DJI’s video transmission system is one of the core competitive advantages of its drone products. But behind the OcuSync protocol and image transmission performance lies an evolutionary story of chips — from using off-the-shelf general-purpose SoCs to custom-designed ASICs.
This article traces the evolution of DJI video transmission chips across three phases: the early OcuSync era using Leadcore LC1860, the transition period with custom P1 and S1 chips, and the current era of specification-defined chips including S2.
Phase 1: Early OcuSync Era — General-Purpose SoC
In the early days of DJI’s digital video transmission, the system used the Leadcore LC1860 SoC. This was a general-purpose application processor with integrated baseband, originally designed for smartphones and tablets.
The LC1860 was a quad-core Cortex-A7 processor with an integrated LTE baseband. DJI chose this chip because it had sufficient processing power for video encoding/decoding and the baseband could be repurposed for custom RF protocols. The OcuSync protocol stack ran on top of the LC1860’s application processor, while the baseband handled the physical layer.
This approach worked but had limitations: the LC1860’s baseband was designed for cellular networks, not optimized for low-latency drone video links. Much of the chip’s capability (cellular protocol stack, SIM interface, etc.) was wasted. Power consumption was higher than necessary because the chip was designed for consumer electronics, not embedded drone applications.
Phase 2: Transition Period — P1 and S1 Custom ASICs
DJI moved to custom chips through collaboration with Leadcore. Rather than designing chips from scratch, DJI worked with Leadcore to create customized variants of existing architectures — stripping unnecessary modules and optimizing for video transmission use cases.
P1 Chip (Codename: Falcon)
The P1 was the first custom chip, built on Leadcore’s architecture but with significant modifications:
- Retained the ARM Cortex-A application processor (needed for video decoding)
- Kept the baseband and RF transceiver (core to video transmission)
- Removed cellular modem components (SIM interface, cellular protocol stack)
- Optimized power management for drone/remote controller use cases
The P1 first appeared in the DJI FPV system and subsequently in all consumer-level drones, continuing through to the Mavic 3 and Mini 3 generations. By removing the cellular modem, DJI reduced power consumption and die area while focusing silicon resources on the video transmission pipeline.
S1 Chip (Codename: Sparrow)
The S1 appeared alongside P1 as a streamlined variant. Where P1 retained the Cortex-A application processor, S1 downgraded to a Cortex-M microcontroller — it only handled RF transceiver functions, not video decoding.
S1 was designed for remote controllers that offload video decoding to a separate SoC (such as the Eagle H3 chipset). Models like the RC231 used this architecture: S1 handles RF communication, while a companion chip handles video decoding and display.
The design logic is clear: if a remote controller doesn’t need to decode video on the same chip as the RF transceiver, don’t waste Cortex-A compute and power on it. A Cortex-M doing pure RF control is sufficient and more efficient. Aomway’s engineering team notes this is a common pattern in wireless system design — separating the RF transceiver from the application processor allows each to be optimized independently.
Phase 3: Current Era (2022–Present) — S2 and Specification Definition Shift
The current phase continues using P1, S1, and the new S2 chip, but the collaboration model has fundamentally changed.
In the transition period, P1 and S1 were more like Leadcore doing customized trimming per DJI’s requirements — the chip architecture was still primarily Leadcore’s. In the current phase, the model has shifted to “DJI defines specifications, chip manufacturers tape out per specs.”
DJI’s role evolved from “selector” to “Specification Definer”:
- Defines which modules the chip needs (keep RF + baseband for video links, remove cellular baseband)
- Specifies module combinations (P1: ARM + baseband; S1: baseband + RF only)
- Provides firmware and protocol stacks (OcuSync 3/4, supporting wider channels and stronger anti-interference)
- Controls production — chips are for internal use only, not sold externally
The difference between phases: the transition period was “Leadcore designs the chip, DJI provides requirements for trimming”; the current phase is “DJI defines specs, Leadcore or foundry tapes out per specs” — the latter is essentially a chip foundry role.
S2 Chip
The S2 is an upgraded S1, supporting wider channel bandwidths — 60MHz and even 80MHz+. This corresponds to the OcuSync 4 era. Expanding channel bandwidth from 40MHz to 60MHz/80MHz means more data can be transmitted per unit time, but requires synchronized upgrades in RF front-end and baseband processing capability.
The S2’s introduction shows DJI has a clear generational roadmap for video transmission bandwidth, with chip specification upgrades paced to match protocol upgrades. Aomway’s analysis of DJI’s roadmap suggests this bandwidth expansion enables higher-resolution video streaming (4K/60fps real-time) and more robust anti-interference through wider spread spectrum.
