Novel Bits

BlueJoule™: How Efficient Is Your Bluetooth LE Chip? Really?

An open, standardized benchmark that reveals the true energy efficiency of Bluetooth LE implementations. Compare chips fairly with real-world measurements, not datasheet specs.

Open Source Benchmark

BlueJoule™: How Efficient Is Your Bluetooth LE Chip? Really?

Vendor datasheets tell one story.
Real-world measurements tell another.

View Benchmark Results Watch the Webinar

A collaboration between The EM Foundation and Novel Bits

The Problem with Comparing Chips

?
Datasheets show best-case specs — Sleep currents and TX power listed in isolation don't tell you how much energy your application will actually consume.
?
No standardized comparison method — Every vendor measures differently. How do you compare a Nordic chip to a TI chip to a Silicon Labs chip fairly?
?
Real-world results vary wildly — SDK optimizations, firmware configuration, and radio scheduling all affect actual energy consumption in ways datasheets can't capture.

BlueJoule™ solves this.

What Is BlueJoule™?

A standardized, open-source benchmark for measuring real-world Bluetooth LE energy consumption

📏

Standardized Workload

Every chip runs the exact same test: Bluetooth LE advertising at a fixed interval and TX power, with a standardized payload containing the "BlueJoule" local name.

⚡

Real Measurements

No calculated estimates. We measure actual energy consumption using precision instruments like the Joulescope JS220 and Nordic PPK2.

⚖️

Fair Comparison

Same test, same conditions, same methodology. Finally compare chips from different vendors on a level playing field.

🔓

Open Source

All sample applications, measurement scripts, and methodology are open source. Verify our results, submit your own, or adapt for your needs.

Meet the EM•erald

A gamified unit of energy efficiency

Instead of comparing raw microjoules or milliamp-hours, BlueJoule™ scores chips in EM•eralds—a unit designed to be intuitive and meaningful.

EM•eralds = 80 ÷ J/day
🔋
1 EM•erald ≈ 1 CR2032-month

A chip scoring 12 EM•eralds could theoretically run for 12 months on a single CR2032 coin cell (under the benchmark workload).

📊
Higher is better

More EM•eralds = more efficient. Simple to compare, easy to understand.

💎
EM•eralds
Benchmark #1

Bluetooth LE Advertising

The first BlueJoule™ benchmark focuses on advertising—the most common Bluetooth LE use case for beacons, sensors, and low-power devices.

Test Parameters

Advertising Interval 1 second (1Hz)
TX Power 0 dBm
Supply Voltage Optimal (varies by device)
Payload Size 19 bytes
Advertising Type Non-connectable, non-scannable
Channels 37, 38, 39 (all primary)

Advertising Payload Structure

02 01 06 | 0a 08 42 6c 75 65 4a 6f 75 6c 65 | 04 ff d3 08 ff
Flags: LE General Discoverable + BR/EDR Not Supported Local Name: "BlueJoule" Manufacturer Data: Novel Bits (0x08D3)
View Current Results →

More Benchmarks Coming

The advertising benchmark is just the beginning. We're developing additional benchmarks to cover the full spectrum of Bluetooth LE use cases.

Live Now

Advertising

Bluetooth LE advertising—the foundation for beacons, asset trackers, and sensor nodes.

Planned

Connections

Connected data transfer—measuring energy for bidirectional communication between central and peripheral.

Planned

Scanning

Active and passive scanning—important for central devices and gateways that discover nearby peripherals.

Planned

High-Speed PHYs

2M PHY and Coded PHY benchmarks—comparing long range and high throughput modes.

Future

Bluetooth LE Audio

LC3 codec streaming—the next generation of Bluetooth audio with improved efficiency.

Future

Direction Finding

AoA/AoD benchmarks—measuring the energy cost of precise indoor positioning.

Have a benchmark idea? Get involved and help shape the roadmap.

Measured with EM•Scope

Our open-source power measurement and analysis tool

1
📡
emscope grab

Capture power data from your measurement hardware

2
🔍
emscope scan

Analyze and extract energy metrics

3
📊
emscope view

Visualize power profiles and results

4
📦
emscope pack

Package results for submission

Supported Measurement Hardware

Joulescope JS220 High-precision energy analyzer
Nordic PPK2 Power Profiler Kit II

Watch the BlueJoule™ Webinar

Learn about the methodology, see live measurements, and understand how to interpret the results

What You'll Learn

  • Why vendor datasheet comparisons fall short
  • The BlueJoule benchmark methodology explained
  • How EM•eralds are calculated and what they mean
  • Live demonstration of EM•Scope measurements
  • Walkthrough of results for popular chips
  • How to submit your own benchmark results

Get Access to the Recording

We'll also keep you updated on new benchmarks and results.

⚠️ Note about benchmark results: The results shown in the webinar recording reflect measurements at the time of recording. BlueJoule is a living benchmark—new chips are added, firmware is optimized, and scores may change. Always check the GitHub repository for the latest results.

Get Involved

BlueJoule™ is an open community effort. Here's how you can contribute.

📤

Submit Your Results

Have a chip we haven't benchmarked? Run the benchmark application, capture your measurements, and submit via pull request. We'll verify and add your results to the leaderboard.

