PLX is a small, general-purpose, subword-parallel instruction set architecture (ISA) designed at Princeton University, Department of Electrical Engineering. PLX is designed to be a simple yet high-performance ISA for multimedia information processing.
There are several versions of the PLX architecture and the toolset. The documentation for each of these is provided below.
This is the current development version of PLX and therefore it is not yet publicly available. PLX developers can login to access these documents with their download password. Please fill out this form and email to rblee@princeton.edu to obtain a password.
For a quick comparison of PLX versions, see this file
The architecture documentation is divided into two parts: (1) ISA reference describing the instructions and general features of PLX, and (2) ISA encoding describing how the PLX instructions map to 32-bit instruction words.
Architecture
Toolset
Not available yet.
Other
A preliminary version of the encoding of PLX floating-point instructions. Floating-point support is planned to be included in the next version of PLX. This encoding is subject to change.
Architecture
Toolset
Architecture
Toolset
PLX 1.0 was released as an ISA specification only and had no accompanying software tools.
| Attachment | Size |
|---|---|
| PLXPAXapplicationform.xls | 18 KB |
| PLX-version-history.txt | 1.73 KB |
PLX architecture and
tools have been used in some of the papers published by PALMS. These are listed
below:
This paper discusses the new permutation instructions included in PLX 1.2:
This paper discusses how to use PLX multimedia instructions to do matrix transpose: