DisplayPort LTTPR: Transparent vs Non-Transparent Mode

This article explains the difference between transparent mode and non-transparent mode in DisplayPort Link Training Tunable PHY Repeaters (LTTPRs). Introduced in the DisplayPort 1.4a and 2.0/2.1 standards, LTTPRs regenerate high-speed video signals over longer physical paths. The fundamental difference between the two operational modes lies in whether the repeater actively communicates with the video source during the link training process (non-transparent mode) or acts as an invisible signal conditioner that passes training communications straight through to the display (transparent mode).

Understanding LTTPR

An LTTPR is an active component, such as a retimer or redriver, placed between a DisplayPort Source (like a GPU) and a DisplayPort Sink (like a monitor). At elevated data rates such as HBR3 and UHBR (Ultra-High Bit Rate), trace attenuation and jitter degrade signal integrity. LTTPRs mitigate these losses, operating in either Transparent Mode or Non-Transparent Mode.

Transparent Mode

In Transparent Mode, the LTTPR is invisible to both the Source and the Sink:

  • Link Training Process: The Source and the Sink execute standard link training directly with one another over the AUX (Auxiliary) channel. The LTTPR does not intercept, modify, or respond to AUX transactions.
  • Repeater Function: The LTTPR operates primarily as an analog amplifier or snooping retimer. It monitors the link training transactions passively to infer data rates, or relies on automatic gain control and fixed equalization settings to clean up and retransmit the high-speed data lanes.
  • Visibility: The DPCD (DisplayPort Configuration Data) registers of the repeater are not exposed to the Source. The system topology appears as a single point-to-point connection between the Source and Sink.
  • Limitations: Because the Source cannot adjust the repeater's internal equalizer or drive levels directly, transparent mode may yield suboptimal link margins on complex channels with severe signal degradation.

Non-Transparent Mode

In Non-Transparent Mode, the LTTPR explicitly reveals itself to the Source and takes an active role in link negotiation:

  • Link Training Process: Link training is divided into discrete physical segments. Instead of a single end-to-end training sequence, training occurs hop-by-hop. For example, in a connection with one repeater, the Source trains the link between itself and the LTTPR, while the LTTPR concurrently or sequentially trains the link to the Sink.
  • DPCD Visibility: The LTTPR exposes a standardized set of DPCD addresses (from 0xF0000 to 0xF00FF for up to eight repeaters). The Source detects the presence and count of repeaters along the transmission path.
  • PHY Optimization: The Source reads the receiver capabilities of the LTTPR and actively adjusts voltage swing, pre-emphasis, and equalization parameters for that specific segment. This ensures optimal signal integrity even at maximum UHBR bandwidths.
  • Latency and Overhead: Non-transparent mode introduces slightly more AUX channel communication during link establishment, but it provides deterministic and reliable link stability.

Key Differences at a Glance

  • Awareness: Non-transparent mode allows the Source to detect and configure each repeater; transparent mode hides the repeater entirely.
  • Training Architecture: Non-transparent mode uses segment-by-segment (hop-by-hop) link training; transparent mode relies on end-to-end link training between the Source and Sink.
  • PHY Control: Non-transparent mode enables software-driven, precise tuning of the repeater's physical layer parameters via DPCD; transparent mode depends on automatic or fixed hardware settings within the repeater.
  • High-Speed Reliability: Non-transparent mode is the preferred and often necessary implementation for DisplayPort 2.1 UHBR speeds, where strict channel parameters require exact equalization tuning across every segment of the link.