Browse Prior Art Database

High-Speed Rendering Method

IP.com Disclosure Number: IPCOM000115595D
Original Publication Date: 1995-May-01
Included in the Prior Art Database: 2005-Mar-30
Document File: 2 page(s) / 51K

Publishing Venue

IBM

Related People

Kohda, Y: AUTHOR [+3]

Abstract

Disclosed is a high-speed rendering methodology which controls wide width memory bus by narrow width local bus and memory controllers on Graphics Card. Memory Controller has a circuit which generates large depth color data from narrow control data. The graphics commands such as Line Draw, Fill Stiple, Text et. al., use a few kind of color at the same time. If their commands are issued, this methodology is very useful.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 89% of the total text.

High-Speed Rendering Method

      Disclosed is a high-speed rendering methodology which controls
wide width memory bus by narrow width local bus and memory
controllers on Graphics Card.  Memory Controller has a circuit which
generates large depth color data from narrow control data.  The
graphics commands such as Line Draw, Fill Stiple, Text et.  al., use
a few kind of color at the same time.  If their commands are issued,
this methodology is very useful.

      Fig. 1 shows the example of this system.  Control data which
correspond to some pixels are transferred from system I/F LSI (1) via
local Bus (2).  Memory control LSIs (3) have color Registers (4)
which are corresponded to one pixel of display.  Its depth is more
than each control data depth and maximum depth depends on Memory Bus
bandwidth and Local Bus Bandwidth.  Memory control LSI selects one
color from each control data by multiplexer (5) and then writes them
to memories (6).  Fig. 2 illustrates the relation of bus width
between local bus and memory bus.  In this case, if cycle time of
local bus is the same with cycle time of memory bus, bandwidth of
memory bus is four times as large as bandwidth of local bus.  But
memory write data is prepared enough to fill memory bus bandwidth.
Therefore, high speed rendering becomes available even though low
speed local bus.

      Since the number of memory controller is not limited, color
depth or number of pixels which are written at the same time c...