NOV152026
Conference
SC26
- November 15 – 20, 2026
- Chicago, Illinois, USA
- Organized by SC Conference Series
Public source
NVIDIA NVLink-C2C
NVIDIA NVLink-C2C is a publicly named high-bandwidth coherent chip-to-chip interconnect for connecting processors, accelerators and custom silicon.
Overview
Official NVIDIA pages describe NVLink-C2C as an interconnect used in chiplet and advanced multi-die systems, including integration in advanced packages. NVIDIA NVLink, listed separately, is the system-level GPU interconnect used within servers and across racks with NVLink Switch. Signaling rates, protocol details and product-generation mappings should be confirmed on NVIDIA's current NVLink-C2C documentation.
Technical capabilities
Distinct from NVLink GPU-to-GPU scale-up fabric, which is listed separately
Lifecycle and categories
- Lifecycle stages
- NoC and interconnect
Related solutions
AMD Infinity Fabric
Compute & AcceleratorsInfinity Fabric is AMD’s coherent on-package and multi-die interconnect, linking CPU core-complex dies to a central I/O die and, in multi-socket systems, across packages. Official EPYC architecture documents describe CCD–IOD GMI links and xGMI.
View solutionArm Neoverse CMN S3
NoC IPLicensable Arm Neoverse CMN S3 coherent interconnect IP. The page documents AMBA 5 CHI and multichip support. A UCIe PHY is not included.
View solutionArteris FlexWay Interconnect IP
NoC IPNon-coherent network-on-chip IP for cost-efficient, low-power IoT and microcontroller SoCs.
View solutionArteris Ncore Cache Coherent Interconnect IP
NoC IPLicensable cache-coherent NoC IP. The page documents AMBA CHI and ACE. A Multi-Die option requires a separate die-to-die controller and PHY. A UCIe PHY is not included.
View solutionAyar Labs TeraPHY Optical I/O Chiplet
PhotonicsAyar Labs TeraPHY is an optical I/O chiplet that combines silicon photonics with CMOS technology to provide high-bandwidth, low-latency chip-to-chip and package-to-package connectivity for large-scale computing systems.
View solutionBaya Systems WeaveIP NoC Fabric IP
NoC IPBaya Systems WeaveIP is chiplet-ready Network-on-Chip fabric IP for SoC and multi-chiplet systems. The official site describes a protocol-agnostic unified transport for coherent, non-coherent, and switch-fabric traffic paired with WeaverPro modeling software.
View solutionRelated events