Cache Coherent Interconnect for Accelerators
GPTKB entity
Statements (35)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:personal_computer
|
| gptkbp:abbreviation |
gptkb:CCIX
|
| gptkbp:announced |
2016
|
| gptkbp:competitor |
gptkb:NVLink
gptkb:OpenCAPI gptkb:CXL gptkb:Gen-Z |
| gptkbp:enables |
low-latency communication
multi-vendor interoperability high-bandwidth communication improved system performance shared memory access cache coherency across devices direct memory access between devices reduced data movement overhead scalable system architectures |
| gptkbp:governedBy |
gptkb:CCIX_Consortium
|
| https://www.w3.org/2000/01/rdf-schema#label |
Cache Coherent Interconnect for Accelerators
|
| gptkbp:purpose |
enabling cache coherency between CPUs and accelerators
|
| gptkbp:supports |
heterogeneous computing
PCI Express physical layer data sharing between processors |
| gptkbp:used_in |
gptkb:machine_learning
cloud computing data centers high-performance computing big data analytics cloud infrastructure AI workloads |
| gptkbp:usedFor |
CPU-GPU interconnect
CPU-FPGA interconnect CPU-NPU interconnect CPU-accelerator interconnect |
| gptkbp:bfsParent |
gptkb:CCIX
|
| gptkbp:bfsLayer |
7
|