TI CC2340 vs nRF54L15 vs InPlay IN120: Which Sub-$1 BLE Chip Actually Wins on Power and Tooling

Comparing power use and developer tooling across three sub-$1 Bluetooth LE chips

You’ve got three datasheets open, all three quote sleep currents in the low microamps, and none of them measured it the same way. One assumes 1.8V, another 3.0V. One quotes TX at 0 dBm, another at +6. Your “comparison” is really three vendors each picking the test bench that flatters their part.

That’s the trap with a sub-$1 BLE SoC pick. The headline numbers look comparable. They aren’t.

There’s no single winner. Each chip leads on a different axis, and the right call depends on what’s actually binding your design. This guide positions the TI CC2340, Nordic nRF54L15, and InPlay IN120 by use case, not by a leaderboard.

One hard caveat up front. Every figure below is synthesized from published datasheets and community measurements as of January 2026. These aren’t original benchmarks. Verify against the latest datasheet revision and, more importantly, on your own hardware with your own duty cycle.

The Contenders at a Glance

Confirm each cell against current datasheets, because revisions move.

+-------------+------------------+------------------+------------------+
| Spec        | TI CC2340        | nRF54L15         | InPlay IN120     |
+-------------+------------------+------------------+------------------+
| Core        | Arm Cortex-M0+   | Arm Cortex-M33   | Arm Cortex-M0    |
| Flash/RAM   | up to 512KB/36KB | up to 1.5MB/256K | ~256KB/(verify)  |
| BLE Ver     | 5.3              | 5.4+             | 5.x              |
| DC-DC       | yes              | yes              | yes              |
| Standout    | TI analog/cost   | tooling+headroom | price/footprint  |
+-------------+------------------+------------------+------------------+

The CC2340 is TI’s cost-down play built on a Cortex-M0+, leaning on TI’s analog and integration story. The nRF54L15 is the heavyweight: a Cortex-M33 with a coprocessor, the biggest memory ceiling here, and Nordic’s tooling behind it. The IN120 is the price/footprint fighter, a leaner part aimed squarely at high-volume, cost-sensitive nodes.

The cores, memory, and ecosystem weight all differ. That’s why a single number can’t rank them. A part that wins your BOM cost can lose your engineering schedule.

Power: Read the Footnotes, Not the Headline

This is where a low-cost BLE chip comparison goes wrong. Vendors quote current draw under conditions tuned to look good, and the differences hide in the footnotes.

Watch four variables. Supply voltage (a 1.8V rail draws differently than 3.0V). TX power (0 dBm versus +6 dBm is a big delta on the radio). DC-DC versus LDO (the buck converter cuts active current but adds an inductor to your BOM). And the sleep mode definition (RAM-retained sleep isn’t the same as a deep system-off state, and vendors love to quote the deepest one).

Here are best-available published figures. Treat them as a starting point for your own measurement, not gospel.

Current Draw — PUBLISHED FIGURES (Jan 2026, verify conditions)
Mode          | CC2340     | nRF54L15   | IN120
--------------|------------|------------|------------
TX (0 dBm)    | ~5.0 mA    | ~3.4 mA    | ~4.5 mA
RX            | ~4.0 mA    | ~3.0 mA    | ~4.0 mA
Sleep (RAM)   | ~0.83 uA   | ~1.0 uA    | ~1.5 uA
Deep/OFF      | ~0.15 uA   | ~0.35 uA   | (verify)
* Test conditions differ per vendor, NOT apples-to-apples
* Sources: TI CC2340 datasheet, Nordic nRF54L15 datasheet,
  InPlay IN120 datasheet, all rev as of Jan 2026

For CC2340 power optimization, the levers are the DC-DC converter for active modes and TI’s standby modes with RAM retention. Get the radio duty cycle low and standby current dominates your battery math, so that’s where to spend tuning effort.

On nRF54L15 battery life, Nordic’s pitch is the sub-microamp System OFF state plus an efficient active radio. If your node spends 99% of its life asleep and wakes briefly to advertise, that deep-sleep floor sets your coin-cell lifetime.

The IN120 holds its own on active draw and competes hard on price, though its deepest sleep figures deserve careful confirmation against the current datasheet.

None of this replaces measurement. Wire up a dev kit, run your actual advertising interval and payload, and put a current meter on it. Your average current is a weighted blend of these modes, and the weights are yours, not the vendor’s. A duty-cycle and power budgeting approach will tell you more than any datasheet table.

