Slides
Page 1
SCALING THE CIRCULAR TECHNOLOGY PLATFORM Extending Our Growth Trajectory 19 September 2026 SLS Change photo
Page 2
Disclaimer The material contained in this document is a presentation of information about the Group’s activities current at the date of the presentation, 19 September 2026, CT. It is provided in summary form and does not purport to be complete. It should be read in conjunction with the Group’s periodic reporting and other announcements lodged with the Australian Securities Exchange (ASX). To the extent that this document may contain forward-looking statements, such statements are not guarantees or predictions of future performance, and involve known and unknown risks, uncertainties and other factors, many of which are beyond the control of the Group, and which may cause actual results to differ materially from those expressed in the statements contained in this release. This document is not intended to be relied upon as advice to investors or potential investors and does not take into account the investment objectives, financial situation or needs of any particular investor. All currency amounts in this presentation are in AUD unless otherwise stated as USD. Authorised for Release by: the Company Secretary, Gretchen Johanns ABN 69 114 838 630 Head Office: level 9, 189 O’Riordan Street, Mascot, NSW, Australia 2020 2
Page 3
Agenda Saturday 19 September 2026 Start Time 10:30 - 10:45 am Presentation – Strategic Position Presenter: Stephen Mikkelsen, Group CEO and MD 10:45 - 11:10 am Presentation – Market Opportunity and Commercial Strategy Presenter: Sean Magann, CCO – SLS Global 11:10 - 11:30 am Presentation – Operational Excellence and Scalability Presenter: Lynn Jacobs, COO – SLS Global 11:30 - 12:00 pm Questions and Answers 3
Page 4
Stephen Mikkelsen Sims Group CEO & Managing Director Strategic Position 4
Page 5
SLS Has Become a Material Earnings Contributor Diversifying the Portfolio and Increasing Exposure to Structural Data-Centre Infrastructure Growth $173.2 $62.7 $148.8 $175.2 $163.5 $102.2 $120.0 $235.5 $8.2 $17.7 $32.6 $172.7 FY23 FY24 FY25 FY26 Sims Metal SAR SLS 1 A Material Earnings Engine SLS has progressed from an adjacent service to a core portfolio contributor. 2 A Broader Earnings Base Technology lifecycle exposure complements the metals portfolio. 3 A Scalable Growth Platform Customer integration and recovery expertise support growth with disciplined capital. A Meaningful Earnings Platform Alongside Metal 1 Sims Metal includes NAM, ANZ and Global Trade. Sims owns 50% of SAR, and Group’s underlying EBIT includes its 50% share of SAR’s profit before tax, adjusted from US GAAP to IFRS. FY26 Underlying EBIT1 ($m) 5
Page 6
A Secure, Integrated Process Routes Equipment to the Best Circular and Commercial Outcome DECOMMISSION PROCESS & PREPARE OPTIMISE THE OUTCOME SECURE DECOMMISSIONING Collection, transport, data security and chain of custody. PROCESS & ASSESS Receive, harvest, test, grade and prepare equipment and components. REDEPLOY Prepare and return equipment or components to the customer. RESALE Sell viable components into global secondary markets. RECYCLE Send residual equipment to shredders for material recovery. More Value Recovered | Less Waste Generated One Integrated Platform, Multiple Revenue Streams 6 FEE PER UNIT % REVENUE SHARE COMMODITY RECOVERY
Page 7
FY26 Gross Margin by Segment ($m) • Only external provider of this capability at scale, developed over three years. • Strengthens customer relationships, supporting access to additional memory volumes. • 60-70% of the memory is recovered and resold, with the remaining eligible for redeployment. • Used selectively, based on customer needs and asset suitability. • Certified, reprogrammed used memory supports CXL (Compute Express Link), enabling CPUs and GPUs to access a shared pool of memory. This pool can incorporate either new memory or certified, reprogrammed used memory. Redeployment FY26 Resale Gross Margin Mix 74% 26% Memory Non-Memory Redeployment Deepens Relationships 7 52 106 20995 89 122 45 34 54 7 12 22 FY24 FY25 FY26 Resale Redeploy/Services Commodity Recovery Other Revenue $199m $241m $407m
