How 128b/132b Encoding Reduces DisplayPort Overhead
DisplayPort standards prior to version 2.0 relied on legacy 8b/10b encoding, a scheme that consumed a significant portion of physical bandwidth purely for signal transmission management. With the introduction of DisplayPort 2.0 and 2.1, the standard shifted to 128b/132b channel encoding for its Ultra-High Bit Rate (UHBR) transmission modes. This architectural change cuts protocol transmission overhead from 20% down to approximately 3%, dramatically increasing the usable payload bandwidth available for higher resolutions, faster refresh rates, and uncompressed color depth.
The Problem with Legacy 8b/10b Encoding
In legacy DisplayPort standards (up to DisplayPort 1.4a), data transmission relies on 8b/10b line coding. To maintain DC balance (ensuring an equal number of ones and zeros) and to guarantee enough signal transitions for the receiver's clock recovery circuit, every 8-bit byte of data is mapped to a 10-bit symbol.
Because two extra bits are added to every single byte, the transmission overhead is fixed at 20%:
\[\text{Overhead} = \frac{10 - 8}{10} = \frac{2}{10} = 20\%\]
Under an 8b/10b scheme, a connection operating at a raw physical layer bandwidth of 32.4 Gbps (like DisplayPort 1.4 HBR3 across four lanes) yields only 25.92 Gbps of actual payload capacity. One-fifth of the entire physical capacity is lost strictly to link maintenance.
How 128b/132b Encoding Operates
DisplayPort's UHBR modes replace byte-by-byte encoding with a block-based structure. Instead of encoding individual 8-bit units, the transmitter groups data into massive 128-bit payloads and prepends a 4-bit header, producing a 132-bit transmission block.
The 4-bit header identifies the block type (such as control data or pixel data) and provides frame synchronization. Because 128 bits of actual data are delivered in every 132-bit packet, the overhead drops significantly:
\[\text{Overhead} = \frac{132 - 128}{132} = \frac{4}{132} \approx 3.03\%\]
This shifts the efficiency of the physical link from 80% under 8b/10b to approximately 96.97% under 128b/132b.
Maintaining Signal Integrity Without Added Bit Overhead
Legacy 8b/10b encoding uses its extra 2 bits per byte to prevent long strings of identical bits (consecutive zeros or ones) and to keep the direct current (DC) bias of the cable neutral.
To achieve the same reliability without adding massive bit overhead, 128b/132b encoding decouples framing from signal conditioning:
- Hardware Scrambling: Before framing, data is processed using a Pseudo-Random Binary Sequence (PRBS) scrambler. This mathematical scrambling randomizes the bit patterns, effectively preventing long runs of identical bits and eliminating DC bias without inflating packet size.
- Header Synchronization: The 4-bit header is purposefully structured with distinct transition states so the receiver can lock onto block boundaries reliably.
Real-World Bandwidth Impact
The transition to 128b/132b encoding directly dictates how much raw wire speed is translated into display performance:
- UHBR 10: Features a raw link rate of 40 Gbps across four lanes. With 128b/132b encoding, it delivers roughly 38.69 Gbps of usable throughput. Under the old 8b/10b standard, that same raw rate would only deliver 32.0 Gbps.
- UHBR 20: Reaches an 80 Gbps raw link rate. After applying 128b/132b overhead, the interface delivers roughly 77.37 Gbps of usable payload bandwidth, losing less than 3 Gbps to overhead.
By utilizing larger data blocks and algorithmic scrambling instead of per-byte symbol expansion, 128b/132b encoding reclaims nearly 17% of total link capacity, enabling DisplayPort to drive ultra-high-resolution setups such as uncompressed 8K at 60Hz or multi-monitor configurations over a single cable.