • Converting unavailable time and forced idle time to true sta
Slide 41: Subscribe to the FabTime Cycle Time Newsletter
For more ideas: subscribe to the FabTime Cycle Time Newsletter If the content in this webinar seems useful to you, consider subscribing to the FabTime Cycle Time Management Newsletter (now distributed by INFICON) • Bi-monthly email publication, currently sent as PDF attachment, 2800+ subscribers, 18
Slide 42: Focusing on the fundamental drivers of CT can help fabs improve profitability
Focusing on the fundamental drivers of CT can help fabs improve profitability Cycle time can be reduced with $$$ (tools and staffing). • This is often infeasible for various reasons. Cycle time can also be reduced with education and behavior changes. • Step 1: Use operating curve concepts to under
Slide 43: INFICON can help via our course and portfolio of Smart Solutions
INFICON can help via our course and complete portfolio of Smart Solutions Learning data warehouse and Real-time factory priorities, factory physics engine FabTime & flexible user-customized reporting Smart Digital Factory Twin Dashboard Optimized maintenance timing Global optimization of Return Fact
Slide 44: Connect with us!
Connect with us! www.inficon.com