UCIe 2.0 for 3D Chip Structures Offers up to 75 Times More Bandwidth
UCIe 2.0 enhances chip design with 75 times more bandwidth than UCIe 1.1, supporting 3D structures, 4 GT/s transfer speeds, and collaboration among major manufacturers, aiming for 2028 advancements.
Read original article
The Universal Chiplet Interconnect 2.0 (UCIe 2.0) is set to revolutionize chip design by offering up to 75 times more bandwidth than its predecessor, UCIe 1.1. This new interconnect specification is tailored for advanced 3D chip structures, where chiplets are stacked vertically, enhancing speed and power efficiency. The UCIe 2.0 protocol allows for more communication channels between chiplets, significantly improving bandwidth density and reducing power consumption. It supports a transfer speed of up to 4 GT/s per channel, with the potential for bandwidth density to reach 300 terabytes per second per square millimeter at a bump pitch of one micron. The UCIe 2.0 also introduces tools for managing and testing chiplets, improving their manufacturability and reliability. Major industry players, including Nvidia, Intel, and TSMC, are part of the UCIe consortium, which aims to standardize this technology. While there is no definitive timeline for market-ready chips based on UCIe 2.0, the consortium is committed to releasing new specifications annually. The shift to 3D designs is expected to surpass traditional monolithic die structures by 2028, marking a significant evolution in semiconductor technology.
- UCIe 2.0 offers up to 75 times more bandwidth than UCIe 1.1.
- The new specification supports advanced 3D chip structures with improved power efficiency.
- It allows for a transfer speed of up to 4 GT/s per channel.
- Major chip manufacturers are collaborating within the UCIe consortium.
- The transition to 3D designs is projected to outpace traditional chip designs by 2028.
Related
TSMC experimenting with rectangular wafers vs. round for more chips per wafer
TSMC is developing an advanced chip packaging method to address AI-driven demand for computing power. Intel and Samsung are also exploring similar approaches to boost semiconductor capabilities amid the AI boom.
Intel CPU with Optical OCI Chiplet Demoed with 4Tbps of Bandwidth and 100M Reach
Intel unveiled a new CPU with an Optical Compute Interconnect chiplet, offering 4Tbps bandwidth and 100m reach. The chiplet merges Photonic and Electronic Integrated Circuits, featuring built-in lasers for improved server communication, especially in AI servers. Intel plans to scale the technology up to 32Tbps, focusing on power savings and enhanced bandwidth. This advancement aligns with the industry's move towards co-packaged optics for better efficiency and performance, marking a significant shift in server architecture towards power efficiency and high-speed data transfer.
What Would You Do with a 16.8M Core Graph Processing Beast?
TSMC collaborates with Intel, MIT, and AWS to develop HIVE, a graph processing unit named PIUMA with 16.8 million cores. The advanced chip aims to efficiently process neural networks, featuring a custom RISC-based architecture and photonics interconnect for scalability.
Intel vs. Samsung vs. TSMC
Competition intensifies among Intel, Samsung, and TSMC in the foundry industry. Focus on 3D transistors, AI/ML applications, and chiplet assemblies drives advancements in chip technology for high-performance, low-power solutions.
DARPA slaps down credit card for 3D military chiplets $840M ought to be enough?
DARPA awarded $840 million to TIE at the University of Texas-Austin for 3DHI technology to enhance military systems like radar and UAVs. The $1.4 billion project involves industry partnerships for defense innovation.
1 comments“We start at 4000 gigabytes per second per square millimeter, and we go all the way up to 300,000 gigabytes per second — or 300 terabytes per second — per square millimeter once we hit one micron, a huge amount of bandwidth,” said Das Sharma."
Related
TSMC experimenting with rectangular wafers vs. round for more chips per wafer
TSMC is developing an advanced chip packaging method to address AI-driven demand for computing power. Intel and Samsung are also exploring similar approaches to boost semiconductor capabilities amid the AI boom.
Intel CPU with Optical OCI Chiplet Demoed with 4Tbps of Bandwidth and 100M Reach
Intel unveiled a new CPU with an Optical Compute Interconnect chiplet, offering 4Tbps bandwidth and 100m reach. The chiplet merges Photonic and Electronic Integrated Circuits, featuring built-in lasers for improved server communication, especially in AI servers. Intel plans to scale the technology up to 32Tbps, focusing on power savings and enhanced bandwidth. This advancement aligns with the industry's move towards co-packaged optics for better efficiency and performance, marking a significant shift in server architecture towards power efficiency and high-speed data transfer.
What Would You Do with a 16.8M Core Graph Processing Beast?
TSMC collaborates with Intel, MIT, and AWS to develop HIVE, a graph processing unit named PIUMA with 16.8 million cores. The advanced chip aims to efficiently process neural networks, featuring a custom RISC-based architecture and photonics interconnect for scalability.
Intel vs. Samsung vs. TSMC
Competition intensifies among Intel, Samsung, and TSMC in the foundry industry. Focus on 3D transistors, AI/ML applications, and chiplet assemblies drives advancements in chip technology for high-performance, low-power solutions.
DARPA slaps down credit card for 3D military chiplets $840M ought to be enough?
DARPA awarded $840 million to TIE at the University of Texas-Austin for 3DHI technology to enhance military systems like radar and UAVs. The $1.4 billion project involves industry partnerships for defense innovation.