Network equipment loses 5 to 10% of resale value every month after OEM End-of-Sale. The math, the lifecycle events that accelerate it, and how to time retirements for maximum recovery.
TL;DR
A $20,000 Cisco Nexus 9500 retired at OEM End-of-Sale recovers $12,000 to $14,000 through specialist channels. The same switch, retired 18 months later, recovers $4,000 to $6,000. That’s not depreciation. That’s recovery decay, and it’s predictable enough to plan around.
Five forces drive the curve:
- OEM lifecycle events (EoS announcements, Last Date of Support) cut value in step-functions, not gradually
- New product releases compress secondary-market pricing within 30 to 90 days of launch
- Refresh cycle waves at large enterprises flood secondary markets with similar equipment
- Supply chain shifts can spike value (current-gen GPU constraints) or collapse it (oversupplied storage)
- Regulatory or standards changes can render specific equipment unsellable overnight
The organizations that retire equipment in the 0 to 12 month window after OEM End-of-Sale capture 50 to 70% of original value. The organizations that wait 24+ months capture 15 to 30%. The difference is the cost of waiting, and on a meaningful refresh project that gap is millions of dollars in recovered capital lost.
This guide covers the decay curve by equipment category, the lifecycle events that move it, the hidden costs of inventory delay, and a decision framework for when to act.
The Network Equipment Recovery Decay Curve
The single most useful concept for thinking about asset recovery timing is the decay curve. Network equipment doesn’t lose value gradually like a depreciating asset on a balance sheet. It loses value in steps tied to specific events, and the curve is steep.
A baseline working model for enterprise routing and switching in 2026:
| Timeline Relative to OEM End-of-Sale | Typical Recovery as % of New Price |
|---|---|
| Pre-EoS (still currently selling) | 30–50% (new is still available; demand is limited) |
| EoS announcement to EoS effective date (3–6 months) | 35–55% (early secondary-market window) |
| 0–12 months post-EoS | 50–70% (peak secondary-market value) |
| 12–24 months post-EoS | 35–55% (decay accelerates) |
| 24–36 months post-EoS | 20–40% (significant value loss) |
| 36+ months post-EoS | 10–25% (approaching scrap) |
| Past Last Date of Support | Under 15% (no vendor support available) |
The shape of this curve is the part most IT leaders underestimate. There’s a single peak window (0 to 12 months post-EoS) where equipment recovers far more than at any other point in its lifecycle. Miss that window and you’re typically recovering 30 to 50% less, on the same equipment, in the same condition.
The peak window exists because of how the secondary market works:
- Buyers can still get vendor support. Once support windows close, demand drops sharply
- New equivalents are no longer available from the OEM. Buyers who need this generation have to buy used
- Refurbishers and specialists are actively rebuilding inventory for their own customers
- The replacement generation hasn’t fully compressed secondary-market pricing yet
Once those conditions shift (typically 12-18 months post-EoS), the curve starts to fall.
The Events That Move the Curve
The decay curve isn’t smooth. It moves in steps triggered by specific events. The events that matter:
OEM End-of-Sale Announcements
The largest single value mover. When Cisco, Juniper, Arista, Nokia, or another major OEM announces End-of-Sale for a product, two things happen simultaneously:
- Buyers who were going to refresh anyway accelerate their purchases (driving up demand for the soon-to-be-EoS equipment)
- Buyers who would have bought new shift to used (creating demand for secondary-market inventory)
This is the start of the peak recovery window. Equipment retired during the 6-12 months after EoS announcement typically recovers the highest secondary-market values.
OEM Last Date of Support
The end of the peak window. Once a product passes its Last Date of Support, vendor patches, security updates, and TAC support are no longer available. Demand drops sharply because most enterprise buyers won’t deploy equipment without support. Recovery values typically drop 30 to 50% in the 6 months following LDoS.
Replacement Product Releases
A new product launch from the OEM compresses secondary-market pricing on the predecessor within 30 to 90 days. The compression is more severe when the replacement offers significant cost-per-port or cost-per-throughput improvements. A buyer who was paying $14,000 for a used previous-gen switch may now buy the new generation at $18,000, which compresses the used market for the predecessor.
Refresh Cycle Waves
Large enterprises, hyperscalers, and carriers refresh equipment on coordinated cycles. When a tier-1 carrier or hyperscaler offloads thousands of units of the same equipment category in a single quarter, secondary-market pricing temporarily drops as supply exceeds demand. Recovery typically rebounds within 60-120 days as the supply absorbs, but equipment sold during the wave window recovers materially less than equipment sold before or after.