Page 8
Customer, capability and commercial shifts position SLS at the centre of DCIS1 1 DCIS refers to Data Centre Infrastructure Services. 2 EBIT comprises $25.2 million from the disposed European compliance scheme operations and $0.8 million from continuing operati ons. Total volumes comprise 326,855 tonnes from the disposed European compliance business and 74,673 tonnes from continuing operations. 3 FY20 Underlying EBIT. Excludes the European compliance scheme operations. FY19 FY20 FY21 FY25 FY26 ITAD Foundation Compliance Scheme Reuse, Redeployment and Recycling of Cloud Infrastructure Repeatable Global Data-Centre Decommissioning Hyperscaler Entry and Integrated Customer Workflows Underlying EBIT Underlying EBIT DCIS-Focused Technology Platform Underlying EBIT Underlying EBIT Underlying EBIT A$26.0m2 A$2.9m3 A$21.8m A$32.6m A$172.7m 401,5282 Tonnes 1.6m Repurposed Units 2.1m Repurposed Units 8.8m Repurposed Units 16.8m Repurposed Units 70.5m GB Memory Sold FY26 Technology-Led Growth 16.8m Units Repurposed A$172.7m Underlying EBIT 8 Evolution Into a High-Value Technology Platform
Page 9
Trust Creates a High Barrier to Entry CUSTOMER OUTCOME ECONOMIC OUTCOME Embedded customer integration Lower execution risk across recurring refresh programs. High retention and recurring volume. Global logistics and processing network One accountable partner across sites and regions. Scalable growth and operating leverage. Security, compliance and chain of custody Confidence that sensitive assets are managed securely. Long onboarding processes create barriers to entry. Proprietary processing and recovery intelligence More assets routed to their highest-value outcome. Higher recovery yield and value per asset. Global redeployment and resale channels Broader demand access and stronger price realisation. Greater margin resilience across asset cycles. Durable Relationships | Volume growth | Higher Recovery Margins | Attractive Capital Returns 9
Page 10
Scale the Circular Technology Platform Secure more assets, recover more value and scale efficiently Deepen Customer Integration Win New Customers Scale with Operating Discipline Higher EBIT and ROCE1 More Assets x More Value Per Asset x Operating Leverage = Sustainable Returns 10 Greater share of wallet Expand across services, sites and asset classes. More customer volume Convert proven performance into new contracts. Use automation, data and the global footprint to grow earnings faster than the capital base. Increase Value Per Asset Higher Margin Extend expertise from DDR4 into emerging, high- value technologies. 1 Return on Capital Employed (“ROCE”) = Underlying EBIT/Average Annual Capital Employed. Capital employed is defined as total a ssets minus current liabilities, excluding right-of-use assets and associated lease liabilities and current tax receivables/payables. DDR4 Demand Provides an Established Foundation for Platform Growth
Page 11
Sean Magann Chief Commercial Officer – SLS Global 11 Market Opportunity & Commercial Strategy
Page 12
Global Data Centre Market We Operate in a Growing Market Installed Infrastructure Growth Expands the Future Recovery Pipeline 12 1 Source: Wall Street research 2 Source: BNP Paribas Equity Research, SankeyMATIC, high level estimates subject to change given rapidly shifting costs and demand -supply dynamics AI CAPEX Flow2 $50.00 $15.00 $20.00 $7.50 $7.50 65% SLS Facilities & Construction Cooling Power Networking Equipment GPUs/Memory Where Does $100 of AI CAPEX Flow? $143 $256 $448 $820 $976 $1,047 2023A 2024A 2025A 2026E 2027E 2028E Hyperscaler CAPEX1 (US$ Billions)
Page 13
Manufacturers Prioritising Advanced Memory DDR4 Prices Expected to Remain Elevated DDR4 Price Outlook1 Primary supply shrinking. Structural shift away from DDR4. DDR5 production more profitable. DDR5 is not Backwards Compatible with DDR4 Upgrading requires a full platform overhaul, making it a costly and time-consuming process that many industrial and enterprise operators defer. Large Enterprise and Industrial Installed Base Millions of active servers, workstations and industrial controllers continue to use DDR4, requiring the same technology for maintenance, repairs and expansion. 13 1 TrendForce, August 2026. $30 $35 $40 $45 $50 $55 $60 Jul-26 Dec-26 May-27 Oct-27 DDR4 8 GB, 1Gx8, 3200DDR4-3200 1Rx8 (US$)DDR4-
