Loading in 2 Seconds...
Loading in 2 Seconds...
Comments for The Proposed Unified Hardware Abstraction PAR. Comments from the Yes Votes. - I think it is important and valuable to have a focus on application, OS/driver and system SW power modeling from a top down approach.
Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.
- I think it is important and valuable to have a focus on application, OS/driver and system SW power modeling from a top down approach.
- It is also imperative that the top down energy proportional standardization encompass a variety of operating system, SW environments inclusive but not restricted to: OSx/OSiand proprietary operating systems/applications/SW, and runtimes Linux and Linux derivatives inclusive of virtual machines such as JVM Java EE, SE, ME, Dalvik etc... and associated runtime components Windows based families inclusive of ACPI and PCI ennumeration and device discovery and system level SW and any/all RTOS, or embedded, server SW stacks
- I would also like to ensure that we minimize any overlap or redundancy with existing efforts in other stds such as P1801, 1801-2013 or others outside of IEEE/Accelerainitiatives
- So a focus on SW, and ensuring compatibility with existing standards are the keys items in the scope. We may need to narrow the scope to produce useful specs.stds
- Also that the efforts are either OS/operating system agnostic or enlist, include and represent multiple industry SW/system implementations and not restricted or preclude other device discovery and hardware management frameworks.
In particular keep in mind compatibility with other low power standards/efforts, without overlapping.
Accompanying Word Document
The scope is unusually broad and appears to be redundant with much found in other standards such as 1801, CPF, and Liberty. I also found the motivation, as described in the purpose - "Energy proportionality is only possible..." and "Energy proportionality requires new design, verification, modeling, management and testing abstractions..." to be overstated and diffuse.
This is not to say that we don't need enhanced modeling standards to accommodate software and system level views of power - we certainly do. I just believe that we don't need to boil the ocean to get there. Moreover, I believe that the likelihood of a successful standard development is compromised without a more sharply defined proposal.