Supply Chain Disruptions
Both directions matter. When new equipment supply tightens (semiconductor shortages, OEM allocation constraints), demand for current-generation used equipment spikes and recovery values climb. This was visible in 2021-2022 during global semiconductor shortages and again in 2024-2025 during GPU allocation constraints. When supply chains normalize, used pricing typically returns to baseline within 90-180 days.
Regulatory and Compliance Changes
The most disruptive event type, though less frequent. A regulatory change, security vulnerability disclosure, or standards update can render specific equipment effectively unsellable overnight. Examples: equipment containing components from sanctioned manufacturers, equipment with disclosed unpatched vulnerabilities, equipment that no longer meets specific industry compliance requirements. These events typically can’t be predicted, but they’re a real source of recovery value loss for equipment held in inventory.
Equipment Categories: How the Curve Varies by Type
Not all network equipment follows the same decay curve. The variations matter for retirement planning:
Equipment with Steep Decay Curves (Sell Fast)
Enterprise switches and routers (Cisco Nexus, Cisco ASR, Juniper MX, Arista 7000 series)
| Timeline | Typical Recovery |
|---|---|
| 0-12 months post-EoS | 50-70% |
| 24 months post-EoS | 30-45% |
| 36 months post-EoS | 15-30% |
The decay is steep because the replacement generation typically offers significant capability improvements, and buyer pools shift to new architectures relatively quickly.
GPU and AI accelerators (NVIDIA H100, H200, A100)
| Timeline | Typical Recovery |
|---|---|
| 0-12 months post-release | 75-90% (supply-constrained) |
| 12-24 months post-release | 50-70% |
| 36+ months post-release | 25-40% |
GPU recovery is dramatic because of allocation constraints and the rapid replacement cycle. The same gear that recovers 80% of new price in year one can recover 30% in year three.
Storage systems (NetApp, Pure, Dell EMC current generation)
| Timeline | Typical Recovery |
|---|---|
| 0-12 months post-EoS | 30-50% |
| 24+ months post-EoS | 15-30% |
Storage decay is more aggressive than networking because of how rapidly capacity and performance specifications improve.
Equipment with Gentler Decay Curves (More Forgiving Windows)
Optical transport (Ciena 6500, Infinera DTN-X, Nokia, ADVA, Fujitsu)
| Timeline | Typical Recovery |
|---|---|
| 0-24 months post-EoS | 40-60% |
| 24-48 months post-EoS | 25-45% |
| 48+ months post-EoS | 15-30% |
Optical equipment has longer practical service lives and stronger international carrier demand, which extends the recovery window beyond what’s typical for enterprise networking.
Carrier-grade routing (Cisco ASR 9000, Juniper PTX, Nokia 7750)
| Timeline | Typical Recovery |
|---|---|
| 0-24 months post-EoS | 45-65% |
| 24-48 months post-EoS | 25-45% |
Carrier-grade equipment moves through tier-2/3 carriers and international operators with longer secondary-market windows than enterprise gear.
Equipment with Limited Recovery (Don’t Wait, But Don’t Expect Much)
RAN equipment (baseband units, remote radio heads, antennas)
OEM software licensing constraints, technology-specific lock-in, and limited buyer markets cap recovery at 5-20% of new value regardless of timing. The right strategy here isn’t to time the retirement but to set realistic expectations.
End-of-life proprietary appliances
Equipment tied to discontinued software typically has no recovery path. The right strategy is acceptance and recycling.
The Hidden Costs of Inventory Delay
The decay curve is the visible cost. Several others compound it.
Storage and Inventory Carrying Costs
Equipment sitting in a storage cage, decommissioned-but-not-sold, costs more than people realize. A working cost model:
- Data center storage: $50-200 per rack per month depending on facility class
- Insurance and security: Variable, but 0.5-2% of equipment book value annually
- Internal tracking and documentation overhead: Often invisible but real
- Opportunity cost of square footage: Particularly significant in expensive data center markets
A retired half-rack of equipment that sits for 12 months waiting to be sold has typically cost $1,500-4,000 in carrying costs before any recovery transaction occurs. On a $40,000 equipment lot, that’s 4-10% of recovery value gone before you start.
Documentation and Configuration Decay
Equipment retired and stored often loses its accompanying documentation, configuration files, and accessories over time. A switch retired with full documentation and original accessories recovers 10-20% more than the same switch retired without those materials. The decay isn’t just in the equipment; it’s in the supporting materials that make the equipment sellable at full value.