Page 14
DDR4 Price Drivers Product mix shapes realised prices 14 Drivers of Sims Realised Prices Capacity Speed Rank Organisation Value increases nearly linearly with capacity 3200, 2933, 2666, 2400 MT/s — mega transfers per second Number of collective units Data width handled by each chip Samsung 32GB DDR4-2666 2Rx4 Manufacturer Module brand Samsung 32GB DDR4-2666 2R x4
Page 15
15 Jul25 Aug25 Sep25 Oct25 Nov25 Dec25 Jan26 Feb26 Mar26 Apr26 May26 Jun26 Jul26 Aug26 Primary Axis: Each scatter plot point represents the average used selling price of a specific Dimm by manufacturer part numbe r sold within the applicable month Secondary Axis: TrendForce DDR4-3200 1Rx8 new spot price DDR4 Average Selling Pricing per Used Dimm (US$) Capacity, speed and configuration determine how much of the supportive market backdrop converts into realised value $45 $40 $35 $30 $25 $20 $15 $10 $ 5 TrendForce pricing historically provided market context; however realised SLS prices vary materially based on capacity, speed, configuration and product mix $450 $400 $350 $300 $250 $200 $150 $100 $ 50 TrendForce DDR4-3200 1Rx8
Page 16
Scarcity Delaying New Builds 16 Lack of new builds delaying decommissioning As Bottlenecks Ease, Upgrade and Replacement Activity will Resume You Can’t Retire What You Can’t Replace Delays don't reduce decommissioning volume; they defer it, then compress it into a shorter, more intense cycle. Delayed and Compressed Decommissioning Decommissioning is inevitable, just delayed. Capital Approved Decommissioning activity is bottlenecked upstream. Assets cannot be retired unless new capacity can take over their workloads. Four P's PeoplePower Parts Perception
Page 17
Expanding Beyond DDR4 Multiple technology pathways extend the growth opportunity DDR4 DDR5 GPU/ Accelerator CPU CY2027 CY2031 Networking 17 Networking
Page 18
DDR5 Extends SLS’ Memory Expertise The same recovery workflow handles both DDR4 and DDR5 memory DDR5 Sales1 DDR5 Expands SLS’ Addressable Market Without Requiring An Entirely New Recovery Model 18 US$ Billions 1. TrendForce / SLS Estimate A Natural Capability Extension DDR5 is already entering SLS asset flows for testing • Existing DDR4 recovery expertise provides the foundation. • Testing protocols are being developed with a hyperscale customer. • Early preparation positions SLS for future volume growth. • Automation capable with both generations of memory. Build Capability Now. Scale as Volumes Grow $0 $50 $100 $150 $200 $250 $300 $350 2026 2027 2028
Page 19
Moving from Training to Inference Pre-2026, Majority of Spend on Training LLMs Inference • Using an already-trained model to produce outputs on new data. • Legacy GPUs are well-suited for inference workloads. Training 19 . "GPUs supporting yesterday's training requirements still serve tomorrow's inference workloads" Jensen Huang, Nvidia1 • Processing large amounts of data and adjusting so predictions get more accurate. • Demand for the most cutting-edge GPUs. Majority of Capital Spend on Inference Additional Market for GPUs Nvidia B100 Nvidia H100 1 https://introl.com/blog/secondary-gpu-markets-buying-selling-used-hardware-guide-2025
Page 20
GPU Recovery A Significant Installed Market is EmergingBuilding Capability Early 20 Demand From Enterprises, Neoclouds and Other Operators Could Support Secondary-Market Absorption Calculated Residual Value Based Upon GPUaaS Leasing Nvidia Data Centre Revenue US$ billions1 The model discounts a GPU's projected rental income over its remaining life minus production costs back to a present value $0 $50 $100 $150 $200 $250 2023 2024 2025 2026 • Material volumes expected in FY2028. NEW GPU Recertification Customer Access Timing • Early visibility into future GPU decommissioning. • Deeper penetration of hyperscalers. • Structured testing, workflows to recertify used GPUs. • Trusted performance drives higher market value. * As of September 10, 2026 A100 US$5,0002 H100 US$20,0002 1 NVIDIA 10-K data is presented on a fiscal-year basis. 2 GPU Performance Data for Companies https://portal.silicondata.com/signin?redirect=https%3A%2F%2Fportal.silicondata.com%2Fresidual -value
