Vendor:
Global UniChip Corp. (GUC)
Category:
Custom
TSMC CLN5FF GLink 2.3LL Die-to-Die PHY
IGAD2DY04A is a high-speed die-to-die interface PHY which transmits data through TSMC packaging solutions: Integrated Fan-Out (In…
Overview
IGAD2DY04A is a high-speed die-to-die interface PHY which transmits data through TSMC advanced packaging solutions: Integrated Fan-Out (InFO) with RDL interconnect and Chip-on-Wafer-on-Substrate (CoWoS®) with silicon interposer. IGAD2DY04A contains 56 TX lanes and 56 RX lanes per slice and supports 8 slices in one PHY. Each TX/RX lane can support up to 17.2 Gbps data rate. In summary, IGAD2DY04A offers a full-duplex data transmission with extremely low power and up to 963.2 Gbps data rate per slice in both directions.
Each TX/RX slice contains PMA and PCS modules. PMA supports serialization, de-serialization, data transmission, eye training, and lane repair functions. PCS provides data bus inversion (DBI), CRC check, and FIFO functions. One PLL is also included in IGAD2DY04A to generate an 8.5 GHz high-speed clock for data transmission.
IGAD2DY04A is designed and fabricated in TSMC 5 nm FF CMOS process with 1.2 V analog supply voltage for PLL/PMA and 0.75 V analog/digital supply voltages. Independent low power mode for PLL and slices is available.
Each TX/RX slice contains PMA and PCS modules. PMA supports serialization, de-serialization, data transmission, eye training, and lane repair functions. PCS provides data bus inversion (DBI), CRC check, and FIFO functions. One PLL is also included in IGAD2DY04A to generate an 8.5 GHz high-speed clock for data transmission.
IGAD2DY04A is designed and fabricated in TSMC 5 nm FF CMOS process with 1.2 V analog supply voltage for PLL/PMA and 0.75 V analog/digital supply voltages. Independent low power mode for PLL and slices is available.
Key features
- 56 full-duplex lanes per slice
- 8 slices are included in the analog hard macro
- 1:8 mode with 448-bit data width or 1:16 mode with 963-bit data width for user interface
- VALID and READY handshake mechanism
- Flow control between TX and RX
- Data bus inversion
- CRC check
- Data replay to ensure no error found in RX
- Programmable data scrambling for smoothing current consumption profile
- Built-in test pattern and checker
- Lane repair
- EHOST: APB, I2C, and JTAG register interface
- Built-in PLL
- ~0.3 (TBD) pJ/bit power consumption
- 1.2 V analog supply voltage for PLL/PMA and 0.75 V analog/digital supply voltage
- Independent low power mode for analog blocks
- Operating junction temperature: -40 °C ~ 125 °C
- Process: TSMC 5 nm 0.75 V/1.2 V CMOS LOGIC FinFET Process
- Metal Scheme: 1P17M (1X_h_1Xb_v_1Xe_h_1Ya_v_1Yb_h_5Y_vhvhv_2Yy2Yx2R)
- Analog hard macro size: 3050 um x 1076 um (3.281 mm2) for North-South direction macro and
- 1113 um x 3050 um (3.39 mm2) for East-West direction macro
- Logic gate count: 4.2 M
- Supports both horizontal and vertical GDS orientation
- Special Layer & Device: High R Resistance, ULVT
Silicon Options
| Foundry | Node | Process | Maturity |
|---|---|---|---|
| TSMC | 5nm | N5FF | Available on request |
Specifications
Identity
Part Number
IGAD2DY04A
Vendor
Global UniChip Corp. (GUC)
Type
Silicon IP
Files
Note: some files may require an NDA depending on provider policy.
Provider
HQ:
Taiwan
Learn more about Custom IP core
Would custom memory ease your SoC design?
A case study shows how custom memory design can adapt to functional, timing, or layout requirements of a particular SoC.
Enabling RISC-V Vector Code Generation in MLIR through Custom xDSL Lowerings
The growing adoption of RISC-V in high-performance and scientific computing has increased the need for performance-portable code targeting the RISC-V Vector (RVV) extension. However, current compiler infrastructures provide limited end-to-end support for generating optimized RVV code from high-level representations to low-level implementations. In particular, existing MLIR distributions lack practical lowering paths that map high-level abstractions to RVV intrinsics, limiting their applicability for production-ready RISC-V kernels. This paper presents a compilation approach that combines MLIR with xDSL to bridge the missing lowering stages required for RVV code generation.
Breaking the Bandwidth Barrier: Enabling Celestial AI’s Photonic Fabric™ with Custom ESD IP on TSMC’s 5nm Platform
As AI workloads scale exponentially, bandwidth has become the new bottleneck. This article explores how Celestial AI’s Photonic Fabric™ redefines interconnect performance, and how Sofics’ low-cap ESD and power clamp IP enabled its integration on TSMC’s 5nm platform.
A New Era for Edge AI: Codasip’s Custom Vector Processor Drives the SYCLOPS Mission
See how Codasip Studio enables SYCLOPS to implement a customizable RISC-V Vector processor, accelerating the future of open-source, high-performance AI compute at the Edge.
RISC-V basics: The truth about custom extensions
The era of universal processor architectures is giving way to workload-specific designs optimized for performance, power, and scalability. As data-centric applications in artificial intelligence (AI), edge computing, automotive, and industrial markets continue to expand, they are driving a fundamental shift in processor design.
Frequently asked questions about Custom Die-to-Die IP cores
What is TSMC CLN5FF GLink 2.3LL Die-to-Die PHY?
TSMC CLN5FF GLink 2.3LL Die-to-Die PHY is a Custom IP core from Global UniChip Corp. (GUC) listed on Semi IP Hub. It is listed with support for tsmc Available on request.
How should engineers evaluate this Custom?
Engineers should review the overview, key features, supported foundries and nodes, maturity, deliverables, and provider information before shortlisting this Custom IP.
Can this semiconductor IP be compared with similar products?
Yes. Buyers can compare this product with similar semiconductor IP cores or IP families based on category, provider, process options, and structured technical specifications.