Chiplet Strategy is Key to Addressing Compute Density Challenges
By Balaji Baktha, Ventana Micro Systems
EETimes (September 28, 2021)
Data center workloads are quickly evolving, demanding high compute density with varying mixes of compute, memory and IO capability. This is driving architectures that are moving away from a one-size-fits-all monolithic solution to disaggregated functions that can be independently scaled for specific applications.
It is imperative to adopt the latest process nodes to deliver the needed compute density. However, doing so with traditional monolithic SoCs presents an inherent disadvantage due to escalating costs and time to market challenges resulting in unfavorable economics. To address this dilemma, chiplet-based integration strategies are emerging where compute can benefit from the most advanced process nodes, while application-specific memory and IO integrations can reside on mature trailing process nodes.
To read the full article, click here
Related Semiconductor IP
- Securyzr™ neo Core Platform
- 112G Multi-SerDes
- SHA3 Cryptographic Hash Cores
- ISO/IEC 7816 Verification IP
- 50MHz to 800MHz Integer-N RC Phase-Locked Loop on SMIC 55nm LL
Related White Papers
- Programmable Logic Holds the Key to Addressing Device Obsolescence
- Reconfiguring Design -> Reconfigurability: Designer's key strategy
- Clock domain modeling is essential in high density SoC design
- Addressing the new challenges of ASIC/SoC prototyping with FPGAs
Latest White Papers
- Practical Considerations of LDPC Decoder Design in Communications Systems
- A Direct Memory Access Controller (DMAC) for Irregular Data Transfers on RISC-V Linux Systems
- A logically correct SoC design isn’t an optimized design
- AI in VLSI Physical Design: Opportunities and Challenges
- cMPI: Using CXL Memory Sharing for MPI One-Sided and Two-Sided Inter-Node Communications