Internal Process Decay
The longer equipment sits, the higher the probability that the people who knew its configuration, deployment history, and condition have changed roles, left the company, or forgotten the details. Equipment that goes to recovery without clear history often sells at discount because buyers can’t verify provenance.
Opportunity Cost on Recovery Capital
Recovery value that could have been redeployed against next-cycle CapEx 12 months ago is recovery value that wasn’t available for that purpose. On organizations with active refresh planning, the opportunity cost of delayed recovery can exceed the recovery decay itself.
Market Signals That Should Trigger Action
The framework most organizations need is a watchlist of trigger events. When any of these fires, equipment retirement and sale should accelerate:
| Signal | Action |
|---|---|
| OEM announces EoS on production equipment | Add to retirement planning watchlist; target sale in 6-18 month window |
| OEM announces replacement product | Evaluate refresh timing against secondary-market compression |
| Major industry refresh wave (visible in OEM earnings, trade press) | Accelerate sales ahead of supply wave |
| Supply chain disruption affects new equipment supply | Premium pricing window for current-gen used; act fast |
| Equipment passes EoS without retirement plan | Schedule for retirement within 12 months |
| Equipment approaches Last Date of Support | Sell now or accept dramatic recovery loss |
| Production capacity exceeds demand by 25%+ | Retire excess immediately; storage costs exceed expected recovery decline |
| Internal budget cycle approaches | Recovery timing should align with capital deployment timing |
The discipline isn’t watching all of these constantly. It’s building a quarterly review process that checks status against the watchlist and triggers action when conditions are met.
The Compounding Effect Across Refresh Cycles
Recovery timing isn’t a one-time decision. Across multiple refresh cycles, the difference between disciplined timing and reactive retirement compounds significantly.
A working model for a mid-size enterprise running 4-year refresh cycles on $5M of network equipment:
| Approach | Average Recovery Per Cycle | 12-Year Cumulative Recovery |
|---|---|---|
| Reactive retirement (sell when convenient, often 24+ months post-EoS) | 25% of new value = $1.25M | $3.75M |
| Disciplined retirement (sell in 0-12 month post-EoS window) | 55% of new value = $2.75M | $8.25M |
| Difference | $1.5M per cycle | $4.5M cumulative |
The compounding gap is the practical case for treating recovery timing as a planning discipline rather than a reactive process. Over a decade, the difference between disciplined and reactive timing on a mid-size enterprise can fund a meaningful portion of a refresh cycle.
For carriers and hyperscalers operating at 10-100x this scale, the math is the same shape but the absolute numbers are dramatically larger.
A Decision Framework: When to Act
For each piece of retired or about-to-be-retired equipment, three questions usually settle the timing decision:
Question 1: Where is this equipment in its OEM lifecycle?
- Pre-EoS, still in production: Hold until EoS announcement (or sell now if equipment is genuinely surplus)
- Within 12 months of EoS: Sell now (peak recovery window)
- 12-24 months post-EoS: Sell within 3-6 months (accelerating decay)
- 24+ months post-EoS: Sell immediately (continued delay loses material value)
- Past LDoS: Sell whatever can be sold; recycle the rest
Question 2: What’s the cumulative cost of waiting another quarter?
For specific equipment, calculate:
- Expected recovery value today
- Expected recovery value in 90 days (typically 3-7% lower for enterprise networking)
- Storage and inventory costs for 90 days
- Opportunity cost on the recovery capital
If the 90-day delay costs more than 5% of equipment value, sell now.
Question 3: Are there active market signals indicating value movement?
Check the watchlist above. If any signal is fired (OEM events, supply chain, refresh waves), the timing decision shifts toward immediate action.
These three questions take 30 minutes per equipment category to answer with reasonable confidence. The output is a clear “act now” or “schedule for next quarter” decision per category, supported by reasoning that survives finance review.
How to Time Retirements for Maximum Recovery
The operational practices that translate the framework into actual capital recovery:
Build OEM Lifecycle Tracking Into Asset Management
Most enterprises don’t systematically track OEM lifecycle status against their installed base. Quarterly tracking of EoS dates and LDoS dates for the full networking inventory is the foundation. Several asset management platforms (ServiceNow, Lansweeper, Device42) can pull OEM lifecycle data; the effort is modest.
Pre-Build the Vendor Relationship
The wrong time to interview asset recovery vendors is during an active retirement. Set up the specialist relationship during normal operations. Test on a small project. Understand the documentation flow. Negotiate engagement model terms. When the trigger event fires, you can move within days instead of negotiating for 4-6 weeks first.
