Bao'An Shenzhen China

IC Unlock Success Rate | How Every Chip Is Assessed

Published · About 7 min read · PCB-Copy

On this page
  1. What Is a Realistic IC Unlock Success Rate?
  2. Why IC Unlock Success Rate Varies So Much
  3. Success-Rate Ranges by Chip Category
  4. How We Assess a Chip Before Accepting the Job
  5. Factors That Clients Can Control
  6. Real-World Success Snapshots
  7. When the Answer Is “No”
  8. Frequently Asked Questions
  9. Related IC Unlock Resources

What Is a Realistic IC Unlock Success Rate?

The IC unlock success rate across the industry typically falls between 70 % and 95 % for well-characterized chip families, but that single number hides enormous variation. A serial EEPROM with no read-protection fuse has a near-100 % extraction rate, while a latest-revision Cortex-M33 MCU with TrustZone and active tamper detection may sit below 30 %. The only honest answer is: it depends on the specific part number, its silicon revision, and the protection mechanism the manufacturer implemented.

Our lab maintains a running feasibility database covering more than 4 000 device variants. Before any engagement begins, we cross-reference your part marking against that database, inspect the physical device if needed, and issue a feasibility verdict — pass, conditional, or decline — at no charge. That front-loaded assessment is what keeps our completed-project success rate above 90 %: we simply do not accept jobs we cannot finish.

Why IC Unlock Success Rate Varies So Much

ic unlock success rate: Microcontroller under stereo microscope during IC unlock feasibility inspection

Protection Scheme Complexity

Chip vendors implement protection at multiple levels. A basic one-time-programmable (OTP) read-protection bit on an 8-bit MCU is a different challenge from a multi-zone, multi-level RDP scheme on an STM32 or the proprietary lock bits on a Renesas R5F series. The more layers a vendor adds — voltage glitch detection, bus encryption, metal shield layers — the lower the statistical success rate for any lab.

Silicon Revision

A technique that works on Rev A silicon may fail on Rev C. Manufacturers patch vulnerabilities between mask revisions without changing the part number. Our engineers check die markings under magnification; a single letter suffix can shift the success probability from 95 % down to 40 %.

Physical Condition of the Device

Corrosion, ESD damage, lifted bond wires, or prior failed decap attempts all reduce the chance of a clean readout. If the flash array is partially degraded, even a successful unlock yields corrupted firmware. We photograph and inspect every incoming device, and we flag condition risks before committing lab time.

Lab Method Selection

Different protection types demand different attack vectors — UV exposure, microprobing, voltage fault injection, laser fault injection, or chemical decapsulation followed by FIB editing. Choosing the wrong method wastes the sample. Our assessment step matches the device to the method with the highest documented yield for that specific revision.

Success-Rate Ranges by Chip Category

Chip Category Example Families Typical Success Range Key Limiting Factor
Serial EEPROM (I²C / SPI) 24C256, AT24C512, 93C56, M95080 92 %–99 % Physical damage; rare write-protect fuses
SPI NOR Flash W25Q64, SST25VF 90 %–98 % OTP lock bits; encrypted payloads
8-bit / 16-bit MCU PIC16, ATmega, MSP430 80 %–95 % Fuse type; die revision patches
Cortex-M0/M0+ MCU MKE02, GD32E230, EFM32HG 75 %–92 % RDP level; glitch-detection countermeasures
Cortex-M3/M4 MCU STM32F4, MK60, nRF52832, CY8C 60 %–88 % Multi-level RDP; JTAG disable; secure boot
Cortex-M4F / M33 MCU EFM32PG, MAX32660 40 %–75 % TrustZone; active tamper mesh; die shields
DSP / FPGA / Secure Element TMS320, Xilinx Spartan, ATECC608 20 %–50 % Bitstream encryption; anti-fuse arrays

These ranges reflect aggregate outcomes across multiple silicon revisions and lab methods. Your specific part may land higher or lower. That is exactly why the feasibility check exists.

How We Assess a Chip Before Accepting the Job

Step 1 — Part Identification & Database Lookup

You send us the full part marking (or a clear photo). We decode manufacturer, family, memory size, package, and — critically — the lot/date code that hints at the silicon revision. Our internal database returns a feasibility score and the recommended attack vector.

Step 2 — Physical Inspection (When Needed)

If the database score is conditional, or the device has been exposed to harsh conditions, we request the physical sample. Under a stereo microscope at 40×–100× magnification, we check for die cracks, corrosion on exposed pads, and signs of prior decap. Devices that fail this inspection are flagged before any lab cost accrues.

Step 3 — Protection-Level Verification

On MCUs with configurable protection, the actual level set by the original programmer matters. An STM32 at RDP Level 1 is a very different job from RDP Level 2. Where possible, we probe the debug port non-destructively to confirm the active protection state. This step alone can shift the predicted success rate by 20 percentage points.