Page 21
21 GPU CPU $0 $20 $40 $60 $80 $100 $120 $140 2026 2027 2028 2029 2030 TAM (US$B) by Year Training AI Models Inference AI Models GPU vs CPU 8 to 1 ratio of work GPU vs CPU 1 to 1 ratio of work GPUs and CPUs – How Workloads Will Change Agentic AI Could Add US$32.5 Billion to US$60 Billion in New Data Centre CPU Demand by 2030 CPU Market Forecast (35%+ CAGR to 2030)1 GPU CPU 1 Based on AMD’s Q1 2026 earnings call, with internal adjustments. 2 Morgan Stanley Estimates. https://valorc3.com/ai-inference-edge-data-center-strategy/
Page 22
Networking 22 Switches and NIC Cards Retain Residual Value Network Processors Optics CableSwitches OCP Ethernet Switch 100G OCP NIC 3.0
Page 23
Breadth of Customer Platform Our client relationships extend across sites, services and refresh cycles Hyperscalers Enterprise OEMs Large, recurring data centre refresh programs. Laptops, network gear, mobile phones. Reverse logistics, product returns, channel management. Global Execution | Speed | Security Compliance | Chain of Custody | Value Recovery Flexible Capacity | Component Expertise | Market Access Customer Priorities Customer Priorities Customer Priorities 19 Circular Centres Growing and Diverse Base of Clients Leverage AI Pipeline 23 40% Of Resale Gross Margin 2.2 x YoY Increase FY26 FY26 Enterprise & OEMs
Page 24
24 Growing Customer Revenue Deeper engagement drives growth across existing and new customers Protect the Core Continue to Expand Service More Touchpoints Across The Lifecycle Expand services for self- processing hyperscalers: • Battery Backup Unit (BBU). • Data eradication. • Migration support. • GPU and hardware testing. • Forward supply: validated hardware sourcing. • Direct sales to NeoClouds. • Visibility into future refresh cycles. Broader Stakeholder Engagement and Multi-Year Agreements Deepen Customer Integration and Improve Revenue Visibility • Agreements aligned to refresh and decommissioning cycles. • Relationships span procurement, operations, infrastructure, and engineering. • We grow alongside our clients.
Page 25
Lynn Jacobs Chief Operating Officer - SLS Global 25 Operational Excellence & Scalability
Page 26
Operational Excellence Execution against customer-specific requirements at scale. 95–98% Hyperscaler Service Level Agreement (SLA) Performance Demonstrates Consistent Execution at Scale SECURE • Certified data destruction. • Auditable certificates. • End-to-end chain of custody. ACCURATE • 10m assets processed in FY26. • Full serialised asset tracking. • Real-time portal and customer system integration. ASSURED • Consistent testing and grading. • Customer-specific quality controls. • Transparent reporting and certification. REPEATABLE • Standardised processes and controls. • Consistent execution across regions. 5+ Million Devices Annually 99%+ DCIS Inventory Accuracy 26
Page 27
Operating Model Built for Scale Critical expertise remains stable while meaningful capacity is flexible and demand -linked Stable Core • Customer integration, site leadership and SLA management. • Technical and engineering expertise. • Quality and security capability. Flexible Capacity • Shift expansion. • Temporary and variable labour. • Cross-training across workflows. • Up to 600m Memory GB available in the current US hyperscaler footprint. • Added an additional 80-100m Memory GB capacity in EMEA. • Additional capacity can be added at other sites with minimal investment. Network Capacity • Workload balancing between SLS sites. • Qualified subcontractors. • Specialist technical partners. Repeatable Expansion • Common process standards. • Standard equipment design. • Geographic replication aligned with customers. Capital-Light Growth Model | A$18 Million Total Capital Expenditure in FY26 | ROACE1 118% 27 1 Return on Capital Employed (“ROACE”) = Underlying EBIT/Average Annual Capital Employed. Underlying EBIT divided by average capital employed (calculated as the average of opening and closing total assets less current liabilities, excluding cash and interest -bearing debt, for the period).