View submission guidelines →
🏭

Chip Vendors

Want your chip on the leaderboard? Provide a sample application that implements the benchmark specification and submit your own measurements—or we can measure it for you. Either way, your customers get real-world efficiency data.

View benchmark specification →
💡

Suggest Benchmarks

Have an idea for a new benchmark? Start a discussion to propose methodology, suggest test parameters, or share improvements to existing benchmarks.

Start a discussion →
🐛

Report Issues

Found a bug in the firmware or measurement tools? Spot an error in the results? Let us know so we can fix it and maintain accuracy.

Report an issue →
📦

BlueJoule Repository

Firmware, measurement scripts, methodology documentation, and results—all open source.

⭐ View on GitHub

Frequently Asked Questions

The Benchmark

What exactly does BlueJoule™ measure?
BlueJoule™ measures the total energy consumption of a Bluetooth LE chip running a standardized workload. For the advertising benchmark, this includes the energy for sleep periods plus the energy for each advertising event (wakeup, radio TX on all three primary channels, and return to sleep).
What do EM•eralds actually mean?
EM•eralds are a unit of energy efficiency. The formula is 80 ÷ J/day, which means 1 EM•erald roughly equals 1 month of runtime on a CR2032 coin cell battery (under the benchmark workload). A chip scoring 24 EM•eralds could theoretically run for 2 years on a single CR2032.
Why might results differ from datasheet specs?
Datasheets typically list individual parameters (sleep current, TX current at specific power levels) under ideal conditions. BlueJoule measures the complete system including firmware overhead, radio scheduling, oscillator startup, and real-world transitions. These factors can significantly impact total energy consumption.
Why does supply voltage vary between devices?
BlueJoule cares about energy (voltage × current), not just current draw. Leaderboard entries use whatever voltage gives the most optimal energy consumption. While we test at 3.0V for datasheet comparisons, lower voltage devices—like emerging 1.5V chips—can achieve better energy efficiency even if their current consumption is slightly higher.
How are measurements taken and analyzed?
We capture approximately 10 raw advertising events at 1Hz and analyze them to extract two key metrics: average sleep current and average event energy. With these values, we can compute energy consumption for any event period—from 1 second to 10 seconds and beyond—then calculate J/day and convert to EM•eralds.
Can vendors optimize their firmware for better scores?
Yes, and that's encouraged! The benchmark uses vendor SDKs with reasonable optimizations enabled. If a vendor releases an improved SDK or someone finds a better configuration, we welcome updated submissions. This drives the industry toward better efficiency.
What about DC-DC vs LDO configurations?
Where chips support both DC-DC converter and LDO regulator modes, we typically test the most efficient configuration (usually DC-DC). The configuration is documented for each result so you can see exactly what was tested.

Getting Involved

How do I submit benchmark results for my chip?
Fork the BlueJoule repository, run the benchmark application on your platform, capture measurements using EM•Scope, and submit a pull request with your results. Chip vendors are especially encouraged to submit results for new or missing chips.
What measurement hardware do I need?
You'll need either a Joulescope JS220 (~$999) or Nordic PPK2 (~$99). The PPK2 is more affordable and sufficient for most measurements. Both are supported by EM•Scope.
My chip isn't supported yet. What can I do?
If you're a chip vendor, provide a sample application that implements the benchmark specification—we'll measure it and add your results. If you're an engineer wanting to test a chip, check the repository for existing implementations or start a discussion to request support for your platform.
Can I use BlueJoule results in my marketing materials?
Yes, with attribution. Please link to the GitHub repository and include the benchmark date, as results may change over time. We recommend phrases like "BlueJoule benchmark, [date]" with a link to the BlueJoule repo.
Who is behind BlueJoule?
BlueJoule is a collaboration between The EM Foundation, a non-profit organization focused on transparent energy efficiency benchmarking, and Novel Bits, a company specializing in Bluetooth LE training and consulting. The project operates independently from chip vendors.
Is BlueJoule affiliated with any chip vendor?
No. Both The EM Foundation and Novel Bits operate independently from chip manufacturers. We don't receive funding from vendors for favorable results. Our goal is accurate, transparent benchmarking that helps engineers make informed decisions.
Where can I get help or ask more questions?
Start a discussion on the GitHub repository for technical questions or general inquiries.

A Collaboration

BlueJoule™ is a joint initiative bringing together benchmarking expertise and Bluetooth LE knowledge.

The EM Foundation

A non-profit organization that promotes, sustains, and evolves programming technology for use within the broader embedded systems community. This charter includes the openly-available EM•Scope tooling used for quantifying energy efficiency within ultra-low-power applications.

Learn more at openem.org →

Novel Bits

A Bluetooth LE consultancy providing development services, technical training, and protocol expertise to product companies and engineering teams worldwide. Novel Bits helps organizations bring wireless products to market while also fostering a thriving community of embedded developers.

Learn more at novelbits.io →

Ready to See How Your Chip Stacks Up?

Explore the benchmark results, watch the methodology webinar, or contribute your own measurements to the community.

View Benchmark Results GitHub Repository