International Business Machines Corporation

Michael A Aranda

Timothy J Ebbers

Yoshio Iida

Andrew D Bowen

Tushar Patel

Robert Tamlyn

Nathan R Hiltebeitel

Terence W Lindgren

Randall L Henderson

Thuy-Linh Tran Bui

Akishi Kamel

James Bernard Keenan, Iii

Taggart H Robertson

A graphics system and method with which thick graphic primitives are efficiently drawn by minimizing dependence on drawing algorithms that require appreciable setup time. The method contemplates drawing a thick primitive in which an offset or displacement value is first calculated, based upon the thickness of the graphic primitive. The offset is approximately one half of the thickness of the primitive. Following calculation of the offset value, line drawing parameter values are determined for a line that is parallel to the origin line and displaced from the origin line in a minor axis direction by the displacement or offset value. A loop is then repeated for each grip point in the major axis range of the line. The loop includes an initial step in which a boundary pixel of the thick graphic primitive is drawn using the line drawing algorithm and the line drawing parameter values calculated for the offset line. After the boundary pixel has been drawn, one or more adjacent pixels are drawn using a stepping routine in which the mirior axis coordinate of the selected pixel is either decremented or incremented, depending upon the slope of the line, to write the pixels adjacent the boundary pixel. In this fashion, the present invention draws a thick primitive as a sequence of segments that are parallel to the minor axis of the origin line. In the preferred embodiment, the line drawing routine is preferably comprised of a Bresenham line drawing algorithm or a similar derivative algorithm. In the preferred embodiment, the displacement D is equal to FLOOR((W&#x2212;1)/2), where W is the thickness of the primitive and FLOOR(X) is the integer portion of X.

Method and apparatus for drawing thick graphic primitives

A modified frame buffer and pixel variable read-modify-write method are described for a high performance computer graphics system. Pixel variables are initially classified as decision variables, intensity variables or decision/intensity variables. Only decision/intensity variables requiring a read-modify-write operation, are stored in dual interleaved DRAMs for improved bandwidth. Decision variables and intensity variables each utilize a single address/data bus per video RAM module in the frame buffer, while decision/intensity variables require dual address/data buses for accessing the interleaved memory banks. Enhanced bandwidth is obtained with a minimization of raster engine I/O requirements.

Categorized pixel variable buffering and processing for a graphics system

The selection in a dual port memory device of data from a serial access memory register having a lower byte and an upper byte of data is described herein. In one embodiment, the register is partitioned lengthwise into two sections, corresponding to, for example, a frame buffer A and a frame buffer B. On each serial clock cycle, frame buffer A or frame buffer B for each byte of data may be selected from the register. Each of the selected bytes of data are then passed to a serial output port. In another embodiment, the lower byte of data corresponds to, for example, a frame buffer A and the upper byte corresponds to a frame buffer B. Then either the upper byte or lower byte of data is selected to be output on the serial port. In yet a further embodiment, the serial access memory register is partitioned lengthwise into two sections, each section corresponding to, for example, a frame buffer and the bytes of data correspond to another buffer, then either the lower byte or upper byte is selected to be output on the serial port.

SAM data selection on dual-ported DRAM devices

A processing system for identifying operator selection of a graphics object in a system using parallel pixel generation. Parallel pixel generation results in a fragment comprising a series of M pixels being generated at each processing cycle. The image fragment must be tested against an operator defined selection area to determine whether the object being generated falls within the operator selection area. Fragments are initially classified as totally within or totally outside of the area. If a fragment intersects the operator selection area, mask is applied for each boundary of the operator defined area. Selection is signalled based upon the logical combination of the masked image fragments. Operator selection or picking occurs without the overhead of maintaining a pick RAM for the X and Y dimensions.

Pick function implementation in a parallel processing system

A system for generating multiple pixels in a single machine cycle employs a plurality of parallel vector generators. Each of the parallel generators is initialized with an error term which is calcualted in accordance with Bresenham's algorithm. The signs of these error terms are then used to determine the region within the first octant of the coordinate system which contains the function for which the pixels are to be generated. The region data, in turn, determine two selectable values for an increment which is to be added to a running error term for each of the parallel generators as multiple pixels are simultaneously generated. The choice of the two possible values to be added to the error term is dependent upon the sign of the error term itself. The sign of the running error term for each vector generator is utilized to form a sequence of binary data which represents the incremental changes in the pixel positions as the pixels are being generated. In a preferred embodiment, the binary data is translated into pixel position data which is then stored in a memory device. Any generated pixel data which cannot be stored in the machine cycle in which it is generated is re-routed for storage in later cycles.

Growing community of inventors

Saugerties, NY, United States of America

Michael A Aranda