A total of seven companies, including ARM, AMD, Huawei, IBM, Qualcomm, Mellanox and Xilinx have joined forces to work together in developing an interface allowing different manufacturers of processors to be able to communicate with each other by sharing main memory.
ARM and IBM will collaborate with AMD to tackle the Data Center Market
Undoubtedly, although in theory the agreement between the seven companies has been easy to carry out, it is a really ambitious project because of the difficulty of performing a consistent interface cache between two or four homogeneous chips and CPUs. Allow devices to communicate and share data across different implementations of CPUs, FPGAs, GPUs and networking chips will be a challenge to overcome.
Anyway, no matter how complex it may be, the benefits that the companies involved would gain can be quite exorbitant. If these companies fail to move forward with this project, then it will provide plug-and-play compute and network acceleration for whatever processor you choose, while providing much better performance than the maximum bandwidth that PCIe is able to deliver right now. The new barrier will bring a consistent speed specification for the Cache Coherent Interconnect for Accelerators (CCIX).
“AMD strongly supports development of open standards to make heterogeneous computing more pervasive,” says AMD’s Gerry Talbot, AMD corporate fellow and vice president of I/O and circuit technologies.
“A ‘one size fits all architecture’ approach to data center workloads does not deliver the required performance and efficiency,” says ARM’s Lakshmi Mandyam, “CCIX enables more optimized solutions by simplifying software development and deployment of applications that benefit from specialized processing and hardware off-load, delivering higher performance and value to data center customers.”
“CCIX will leverage existing server interconnect infrastructure and deliver higher bandwidth, lower latency, and cache coherent access to shared memory,” said Gaurav Singh of Xilinx, “this will result in a significant improvement in the usability of accelerators and overall performance and efficiency of data center platforms.”
Applications in the data center have created to a point where processing power won’t have the capacity to stay aware of these prerequisites. Below you can see some examples of these applications:
- Data analytics
- Machine learning
- Wireless 4G/5G
- In-memory database processing
- Video analytics
- Network processing