Graduate Thesis Or Dissertation

 

Offloading of I/O transactions in current CPU architectures Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/s1784r46j

Descriptions

Attribute NameValues
Creator
Abstract
  • IO transactions within a computer system have evolved along with other system components (i.e., CPU, memory, video) from programmed IO (PIO). In current mainstream systems (spanning from HPC to mobile) the IO transactions are CPU-centric descriptor-based DMA transactions. The key benefit is that slower IO devices can DMA write system receive traffic to system memory and DMA read system transmit data at slower device throughput relative to the CPU. With the advent of more cores in a CPU, power restrictions and latency concerns, we show this approach has limitations and based on measurements we propose alternatives to descriptor-based DMA IO transactions. We explore and quantify performance improvement in three options:1) iDMA: Embedded smalller core to offload DMA descriptor processing from the larger application-oriented cores, reducing latency up to 16% and increasing bandwidth per pin up to 17%.2) Hot-Potato: Where latency is a concern we re-visit using WC-buffers for direct IO CPU transactions and avoiding CPU hardware changes. While keeping a specialized receive IO device DMA engine, we reduce latency for small messages by 1.5 μs.3) Device2Device: For applications moving data between devices, we propose how to bypass the CPU, improving latency, power, and CPU utilization.
  • Keywords: I/O, Computer input-output equipment, Computer architecture, CPU, Computer storage devices, system, performance
License
Resource Type
Date Available
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Rights Statement
Publisher
Peer Reviewed
Language
Replaces

Relationships

Parents:

This work has no parents.

In Collection:

Items