Hi All
Going back over old ground can yield much of interest. In the various interviews shared of Pat Gelsinger speaking at and being interviewed at the Intel Foundry event he mentions the intention to challenge Nvidia’s GPU dominance in the Cloud which has eroded Intel’s CPU market share. He did not say in detail how Intel intended to revive its CPUs to achieve this resurgence.
The following previously highlighted findings by these German Researchers could be pointing to what Intel has planned for AKIDA technology:
“C. Research Directions for SNNs in Machine Learning
As we approach the forefront of the current research, we suggest that the future lies in incorporating SNN-based models into mainstream applications,
focusing on extending large- scale applications beyond image classification, refining training techniques, and optimizing mapping strategies for diverse hardware platforms. Efforts in standardization and compati- bility with existing frameworks will enhance SNN adoption across industries. Additionally, delving into novel applications, particularly in real-time scenarios, and further integration with transformer-based models could unlock new frontiers for SNN research. Emphasizing energy efficiency, economic viability, and scalability will be pivotal for solidifying SNNs as a key player in the evolving landscape of machine learning and neuromorphic computing.
V. HARDWARE AND SOFTWARE INTEGRATION
In this section, we discuss the hardware and software challenges for integrating neuromorphic hardware platforms into AI data centres.
A. Hardware integration
1) Status quo: So far, the following neuromorphic sys- tems have been integrated at large scale into data centers: SpiNNaker 1 [20], TrueNorth (NS16-4e) [83], Loihi (Po- hoiki Springs) [84], BrainScaleS-1 [85] and Tianjic [32]. All systems use slide-in modules with custom printed circuit boards (PCBs) for integration into standard 19” server racks. Typically, the neuromorphic chips are accessed via Ethernet, only the TrueNorth NS16e-4 uses PCIe for communication with the host chip. Baseboard Management Controller (BMC) or similar controllers are used for booting and monitoring the boards. All platforms also include field-programmable gate arrays (FPGAs) or system-on-chips (SoCs), most often as middleware between host computers and neuromorphic sys- tems. Some systems have already integrated a host CPU, e.g., Pohoiki Springs or the Tianjic server, while the other systems require external host CPU servers for the configuration and control of the neuromorphic systems and for preprocessing.
As an exception, BrainChip offers PCIe boards for integrating their Akida chips with CPU servers. This represents an- other option for integrating neuromorphic computing systems into data centres, similar to normal GPUs. Note, however, that this might limit the size of neuromorphic models that can be implemented compared to the larger systems discussed above.
2) Conclusion: The above examples show that a variety of neuromorphic systems have been successfully integrated into standard data center server racks. Thus, technically, the hardware integration does not pose a problem. Yet, we observe a diversity in how large neuromorphic systems are assembled into server boards, e.g., many of them leverage FPGAs or SoCs as middleware. These extra devices and the host CPU add a power overhead to the very energy-efficient neuromorphic systems. Optimizing for system-level efficiency of AI compute servers, these components need to be included when perform- ing benchmarking on AI workloads. Another requirement for the industry-level deployment of neuromorphic chips is high reliability and robustness. The chips and boards need to be designed for a 24/7 operation, e.g., the server board should keep working if a single chip or processor fails. Replacement parts should be available for a long period.“
https://arxiv.org/pdf/2402.02521
Watching a little bit of AKIDA magic being added to every Intel CPU on one of the new Intel process nodes would be fun to see as a Brainchip shareholder.
My opinion only DYOR
Fact Finder
With love from Russia Vlad.
..