Operate on Engagement Model Discipline
For equipment in the peak recovery window, outright buyback or consignment typically delivers the best outcomes. For equipment in slower decay categories, consignment can extend the marketing window without significant carrying cost. The right engagement model is set by equipment timing, not by vendor convenience.
Document Equipment During Production, Not at Retirement
Equipment retired with full documentation (configuration files, deployment history, accessory inventory, photographs of condition) recovers 10-20% more than equipment retired without. Building documentation discipline into the production lifecycle (rather than retroactively at retirement) pays back materially.
Match Recovery Timing to Capital Deployment
Recovery capital from equipment retirement is more valuable when it’s available against active CapEx commitments. Aligning retirement timing to next-cycle equipment purchase timing means the recovered value is actually deployed against the refresh, not held as accounting receipts. This is finance discipline more than IT discipline, but the IT decisions enable or constrain it.
Frequently Asked Questions
How much value does network equipment lose by waiting to sell?
Enterprise networking equipment typically loses 5 to 10% of resale value per month after OEM End-of-Sale, though the curve isn’t linear. The biggest single drop usually happens around 18-24 months post-EoS as Last Date of Support approaches and buyer support concerns surface. Equipment retired in the 0-12 month post-EoS window typically recovers 50-70% of new value; equipment retired 24+ months post-EoS typically recovers 15-30%. The 30-50 percentage point difference is the cost of waiting.
What is OEM End-of-Sale and why does it matter for resale value?
OEM End-of-Sale (EoS) is the date a manufacturer stops selling new units of a specific product. The EoS announcement is typically the start of the peak secondary-market window for that product because new equipment is no longer available from the OEM, but vendor support continues for several years afterward. Equipment retired in the 0-12 month window after EoS typically recovers the highest secondary-market value because buyers who need that generation can still get vendor support but can’t buy new from the OEM.
What’s the difference between OEM End-of-Sale and Last Date of Support?
End-of-Sale (EoS) is when the OEM stops selling new units of a product. Last Date of Support (LDoS) is when the OEM stops providing technical support, security patches, and software updates. The gap between EoS and LDoS is typically 3 to 7 years depending on the OEM and product. The peak secondary-market value window is during this gap (after EoS but well before LDoS). Once equipment passes LDoS, recovery values typically drop 30-50% because most enterprise buyers won’t deploy unsupported equipment.
When should I sell decommissioned network equipment?
The general answer is: within 0-12 months of OEM End-of-Sale for enterprise networking and switching, within 18-24 months for carrier-grade routing and optical transport, and as quickly as possible for GPUs and AI accelerators. Specific timing depends on OEM lifecycle status, current secondary-market conditions, and whether any trigger events have fired (new product launches, supply chain disruptions, refresh waves). Most enterprises should run a quarterly review against OEM lifecycle data to identify equipment approaching retirement windows.
Does waiting really cut equipment value in half?
In many cases, yes. A $20,000 enterprise switch retired in the peak window (0-12 months post-EoS) typically recovers $12,000-$14,000 through specialist channels. The same switch retired 36 months later typically recovers $4,000-$6,000. That’s not unusual, that’s the expected curve for enterprise networking. The pattern is even more dramatic for AI accelerators and GPUs, where 12-month delays can cut recovery by 50-70%.
What network equipment loses value fastest?
GPUs and AI accelerators have the steepest decay curves in 2026 because of rapid replacement cycles and allocation-driven supply dynamics. Enterprise switches and routers (Cisco Nexus, Juniper QFX, Arista 7000 series) decay faster than carrier-grade or optical equipment. Storage systems decay aggressively because capacity and performance specs improve rapidly between generations. RAN equipment has a near-flat decay curve at low values because secondary-market demand is limited regardless of timing.
What network equipment holds value longer?
Optical transport (Ciena, Infinera, Nokia, ADVA, Fujitsu) has the gentlest decay curves because of long practical service lives and international carrier demand. Carrier-grade routing (Cisco ASR 9000, Juniper PTX, Nokia 7750) holds value longer than enterprise routing because of tier-2/3 carrier and international operator demand. Microwave backhaul equipment has stable demand from specific buyer markets. Equipment from less common OEMs sometimes holds value longer because secondary-market supply is constrained.
How do I know when to sell my network equipment?
