---
title: "Secure Element Chips in Hardware Wallets: EAL Ratings vs Open Source"
description: "Hardware wallets rely on tamper-resistant chips or transparent code to protect private keys. Here is how specialized silicon defends against physical attacks, and how open-source designs navigate the trade-off."
url: https://basisdesk.news/learn/secure-element-chips-hardware-wallets
published: 2026-11-20T20:30:31.893Z
modified: 2026-11-20T20:30:31.893Z
section: Security & Hacks
author: Basis Desk Newsroom (AI-generated, source-verified)
sentiment: neutral
tickers: [BTC, ETH]
tags: [hardware wallets, secure element, common criteria, eal6, eal5, cryptography, open source, fault injection, side channel attacks, self custody]
license: Quote with attribution to Basis Desk (basisdesk.news). Not financial advice.
---

# Secure Element Chips in Hardware Wallets: EAL Ratings vs Open Source

Hardware wallets rely on tamper-resistant chips or transparent code to protect private keys. Here is how specialized silicon defends against physical attacks, and how open-source designs navigate the trade-off.

## Key points

- Secure Elements are specialized microchips built with active shields, sensors, and bus encryption to stop physical and side-channel extraction of private keys.
- Common Criteria EAL ratings (such as EAL5+ and EAL6+) measure the depth and rigor of physical penetration testing conducted by independent labs.
- Non-disclosure agreements required by chip foundries prevent vendors from open-sourcing low-level Secure Element code, creating an architectural trade-off between physical security and code transparen
- Standard microcontrollers allow 100% auditable open-source firmware but are susceptible to low-cost physical attacks like voltage glitching if the device is stolen.

Hardware wallets safeguard crypto assets by keeping private keys completely isolated from internet-connected devices. The foundational architecture of these devices centers on a single design tension: using closed-source, hardened silicon known as a **Secure Element** to stop physical extraction, versus using transparent, general-purpose microcontrollers that permit independent verification of every line of code.

Understanding this distinction requires examining how microchips store secrets, how attackers extract them when they gain physical possession of a device, and how security certifications quantify hardware resilience.

## The Anatomy of Hardware Key Isolation

A cryptocurrency wallet does not store coins; it stores the cryptographic private keys required to authorize transactions on decentralized ledgers like Bitcoin ($BTC) or Ethereum ($ETH). When a user approves a transfer, the host computer or smartphone transmits an unsigned transaction payload to the hardware device. The device parses the transaction, displays the details on its internal screen, signs the payload internally using the private key, and returns only the digital signature to the host. The private key never leaves the device's internal memory boundary.

To maintain this boundary, hardware manufacturers must decide what type of silicon houses the key. A standard general-purpose microcontroller unit (MCU)—such as those found in smart appliances, toys, and basic electronics—executes instructions linearly and stores data in standard flash memory. These chips were engineered for low cost and processing efficiency, not for adversarial environments where an attacker has physical possession of the hardware.

A **Secure Element** (SE), by contrast, is a dedicated integrated circuit engineered from the physical transistor layer up to defend against unauthorized extraction. These chips originate from the banking, smart card, and passport industries, where secrets must remain intact even if an adversary disassembles the casing.

To understand the broader signing pipeline, readers can explore [how hardware wallets work](https://basisdesk.news/learn/hardware-wallets-how-they-work-5ukbz).

## Physical Extraction Attacks and Silicon Defense

When an attacker gains physical custody of a wallet, standard software protections such as numeric PINs can be bypassed if the chip lacks hardware hardening. Attackers deploy several categories of physical attacks against unhardened microcontrollers:

Fault injection involves disrupting the chip's electrical environment to induce computational errors. By briefly dropping the core operating voltage (voltage glitching) or hitting the chip with a precise laser pulse (clock or optical glitching) at the exact nanosecond the processor evaluates a PIN check, an attacker can cause the CPU to skip the conditional branch instruction that verifies the credential. Once skipped, the chip grants unrestricted access to onboard flash memory.

