If you manage security cameras or dash cams, you have probably seen a microSD card die at the worst possible moment — right after an incident, with no warning. The card worked fine for months, then suddenly became unreadable. The footage you needed is gone.
This is not a manufacturing defect. It is the predictable result of flash memory wear, and it can be estimated in advance with basic math. This article gives you the formula, the variables, and a step-by-step method for estimating microSD card lifespan in the two most write-intensive consumer applications: security cameras and dash cams.
Why Continuous Recording Kills MicroSD Cards
Security cameras and dash cams use loop recording: when the card fills up, the oldest footage is overwritten with new footage. The card is written to continuously whenever the device is on. A 24/7 security camera is writing to the card 24 hours a day. A dash cam is writing whenever the vehicle is running.
Flash memory wears out with every write-and-erase cycle. Each NAND cell can only be programmed and erased a limited number of times before it can no longer reliably store data. Regular microSD cards are rated for light, intermittent writes — a few photos, occasional file transfers. When a regular card is used in a dash cam, it can rewrite its entire capacity every few days, burning through those cycles in months.
The solution is not a bigger card. It is understanding how much endurance your application actually needs, and choosing a card that meets that requirement.
The Core Formula: TBW = (Capacity × P/E Cycles) ÷ WAF
The industry-standard method for estimating flash storage endurance is TBW (Total Bytes Written). The formula is:
TBW = (Capacity × P/E Cycles) ÷ WAF
Where:
Capacity = the card’s usable capacity in gigabytes or terabytes
P/E Cycles = the NAND flash endurance rating (program/erase cycles per cell)
WAF = Write Amplification Factor, the ratio of physical writes to the flash versus logical writes from the host
This formula is published by KIOXIA and used across the flash storage industry. The result is the total amount of host data that can be written to the card before the NAND is expected to wear out.
Understanding WAF
Write amplification is the invisible multiplier that determines how much of your card’s endurance you actually get. It exists because the controller cannot simply overwrite a single cell when the host writes a small block of data. Instead, it must read an entire erase block, modify the relevant pages, and write the whole block to a new location. The old block is then erased — a process that consumes P/E cycles on both the old and new blocks.
WAF depends on the write pattern:
Sequential writes (video recording) typically produce WAF values close to 1 — the ideal case. Video data is written in large, sequential blocks, so the controller can write efficiently without excessive read-modify-write operations.
Random writes (data logging, database operations) produce WAF values of 2 to 5 or higher, because small writes force the controller to move larger blocks of data.
Mixed workloads fall somewhere in between.
For security cameras and dash cams, sequential video recording means WAF is typically low — usually between 1 and 2. This is one of the few advantages these applications have: the write pattern is friendly to the NAND.
Step 1: Estimate Daily Write Volume
Before you can estimate lifespan, you need to know how much data your device writes per day.
Security cameras
Daily write volume depends on resolution, frame rate, and whether the camera records continuously or only on motion events.
For continuous 24/7 recording, use the bitrate to calculate daily data:
Daily data (GB) = Bitrate (Mbps) × 3600 × 24 ÷ 8 ÷ 1024
Here are reference values for common configurations:
| Resolution | Typical Bitrate | Daily Write Volume |
|---|---|---|
| 1080p (H.265) | 2 Mbps | ~21 GB |
| 1080p (H.264) | 4 Mbps | ~42 GB |
| 2K (H.265) | 4 Mbps | ~42 GB |
| 4K (H.265) | 8 Mbps | ~84 GB |
| 4K (H.264) | 16 Mbps | ~168 GB |
Real-world measurements confirm these ranges. EZVIZ reports that a 1080p camera consumes roughly 5–7.5 GB per day for motion-triggered recording, while 4MP cameras under continuous recording can write 32–53 GB per day. Reolink’s storage tables show that a 4 Mbps stream produces approximately 42 GB per day of continuous recording.
For motion-triggered recording, the daily volume is much lower — typically 1.5–5 GB per day for a residential camera with average activity.
Dash cams
Dash cams typically record at 1080p or 4K while the vehicle is running. A typical commuter drives 1–2 hours per day. At a 10 Mbps bitrate:
1 hour/day: ~4.5 GB
2 hours/day: ~9 GB
Rideshare or fleet use (8+ hours): ~36 GB+
Some dash cams also record while parked (parking mode), which can add several more hours of low-bitrate or event-triggered recording per day
Step 2: Determine the Card's TBW Rating
If the card’s manufacturer publishes a TBW rating, use it directly. If not, calculate it from capacity and P/E cycles.