Step 4 — Feasibility Verdict

We issue one of three outcomes:

  • Pass — high confidence (≥ 85 %). We quote a fixed price, and you pay only on successful readout.
  • Conditional — moderate confidence (50 %–84 %). We quote with a clear risk disclosure and often a reduced fee structure reflecting the uncertainty.
  • Decline — confidence below our threshold, or the device uses a protection class we do not currently support. No charge, no wasted samples.

This three-tier model is why our post-acceptance success rate stays above 90 %, even though the raw, all-comers rate for the industry hovers around 70 %–80 %. We filter before we cut. If you want to start this process, submit a free feasibility request with your part details.

Factors That Clients Can Control

Not every variable is on the lab side. Here is what you can do to maximize the IC unlock success rate for your project:

  • Provide multiple samples. Two or three units let us attempt different methods without total loss. For conditional-tier devices, a second sample can raise effective success from 65 % to over 85 %.
  • Share known firmware version or compiler. If you originally programmed the device and have partial source or a hex map, post-readout verification is faster and more reliable.
  • Preserve physical condition. Ship in anti-static packaging. Do not attempt home decap with acetone or a heat gun — botched decaps are the number-one cause of unrecoverable devices in our intake log.
  • Confirm authorization early. We require proof of ownership or written authorization before starting work. Having that documentation ready avoids delays. Read more about the legal framework for IC unlock and the jobs we decline.

Real-World Success Snapshots

IC samples and lab equipment prepared for chip unlock assessment

NXP Kinetis MK60 — Cortex-M4, RDP Active

A client needed firmware recovery from an MK60 series MCU on a legacy industrial controller. The device was at JTAG-disabled state with flash security enabled. Our database rated it conditional (78 %). Using voltage fault injection on the boot sequence, we achieved a clean 512 KB readout in 4 working days. Two samples were provided; the first succeeded on the third glitch attempt.

Winbond W25Q64 — SPI Flash, OTP Lock

An OEM sent a W25Q64 SPI flash from a discontinued networking appliance. The status register showed OTP write-protect bits set, but no hardware read-lock. Direct SPI readout returned the full 8 MB image on the first pass — a textbook high-confidence case completed in under 24 hours.

Cypress CY8C — PSoC with Flash Protection

A CY8C PSoC device with user-set flash protection required microprobing after chemical decapsulation. Success rate for this family sits around 72 %–80 % depending on die revision. The client supplied three samples; the second yielded a verified 64 KB dump after bond-wire rerouting under FIB.

When the Answer Is “No”

Transparency matters more than a sale. We decline engagements when:

  • The protection scheme has no known viable attack (e.g., certain secure elements with active mesh and encrypted NVM).
  • The device is physically destroyed beyond recovery — cracked die, vaporized bond pads.
  • The client cannot demonstrate legitimate ownership or authorization.
  • Legal restrictions in the destination country prohibit the work.

Declining these jobs is what keeps our published success metrics honest. A lab that claims 99 % success on all devices is either cherry-picking numbers or not disclosing failed attempts.

Frequently Asked Questions

What is the average IC unlock success rate across all chip types?

Across our full intake log — including EEPROMs, flash memories, and MCUs from 8-bit to Cortex-M4 — the aggregate success rate on accepted jobs is approximately 91 %. The raw inquiry-to-success rate (including declined jobs) is closer to 72 %. The gap reflects our pre-screening: we decline roughly 20 % of inquiries at the feasibility stage.

Do I pay if the IC unlock attempt fails?

For high-confidence (Pass) engagements, our standard terms are success-based: you pay only when we deliver a verified readout. For conditional-tier jobs, we negotiate a reduced-risk fee structure upfront, typically a small lab-access fee plus the full fee on success. Declined jobs incur no cost at all.

How many samples should I send to maximize success?

We recommend a minimum of two identical samples for any MCU unlock and three for conditional-tier devices. Each additional sample raises the cumulative probability of at least one successful extraction. For serial EEPROMs and SPI flash, one sample is usually sufficient.

Does silicon revision really affect the outcome that much?

Yes. We have documented cases where a single-letter revision change (e.g., Rev B to Rev D on an STM32F1) shifted our internal success score from 90 % to 55 %. Manufacturers specifically patch fault-injection and UV vulnerabilities between revisions. Always send us the full lot code so we can check before you ship.

Is IC unlock legal for my situation?

IC unlock is legal when performed on devices you own or are authorized to access — for legacy maintenance, firmware recovery, or second-source qualification. We require proof of ownership and operate under NDA. We decline any engagement that appears to target third-party proprietary IP. See our detailed legal and authorization policy.

Related IC Unlock Resources

Check feasibility for your project

Send the chip marking or board details and tell us what you need to recover. Add clear photos if available.

Request a feasibility review
TURN THE GUIDE INTO A PROJECT QUESTION

What do you need to recover?

Send the full marking or board photos, sample condition and required outcome. Ask for an NDA before sharing confidential files. Scope and sample handling are confirmed individually.

The article title is carried into the contact form. Add your own details and review the enquiry before sending.

WhatsApp Send board details