Side-channel analysis measures the physical side effects of cryptographic computation. **Differential Power Analysis** (DPA) records micro-variations in electrical current consumption while the chip computes an elliptic curve digital signature. Because binary ones and zeros consume slightly different amounts of power during mathematical operations, statistical analysis across thousands of traces can reconstruct the private key without opening the chip package.

Decapping and microprobing represent invasive physical attacks. The attacker dissolves the chip's epoxy resin packaging with fuming nitric acid to expose the bare silicon die. Using high-powered optical microscopes and micromanipulators with microscopic needles, the attacker taps directly into internal communication buses to eavesdrop on data traveling between memory and the processor core.

Secure Elements mitigate these vectors through specialized physical countermeasures built directly into the silicon substrate:

Active protective shields cover the entire upper layer of the silicon die with an electrical mesh. If an attacker attempts to etch or probe the surface, the circuit breaks or shorts, triggering an immediate and irreversible zeroization of all stored cryptographic secrets.

Internal bus encryption ensures that data moving between the arithmetic logic unit and internal memory cells is cryptographically scrambled, rendering direct probing of traces useless without the hardware-level key baked into the silicon fabric.

Hardware sensors continuously monitor environmental parameters, including clock frequency, supply voltage, ambient temperature, and light exposure. If voltage fluctuates outside a narrow tolerance, or if decapping exposes internal light sensors, the chip halts operation instantly.

Internal clock jitter and randomized dummy operations inject artificial electrical noise, flattening the power consumption profile and defeating side-channel analysis.

## Decoding Common Criteria and EAL Ratings

Manufacturers quantify a chip's resistance to physical intrusion through the Common Criteria for Information Technology Security Evaluation (ISO/IEC 15408), an internationally recognized framework. The framework assigns an **Evaluation Assurance Level** (EAL), scaled from EAL1 to EAL7.

An EAL rating does not measure how many security features a device possesses; it measures the rigor and depth with which an independent, accredited laboratory tested the chip against a standardized security target.

EAL4 is common for commercial operating systems and standard enterprise hardware. It denotes that the system has been methodically designed, tested, and reviewed, but it does not guarantee protection against advanced physical laboratory attacks.

EAL5 indicates that the chip was semi-formally designed and tested. When appended with a plus sign (EAL5+), it signifies that the testing facility conducted vulnerability assessments beyond the standard baseline, specifically verifying resistance against high-potential penetration attacks.

EAL6 and EAL6+ represent the upper echelon of commercially viable security silicon, typically utilized in biometric passports, defense communication modules, and high-volume banking chips. Achieving EAL6+ requires formal mathematical verification of the security architecture and extensive physical penetration testing where evaluation teams spend months attempting side-channel extraction, decapping, and fault injection.

## The Open-Source Dilemma and NDAs

While Secure Elements provide superior resistance to physical laboratory probes, they introduce a distinct governance challenge: transparency.

The dominant semiconductor foundries that manufacture certified Secure Elements require hardware wallet vendors to sign strict Non-Disclosure Agreements (NDAs). These NDAs forbid vendors from publishing the detailed data sheets, register maps, and low-level cryptographic driver code that interface with the secure chip. Consequently, a wallet vendor using an EAL6+ Secure Element cannot release its complete firmware as fully auditable, open-source code.

This structural restriction divides the hardware wallet industry into distinct architectural philosophies:

Closed Secure Element with open application layer: Vendors like Ledger pair a certified Secure Element with a custom operating system. While the top-level application code handling transaction logic is open source, the underlying platform firmware that interacts directly with the silicon remains closed. Users must trust the manufacturer's cryptographic attestation that the closed binary does not contain backdoors or key-export mechanisms.