Reading the P/E cycles from the NAND type
The P/E cycle rating is determined by the NAND flash technology:
| NAND Type | Typical P/E Cycles | Example MicroSD Cards |
|---|---|---|
| QLC | 100–1,000 | High-capacity consumer cards (1TB+) |
| TLC (standard) | 500–1,000 | WD Purple QD101 (500 P/E cycles) |
| TLC (high-endurance) | 1,000–3,000 | Samsung Pro Endurance, SanDisk High Endurance |
| MLC | 3,000–5,000 | Industrial-grade cards |
| pSLC | 20,000–30,000+ | Kingston Industrial microSD (30K P/E cycles) |
Worked example: calculating TBW from scratch
Take a 128GB TLC card with 1,000 P/E cycles, recording sequential video (WAF = 1.5):
TBW = (128 GB × 1,000) ÷ 1.5 = 85,333 GB = ~85 TB
Now compare this to published ratings. The WD Purple 128GB surveillance card is rated at 60 TBW. The Sony 128GB industrial card is rated at 274 TBW. The difference reflects not just P/E cycles, but controller firmware, over-provisioning, and the quality of the NAND die.
Published TBW ratings for reference
| Card | Capacity | TBW Rating | P/E Cycles |
|---|---|---|---|
| WD Purple QD101 | 512GB | 256 TBW | 500 |
| WD Purple QD101 | 128GB | 60–128 TBW | 500 |
| SanDisk MAX Endurance | 128GB | 170 TBW | 1,000+ |
| Samsung Pro Endurance | 128GB | 70,080 hours (FHD) | 1,000+ |
| Kingston Industrial | 64GB | 1920 TBW | 30,000 (pSLC) |
Step 3: Calculate Expected Lifespan
Once you have daily write volume and TBW rating, the lifespan calculation is straightforward:
Lifespan (years) = TBW ÷ (Daily Write Volume × 365)
Worked example: 1080p security camera, continuous recording
Card: 128GB TLC, TBW = 60 TB
Daily write volume: 42 GB/day (4 Mbps continuous)
Annual write volume: 42 GB × 365 = 15,330 GB = ~15.3 TB/year
Estimated lifespan: 60 TB ÷ 15.3 TB/year ≈ 3.9 years
Worked example: 4K security camera, continuous recording
Card: 128GB TLC, TBW = 60 TB
Daily write volume: 84 GB/day (8 Mbps continuous)
Annual write volume: 84 GB × 365 = 30,660 GB = ~30.7 TB/year
Estimated lifespan: 60 TB ÷ 30.7 TB/year ≈ 2.0 years
Worked example: dash cam, 2 hours/day
Card: 128GB high-endurance TLC, TBW = 128 TB
Daily write volume: 9 GB/day (2 hours at 10 Mbps)
Annual write volume: 9 GB × 365 = 3,285 GB = ~3.3 TB/year
Estimated lifespan: 128 TB ÷ 3.3 TB/year ≈ 39 years
This illustrates an important point: dash cams that record only while driving are far less demanding than 24/7 security cameras. The same card that lasts 2 years in a 4K surveillance camera could last decades in a commuter dash cam.
The Dash Cam Problem: Why Cards Fail Faster Than the Math Predicts
If the math says a dash cam card should last 39 years, why do so many dash cam cards fail in 6–12 months?
There are three reasons the simple TBW calculation understates the risk:
1. Loop recording concentrates writes on the same blocks
In loop recording, the camera deletes the oldest clips and overwrites them. If the controller’s wear leveling algorithm is not aggressive enough, the same physical blocks are erased and rewritten repeatedly while other blocks remain relatively untouched. This creates localized wear that can kill the card long before its rated TBW is reached.
2. Heat accelerates wear
Dash cams sit on windshields, where summer temperatures can exceed 60°C. Heat increases charge leakage in NAND cells and reduces the effectiveness of error correction. A card operating at 60°C wears faster than the same card at 25°C, even with identical write volumes.
3. Low-quality cards lack the firmware to manage wear
The controller firmware — wear leveling, garbage collection, ECC strength, read-refresh — determines how close the card gets to its theoretical TBW. A cheap card with 1,000 P/E cycles but poor wear leveling may fail at 20% of its rated TBW. A high-endurance card with the same P/E rating but sophisticated firmware may exceed it.
