Scalable Ultra-High Throughput VESA DSC 1.2b Decoder

Overview

The UHT-DSC-D core is a scalable, ultra-high throughput, advanced DSC 1.2b decoder, compliant to the VESA Display Stream Compression (DSC) 1.2b standard. It supports decoding of 4:4:4, 4:2:2 and 4:2:0 video streams, in 8 to 16 bits per component color depths.

The core is designed for enabling ultra-high frame rate SD, HD and Ultra HD video decoding up to 10K resolutions, even in medium-end ASIC or FPGA silicon. The scalability of this IP core enables highly cost-effective silicon implementations of applications that need to handle massive pixel rates and resolutions. The UHT-DSC-D is available for ASIC or AMD-Xilinx, Efinix, Intel, Lattice and Microchip FPGA and SoC based designs.

The UHT-DSC-D is very easy to use and integrate in a system, designed for using only internal memory blocks and with simple, fully controllable and FIFO-like, streaming input and output interfaces. It requires minimal host intervention as it only needs to be programmed once per video sequence. Once programmed, it can decode an arbitrary number of video frames without the need of any further intervention or assistance by the host system CPU.

Key Features

VESA DSC 1.2b Compliant, Complete and Standalone Operation
- Full compliance with the VESA DSC 1.2b specification
- Backwards compatible with VESA DSC 1.1
- RGB and YCbCr color space formats
- 8, 10, 12, 14 and 16 bits per color component dynamic range
- Native support for 4:4:4, 4:2:2 and 4:2:0 sampling formats
- Up to 16 slices per line
- Scalable architecture with configurable number of internal, parallel decoding engines
  - 3 pixel/clock per decoding engine processing for 4:4:4 sampling format
  - 6 pixels/clock per decoding engine processing for 4:2:2 and 4:2:0 sampling formats
- Operation without external memory
  - Very low internal memory requirements (a few image lines)
- Ultra-low latency performance (sub-line latency)
- CPU/GPU-less, complete and standalone implementation
Trouble-Free Technology Map and Implementation
- Self-contained RTL design
- No internal tri-states
- Strictly positive edge triggered design
- D-type only Flip-Flops
- Fully synchronous operation per clock domain
- Safe CDC transfers between clock domains
- No need for special timing constraints
  - No false or multi-cycle paths within the same clock domain
  - No CDC transfers that need to be specially constrained
  - No other specially constrained timing paths

Block Diagram

Deliverables

Clear-text RTL sources for ASIC designs, or pre-synthesized and verified Netlist for FPGA and SoC devices
Release Notes, Design Specification and Integration Manual documents
Bit Accurate Model (BAM) and test vector generation binaries, including sample scripts
Pre-compiled RTL simulation model and gate-level simulation netlist for the FPGA Netlist license
Self-checking testbench environment sources, including sample BAM generated test cases
Simulation and sample Synthesis (for ASICs) or Place & Route (for FPGAs) scripts

Technical Specifications

Request Info