Three signals usually settle the question: OEM lifecycle status (where is this equipment relative to EoS and LDoS), active market triggers (supply chain disruptions, refresh waves, new product launches), and internal operational signals (is the equipment actually retired, is it taking up valuable rack space). When equipment is past EoS, when a market trigger has fired, or when internal carrying costs are exceeding 5% of equipment value annually, sell.
Can I sell network equipment that’s already past Last Date of Support?
Yes, but recovery values are typically under 15% of new price. Past-LDoS equipment moves to specific buyer markets: international operators with longer hardware reliability tolerance, refurbishers who can repurpose components, lab and development environments, and parts recovery for spare inventory at sites still running the equipment. The right channel matters: generic ITAD vendors typically can’t move this equipment at meaningful value, while specialists with parts and international networks sometimes can.
What’s the right partner for equipment recovery timing decisions?
The right partner is one that proactively tracks your installed base against OEM lifecycle data and notifies you when peak recovery windows approach, not one that only responds when you call. Specialist asset recovery firms with carrier and hyperscaler buyer networks typically deliver 30-50% more recovery value than generalist ITAD providers because of direct buyer relationships. The right partner also has multiple engagement models (buyback, trade-in, consignment, cost-offset decommissioning) so the right model can be matched to the right timing situation.
The Bottom Line
Network equipment recovery value isn’t a vague depreciating asset. It’s a curve with specific shapes, specific events that move it, and specific timing windows that determine the recovery percentage. The organizations that treat this as a planning discipline capture 2 to 3 times the recovery value of organizations that retire equipment when it stops being convenient to keep.
The framework isn’t complicated:
- Track OEM lifecycle status against your installed base quarterly
- Build a watchlist of trigger events that should accelerate retirement
- Establish a specialist asset recovery relationship before you need it
- Operate on equipment-specific timing rather than convenience timing
- Document equipment during production, not at retirement
- Match recovery timing to capital deployment timing
What it requires is recognition that retired infrastructure has value that decays predictably, that the channel through which it’s recovered matters, and that disciplined timing produces compounding returns across multiple refresh cycles. The cost of treating retirement as an operational afterthought is millions of dollars across a refresh program. The cost of treating it as a planning discipline is a quarterly review meeting.
The clock is ticking on every piece of equipment in your installed base today. Some of it is in its peak recovery window right now. Some of it has already passed peak and is in active decay. The organizations that know which is which are the ones that capture the value before it disappears.
How ROC Telecom Helps
ROC Telecom is an R2v3, RIOS, NIST 800-88, and ITAR-compliant asset recovery specialist with the buyer network and operational discipline to capture peak recovery windows:
- Active OEM lifecycle tracking integrated into retirement planning, so peak recovery windows are identified before they close
- Specialist per-asset valuation across Cisco (ASR, NCS, Nexus), Juniper (MX, PTX, QFX), Arista, Ciena, Infinera, Nokia, ADVA, Fujitsu, Calix, and NVIDIA platforms
- Direct buyer relationships across carriers, hyperscalers, neoclouds, and international operators (not broker channels), which is what enables 30-50% recovery premiums vs generalist ITAD
- All four engagement models (outright buyback, trade-in credit, consignment, cost-offset decommissioning) matched to your timing situation
- 48-hour rapid-response mobilization nationwide for compressed-timeline retirements where a window is closing
- Cost-offset model that often delivers data center decommissioning at no net cost on the right project volume
- Full documentation including serialized Certificates of Destruction, mass-balance recovery reports, and Scope 3 reporting inputs
15+ years of ITAD experience, $25M+ in client capital recovered for clients, 45M+ pounds diverted from landfill.
If you have retired or about-to-be-retired networking equipment sitting in storage, every quarter it sits is a meaningful percentage of recovery value walking away. A 30-minute valuation conversation can show you what the equipment is worth today vs. what it’s likely to be worth in 6 or 12 months, and what the right engagement model is for your situation. Prefer to talk directly? Call 585-406-1249 or email info@roctelecom.com.
"*" indicates required fields
Related reading:
- Data Center Asset Recovery: The 2026 Strategic Guide
- How to Sell Decommissioned Network Equipment: The Enterprise Buyer’s Guide for 2026
- How Long Does Networking Equipment Really Last? (2026 Data & OEM Lifecycle Guide)
- The 10 Data Center Liquidation Mistakes That Drain Your IT Budget
- Why AI Is Shortening Your Hardware Lifecycles (and What to Do About It)
- Top 10 ITAD Companies for Data Center Decommissioning & Asset Recovery in 2026
- ITAD for Telecom and Mobile Operators: The 2026 Strategic Guide