This is why high-endurance microSD cards are not just marketing. They use better NAND die (higher-binned TLC or MLC), more aggressive wear leveling, and stronger ECC. For continuous-recording applications, the endurance premium is almost always justified.
A Practical Lifespan Estimation Checklist
Use this sequence when estimating lifespan for a specific deployment:
1. Determine the write workload.
What is the bitrate? (Check the camera’s recording settings.)
Is recording continuous or motion-triggered?
For dash cams, how many hours per day is the camera recording?
2. Calculate daily write volume.
Continuous: Bitrate (Mbps) × 10.8 = GB per day (approximately)
Motion-triggered: Use the camera manufacturer’s daily storage estimates or measure actual usage.
3. Identify the card’s TBW.
Check the datasheet. If no TBW is published, calculate from capacity and P/E cycles: TBW = (Capacity × P/E Cycles) ÷ WAF.
For sequential video recording, use WAF = 1.5 as a conservative estimate.
4. Calculate lifespan.
Lifespan (years) = TBW ÷ (Daily GB × 365 ÷ 1000)
5. Apply a safety margin.
Derate the calculated lifespan by 30–50% to account for heat, firmware quality, and real-world wear variation.
If the derated lifespan is shorter than your product’s intended service life, upgrade to a higher-endurance card.
6. Monitor card health if possible.
Some cards and cameras support S.M.A.R.T. monitoring or a health status register that reports remaining endurance as a percentage. This is the most reliable way to know when a card is approaching end-of-life.
If health monitoring is not available, replace the card proactively at 70–80% of its estimated lifespan.
Quick Reference: Estimated Lifespan by Application
The following table assumes a 128GB card with TBW ratings typical of each card class, and continuous recording where applicable.
| Application | Daily Write | Card Type | TBW | Est. Lifespan |
|---|---|---|---|---|
| 1080p security cam, motion only | ~3 GB | Standard TLC | 60 TB | ~54 years |
| 1080p security cam, 24/7 | ~42 GB | Standard TLC | 60 TB | ~3.9 years |
| 4K security cam, 24/7 | ~84 GB | Standard TLC | 60 TB | ~2.0 years |
| 4K security cam, 24/7 | ~84 GB | High-endurance TLC | 128 TB | ~4.2 years |
| Dash cam, 2 hr/day | ~9 GB | High-endurance TLC | 128 TB | ~39 years |
| Dash cam, 8 hr/day (fleet) | ~36 GB | High-endurance TLC | 128 TB | ~9.7 years |
| Dash cam, 2 hr/day | ~9 GB | Standard TLC | 60 TB | ~18 years |
Note: These estimates assume ideal conditions. Real-world lifespan may be 30–50% shorter due to heat, firmware quality, and loop-recording wear concentration.
Key Takeaways
TBW is the right metric for lifespan estimation, not capacity or speed class. Calculate it as (Capacity × P/E Cycles) ÷ WAF.
Continuous recording is the killer. A 4K security camera recording 24/7 writes ~84 GB per day, consuming a 60 TB card in about 2 years. Motion-triggered recording extends this by an order of magnitude.
Dash cams are less demanding than 24/7 cameras — but heat, loop recording, and cheap cards still cause premature failures. High-endurance cards with strong firmware are the safest choice.
WAF matters. Sequential video recording has low WAF (1–2), which is favorable. But poor wear leveling can still concentrate writes and cause localized failure.
Published TBW ratings vary widely. A 128GB card can be rated from 60 TBW to 274 TBW depending on NAND quality and controller firmware. For continuous-recording applications, the higher-rated card is almost always worth the premium.
Monitor card health when possible. If your camera supports S.M.A.R.T. or a health register, use it. If not, replace cards proactively at 70–80% of estimated lifespan.
Still Unsure Which Card Fits Your Application?
The lifespan estimate depends on your specific bitrate, recording schedule, thermal environment, and card choice. If you are selecting a microSD card for a security camera or dash cam product and want a straight answer, our application engineers can review your workload and recommend the right endurance class.
Tell us about your application:
Camera resolution and bitrate
Continuous or motion-triggered recording
Operating temperature range
Expected product service life
Leave your email below. Our team will reach out with a written assessment — including a TBW calculation for your specific workload and a card recommendation. No obligation, no generic brochure.