Dual-chip architectures: Some devices employ two separate microchips: a general-purpose, fully open-source MCU that runs the user interface and coordinates device state, coupled to a closed-source Secure Element that acts strictly as a cryptographic vault. In this model, the general MCU requires the Secure Element to release or sign data, mitigating single-point physical compromise while keeping the primary application stack auditable.

Pure open-source microcontrollers: Vendors like Trezor historically opted to avoid Secure Elements entirely, building wallets around standard microcontrollers (such as STMicroelectronics STM32 chips). Because these chips carry no NDA restrictions, 100% of the bootloader, operating system, and application firmware is published. Anyone can inspect the code, compile it independently, and verify reproducible builds. The trade-off is physical security: standard MCUs lack active shields, voltage monitors, and side-channel countermeasures, rendering them vulnerable to physical extraction techniques like fault injection if stolen.

Recent developments include custom-tailored silicon initiatives (such as open-access secure elements developed specifically to provide physical resistance without proprietary NDA encumbrance), but these designs remain less established than traditional banking-grade chips.

## A Worked Attack-Cost Scenario

To evaluate whether a user requires Secure Element protection, one must weigh the replacement cost of an asset against the capital cost required to execute a physical hardware breach.

Assume an adversary steals a physical hardware wallet. The target device is protected by an eight-digit numeric PIN, which normally allows an adversary a maximum of three attempts before wiping the memory.

Scenario A: Standard General-Purpose MCU (No Secure Element)
- Equipment required: A commercial digital storage oscilloscope, an arbitrary waveform generator or custom FPGA-based glitch board, and a stable benchtop power supply. Total hardware cost is approximately $1,500 to $3,000.
- Skill and labor: Intermediate embedded-hardware expertise. An experienced reverse engineer can decap the casing, identify the VDD power lines supplying the CPU core, and establish a voltage glitching script within 40 to 80 hours of bench testing.
- Outcome: Once the glitch successfully bypasses the PIN verification loop instruction, the flash memory contents—including the encrypted seed phrase and the operational key—are read directly via standard debug interfaces (such as SWD/JTAG). The recovery seed is obtained.

Scenario B: High-Assurance Secure Element (EAL6+)
- Equipment required: A focused ion beam (FIB) workstation for trace cutting and microprobing, a transmission electron microscope for silicon mapping, and multi-channel differential power analysis rigs equipped with synchronized laser emitters. Equipment acquisition or specialized laboratory rental costs exceed $250,000 to $500,000.
- Skill and labor: Advanced semiconductor security teams requiring deep domain knowledge of custom hardware obfuscation, active shield topologies, and proprietary crypto coprocessors. Analysis often demands several months with a high probability of destroying the single target die during chemical etching or circuit cutting.
- Outcome: Attempting to bypass the active shield severed the top-layer mesh, permanently triggering internal tamper switches that erased the secure storage registers. The private key remains unrecoverable.

For high-net-worth individuals or institutional custodians, the financial cost of an attack on an unhardened MCU is orders of magnitude lower than the value of the stored assets, making specialized silicon essential.

## Supply Chain Auditing and Attestation

Physical resistance at rest does not solve the risk of upstream tampering before the device reaches the user. Both open-source and closed-source devices must verify device integrity.

Certified Secure Elements use cryptographic **attestation** keys injected during the manufacturing process inside a secure cleanroom. When the user sets up the device, the host management software challenges the Secure Element with a cryptographic nonce. The chip signs this nonce using its factory-injected private key, which chains back to the manufacturer's root Certificate Authority.

This verification proves that the chip inside the casing is genuine, that the manufacturer flashed the device, and that no intermediary swapped the board for a modified clone during transit. Verification methods and counterfeit indicators are detailed in [how to verify a new hardware wallet](https://basisdesk.news/learn/fake-hardware-wallets-supply-chain-tampering).

In pure open-source hardware, attestation is more difficult because standard microcontrollers lack factory-sealed cryptographic identities. Instead, open-source designs rely heavily on reproducible builds, where users compile the open source code on their own machines to verify that the resulting binary hash matches the binary installed on the device byte for byte.