Page 28
Automation Already Delivering Value Reliability Quality Cost Scale ~96% uptime achieved. Automation is a more reliable operation model over manual processing. High-quality output and only ~1.5% defect rate. Automation protecting quality as volumes scale. ~47% reduction in cost for processing, not fully realised in all phases Fewer manual touches reduce cost and improve throughput. Enabled by modular design. ~5.4x increase in speed of Dimm sorting Automation expands practical capacity while reducing the risk that quality deteriorates as volume grows Proven Uniquely Designed Automation Expands Practical Capacity While Protecting Quality, Reliability and Traceability 28
Page 29
Complexity Raises The Operational Bar Increasing complexity strengthens SLS’ competitive advantage 29 Technology Change Operational Requirement Liquid-cooled infrastructure. Specialist disconnection, fluid management and safe handling. GPUs and advanced accelerators. Configuration-level identification, secure handling and more extensive testing. More integrated systems. Co-ordinated removal across compute, networking, storage, power and cooling. Greater variation in asset condition. Component-level assessment, grading and routing for reuse, resale or recycling. SLS’s Technical Capability Enables Safe Execution and Greater Value Recovery
Page 30
30 Customer Integration Keeps SLS Technically Ahead Early visibility allows SLS to develop and validate recovery processes while markets are still forming Customer Insight Shortens the R&D Cycle 2 Gain Insight See emerging assets, timing and requirement earlier. 3 Engineering & R&D Develop testing methods, equipment and secure handling procedures. 4 Live Validation Pilot processes using customer assets and operating data. 5 Operational Readiness Establish repeatable workflows, quality controls and trained teams. 6 Scale Service Replicate proven capability across sites as volumes develop. 1 Integrate Operate inside customer workflows.
Page 31
Advancing Opportunities in Emerging Technologies GPU testing infrastructure and validation frameworks are progressed, with further capability development underway 31 GPU Capability Established and Development Progressing Operational Reliability • Established a testing validation framework addressing functionality risks that may not surface through basic diagnostics. Testing Infrastructure • Engineered a Micro data centre to replicate standard operating conditions. Condition Validation • Developed ValidCore certification (White paper) and grading criteria that combine logical test results with physical condition to support informed disposition decisions. Operational Efficiency • Developing a purpose-built system to automate testing and integrate inspection results. DDR5 Capability in Development Customer Collaboration • Partnering with customers to access DDR5 units and develop recovery processes. Testing Capability • Built automated testing capability to support validation at scale. Process Development • Applying established memory-recovery expertise to develop DDR5 capability. GPU Testing Environment GPU Testing and Diagnostic Platform
Page 32
32 Controlled Execution Technically Ahead Scalable Operating Model Low Capital Intensity Common standards, traceability and assurance protect delivery across regions. Customer visibility and engineering capability prepare SLS for increasingly complex assets. Flexible capacity and modular automation support growth while managing volume variability. Existing infrastructure and modular automation support growth with limited incremental capital. Built To Scale Into The Next Technology Cycle
Page 33
Stephen Mikkelsen Sims Group CEO & Managing Director 33
Page 34
34 A Growing Market A Strong Customer Position A Broader Opportunity Operating Leverage More infrastructure creates a larger future asset-recovery pool. Embedded relationships support recurring volume and further expansion. DDR4, DDR5, GPUs and networking extend the growth runway. Automation and existing infrastructure support growth with limited incremental capital. Positioned To Scale A Capital- Efficient Growth Model Sustainable earnings growth with attractive capital returns.
Page 35
35 Questions & Answers