IEEE Intelligent Vehicles Symposium, 2004
DOI: 10.1109/ivs.2004.1336381
|View full text |Cite
|
Sign up to set email alerts
|

RSVP/sub TM/ : an automotive vector processor

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
5
0

Publication Types

Select...
2
2

Relationship

0
4

Authors

Journals

citations
Cited by 4 publications
(5 citation statements)
references
References 3 publications
0
5
0
Order By: Relevance
“…In [7,8], stream register files are used to store and assemble data as contiguous streams. In [9,10,11], dedicated stream units prefetch and gather stream elements from different memory locations. These stream units use stream descriptors, which are described later in this paper.…”
Section: Stream Descriptorsmentioning
confidence: 99%
See 2 more Smart Citations
“…In [7,8], stream register files are used to store and assemble data as contiguous streams. In [9,10,11], dedicated stream units prefetch and gather stream elements from different memory locations. These stream units use stream descriptors, which are described later in this paper.…”
Section: Stream Descriptorsmentioning
confidence: 99%
“…Fig. 7 shows a block diagram of RSVP, a detailed description of its internal architecture and programming model is beyond the scope of this paper, but more information can be found in [9,10]. Stream descriptors are used in the Vector Stream Unit (VSU) to prefetch data ahead of the processing unit.…”
Section: The Rsvp™ Architecturementioning
confidence: 99%
See 1 more Smart Citation
“…It complements the data processing instructions of the host, but does not participate in the control flow. In programming the RSVP™ architecture, the user describes the shape and location of input and output streams in memory and describes a data flow graph of the operations to be performed, thus allowing programmers to provide mechanisms that exploit data and instruction level parallelism [6,8].…”
Section: A Stream Processorsmentioning
confidence: 99%
“…These access patterns are fixed and known at compile time so it is possible to minimize data movement and overlap it with computation. Speedups from 2 to 20 over a scalar processor for stream kernels and applications have been found using the stream computation model [6,8].…”
Section: A Stream Processorsmentioning
confidence: 99%