## Common Misconceptions

"Open-source code automatically makes a device more secure than closed silicon."
Transparency allows independent reviewers to audit software logic for memory leaks, integer overflows, and malicious backdoors. However, open-source code cannot alter the underlying laws of semiconductor physics. If code runs on a standard chip devoid of internal hardware shields and power balancing, the most elegant software architecture can still be defeated in seconds by physical glitching or differential power analysis.

"A Secure Element prevents remote malware on a computer from draining funds."
A Secure Element protects private keys from unauthorized extraction; it does not dictate what a user chooses to sign. If an attacker's computer malware swaps the destination address in a transaction payload, and the user approves the transaction on the hardware wallet without verifying the address on the device's physical screen (known as blind signing), the funds are lost. The Secure Element signed the transaction precisely as authorized.

"EAL ratings certify the entire hardware wallet as unhackable."
Common Criteria certifications evaluate a specific "Target of Evaluation" (TOE), which is typically the bare microchip and its foundational crypto library—not the assembled wallet. An EAL6+ rating on a Secure Element does not guarantee that the wallet's secondary microcontroller, display screen, USB controller, or vendor desktop software is free of vulnerabilities.

## How This Connects to the Market

The debate between Secure Elements and open-source firmware directly influences institutional custody, retail self-custody adoption, and regulatory compliance.

Institutional capital allocators operating under regulatory frameworks (such as the EU's Markets in Crypto-Assets regulation, or MiCA) face specific compliance requirements regarding operational risk and asset storage. Custodians and exchange operators overwhelmingly favor hardware security modules (HSMs) and devices anchored by EAL6+ silicon because these certifications align with traditional banking compliance standards (like FIPS 140-2/140-3 and ISO/IEC 19790).

Conversely, retail decentralization advocates prioritize sovereign verification, arguing that trusting a chip vendor's proprietary silicon contradicts the trust-minimized ethos of public blockchains. As physical extraction tools become cheaper and more automated, the market is steadily moving toward hybrid architectures: hardware platforms that implement physical silicon hardening while isolating and open-sourcing the higher-level transaction logic and signing repositories.

## FAQ

**What is the difference between an MCU and a Secure Element?**

A microcontroller unit (MCU) is a general-purpose processor designed to execute standard code efficiently, but it lacks physical defenses. A Secure Element (SE) is a hardened chip designed specifically to protect cryptographic secrets against physical probing, side-channel analysis, and fault injection.

**What does the '+' mean in EAL5+ or EAL6+?**

The plus sign indicates that the testing laboratory evaluated the chip against security requirements exceeding the standard baseline for that Evaluation Assurance Level, specifically validating resilience against advanced, high-potential penetration attacks.

**Can someone steal my crypto from a hardware wallet without the PIN?**

If the wallet uses an unhardened general-purpose microcontroller and the attacker has physical custody and specialized lab equipment, they may bypass the PIN using fault injection. If the wallet stores keys inside a certified Secure Element, extracting the key without the PIN is extremely difficult and cost-prohibitive.

**Why don't all hardware wallet vendors use Secure Elements?**

Major Secure Element manufacturers require strict non-disclosure agreements, which prevents vendors from publishing the chip's complete underlying driver code. Vendors prioritizing complete open-source transparency historically chose general microcontrollers to keep all code publicly auditable.

## Sources

1. [Bitcoin Developer Documentation: Wallets](https://developer.bitcoin.org/devguide/wallets.html) — Bitcoin Project
2. [Common Criteria for Information Technology Security Evaluation](https://www.commoncriteriaportal.org/) — Common Criteria
3. [FIPS 140-3: Security Requirements for Cryptographic Modules](https://csrc.nist.gov/pubs/fips/140-3/final) — National Institute of Standards and Technology

---
Basis Desk Newsroom · AI-generated, source-verified · https://basisdesk.news/about/how-we-use-ai