Evolution Summary
| Phase | Chip Source | DJI’s Role | Representative Chips |
|---|---|---|---|
| OcuSync Early | General SoC (Leadcore LC1860) | Selection + Protocol Development | LC1860 |
| Transition | Custom ASIC (based on Leadcore architecture) | Trim Requirements + Protocol Development | P1, S1 |
| Current | Spec-defined custom chips | Spec Definition + Protocol Development + Production Control | P1, S1, S2 |
Core Strategy: Protocol and Algorithm Over Chip Design
DJI never developed communication chips from scratch — semiconductor design and manufacturing is not their core competency. Instead, through a three-step approach — first use general-purpose chips to validate the protocol, then do customized trimming, finally master specification definition — DJI gained increasing control over the chip layer.
What truly determines video transmission performance is the protocol and algorithms; the chip is the hardware carrier. DJI’s strategy is to make chips serve their needs, not be limited by them. This is a lesson applicable to any company building wireless communication systems: control the protocol layer, and you control the product’s competitive advantage — even if you never design your own silicon.
Have questions about video transmission chip selection, custom ASIC development, or wireless protocol design for drone applications? Contact Aomway at [email protected] — our engineering team provides wireless system consulting, chip selection guidance, and complete video transmission system design for industrial UAV applications.
Frequently Asked Questions
1. Why did DJI use a smartphone SoC (LC1860) for video transmission initially?
In the early OcuSync era, no purpose-built chip existed for drone video transmission. The Leadcore LC1860 was selected because it combined an application processor (quad-core Cortex-A7) with an integrated baseband — the application processor could run the OcuSync protocol stack while the baseband handled physical layer processing. This was a pragmatic choice: use what’s available to validate the protocol, then move to custom silicon once volume justifies it. The downside was wasted silicon (cellular modem, SIM interface) and higher power consumption, but for an initial product launch, time-to-market mattered more than chip optimization.
2. What is the architectural difference between P1 and S1 chips?
P1 (Falcon) retains the full ARM Cortex-A application processor alongside the baseband and RF transceiver — it can handle both RF communication and video decoding on a single chip. S1 (Sparrow) downgrades to a Cortex-M microcontroller, handling only RF transceiver functions. S1 is used in remote controllers where video decoding is offloaded to a companion SoC (like the Eagle H3). This separation allows DJI to optimize cost: controllers that need integrated video decoding use the more expensive P1; controllers with a separate video decoder chip use the cheaper, more efficient S1.
3. How does S2 improve over S1 in terms of video transmission performance?
S2 supports wider channel bandwidths (60MHz/80MHz vs. S1’s 40MHz). Wider channels mean higher data throughput — enabling 4K/60fps real-time video streaming where S1 might be limited to 1080p or lower bitrate 4K. However, wider channels also require more RF spectrum, better anti-interference algorithms, and upgraded RF front-end components. S2’s introduction aligns with OcuSync 4, which includes improved spread spectrum and frequency hopping algorithms to maintain reliability even with wider channels in congested spectrum environments.
4. Did DJI design these chips themselves or just customize Leadcore’s designs?
DJI did not design the chips from scratch. The evolution went through three phases: (1) Selection — picked an off-the-shelf Leadcore SoC and wrote the protocol stack on top of it. (2) Customization — worked with Leadcore to trim unnecessary modules (cellular modem) from existing architectures. (3) Specification definition — DJI now defines the complete chip specification (which modules, what interfaces, what performance targets) and Leadcore or other foundries tape out chips to those specs. DJI controls the specification, firmware, and protocol stack — the foundry handles physical implementation. This gives DJI chip-level differentiation without requiring a full semiconductor design team.
5. What can other drone/robotics companies learn from DJI’s chip strategy?
The key lesson: control the protocol layer, not necessarily the silicon. DJI never built their own chip from scratch — they progressively increased control over specifications while relying on partners for fabrication. What gave DJI its competitive moat was OcuSync’s protocol and algorithms (low-latency coding, adaptive modulation, frequency hopping), not the chip itself. Companies building wireless systems should focus on protocol innovation first, then optimize the hardware platform through customization once volume justifies the investment. Aomway applies this same principle in our wireless product development — we prioritize protocol and algorithm innovation over custom silicon, partnering with chip vendors for hardware optimization when needed.
Building a wireless video transmission system or selecting chips for drone applications? Contact Aomway at [email protected] — we provide wireless system architecture consulting, chip selection guidance, and complete video link design for industrial UAV platforms.