BLE SoC Tooling: Where Engineering Time Actually Goes

The spec sheet won’t tell you this, but tooling is often the deciding cost. A chip that saves 30 cents in the BOM can burn weeks of engineer time in bring-up, and at volume the math still favors the part that ships sooner.

nRF54L15 has the deepest ecosystem of the three. The nRF Connect SDK is built on Zephyr, which means a mature RTOS, broad peripheral driver coverage, real debug and power-profiling tools, and a large community to search when you hit a wall. If you’ve shipped on a Nordic part before, much of that knowledge carries over. Zephyr maturity alone shortens a lot of schedules.

CC2340 runs on TI’s SimpleLink SDK with Code Composer Studio (and the newer Theia-based IDE). If your team already lives in the TI world, the familiarity is worth real money, and TI’s documentation and example coverage are solid. The ecosystem is narrower than Nordic’s BLE community but deep where TI is strong.

InPlay IN120 is the risk here, and you should price that risk in. The SDK is leaner and the community is smaller, which means fewer worked examples, fewer Stack Overflow answers, and more time spent reading source. None of that is disqualifying. It just means you budget extra bring-up time and lean harder on InPlay’s support.

A cheaper chip with thin tooling can quietly erase its BOM savings before you ship.

Pick by Use Case

Here’s the decision flow, then the verdicts.

Range a problem? --YES--> Solve at network layer first
       |
       NO
       v
Need tooling depth? --YES--> nRF54L15
       |
       NO
       v
TI analog / cost-down? --YES--> CC2340
       |
       NO
       v
Lowest unit cost/footprint --> IN120

Choose nRF54L15 if tooling maturity and RF feature headroom drive your decision. The Cortex-M33, the larger memory ceiling, and the Zephyr-based SDK give you room to grow features without a chip swap later. Good fit for products that’ll get more complex over their life, or teams that value a deep community.

Choose CC2340 if TI analog integration, aggressive cost-down, or existing SimpleLink experience dominate. If your design already pulls in TI parts, or your team knows Code Composer cold, the CC2340 lets you reuse what you’ve got and squeeze the BOM.

Choose IN120 if unit cost and footprint at volume are the hard constraint, and you can absorb leaner tooling. For a simple, high-volume node where every cent and square millimeter counts, and the firmware feature set is modest, the IN120’s price and size make a strong case. Just plan for the extra bring-up.

When the Chip Doesn’t Matter, Range Does

If your node can’t reach a gateway, it doesn’t matter which sub-$1 SoC you soldered down. A perfectly power-optimized BLE node that’s out of range of any receiver is a dead node.

BLE’s range is short by design. The usual fix is deploying gateways everywhere your assets might roam, which is expensive infrastructure you have to buy, install, and maintain. For anything that moves (asset trackers, fleet sensors, supply-chain tags), gateway coverage is the real constraint, not the silicon.

A satellite-backed BLE network changes that calculation. Hubble’s network has 7 satellites in orbit today (as of January 2026), and a standard BLE advertising packet from your node can reach them directly. No gateway hardware to deploy, no cellular module, no LoRa infrastructure. Your node keeps advertising the way it already does.

Solve range at the network layer, and your chip choice goes back to being a pure power-and-tooling decision. The CC2340, nRF54L15, and IN120 all speak standard BLE, so any of them can sit in a node that reaches orbit. You pick the silicon that fits your power budget and your team’s tools, and let the network handle reach.

Verify Before You Buy

No single chip wins. The nRF54L15 leads on tooling and headroom, the CC2340 on TI integration and cost-down, the IN120 on price and footprint. The right one depends on your binding constraint.

Before you finalize the BOM, run this checklist:

  • Confirm the datasheet revision date for every figure you’re relying on. Numbers shift between revs.
  • Measure current on a dev kit using your actual duty cycle and payload, not the vendor’s test conditions.
  • Validate that the SDK supports every peripheral and BLE feature your design needs, including the ones you’ll add in v2.
  • Confirm your range and network plan before locking the design, because that decision outranks the chip.

If your nodes need to reach beyond local gateways, sort out the network layer first. You can sign up at hubble.com to see how a satellite-backed BLE network fits your architecture, then go optimize the node chip with a clear head.

Hubble Network lets standard BLE nodes reach orbit directly, so your range plan isn’t bound by local gateway placement. See how it works →