KR102261250B1 - Ray tracing device and method for reducing the amount of data transmitted between devices - Google Patents

Abstract

The present invention relates to a ray tracing device and method for reducing the bandwidth between devices. The ray tracing device comprises a software unit and a dedicated hardware unit, wherein the software unit, through software, processes (a) a ray information construction operation, (b) a ray generation operation, (c) a collision point calculation operation and (d) a shading operation, and the dedicated hardware unit, as dedicated hardware interlocking with the software unit, includes a ray generation unit, a traversal and intersection test unit, a reorder buffer unit, a collision point calculation unit and a ray setup unit. Therefore, the present invention allows the dedicated hardware to independently process the collision point calculation and the ray setup separately from the operations of the software, thereby minimizing data transmission between a hardware device and a software device.

Description

RAY TRACING DEVICE AND METHOD FOR REDUCING THE AMOUNT OF DATA TRANSMITTED BETWEEN DEVICES

The present invention relates to a ray tracing technology for reducing bandwidth between devices, and more particularly, to a ray tracing apparatus and method for reducing bandwidth between devices for improving ray tracing performance by reducing large-scale data transmission between devices. .

Ray tracing is largely composed of the steps of ray generation, tree traversal & intersection test, hit point calculation, and shading, and is recursive. can be performed with

Recently, as high-quality CG images are required in movies and games, the tree visit & intersection test and collision point calculation, which require the most computational amount in ray tracing, are processed in dedicated hardware (HW), and the rest of the ray generation and shading is performed. A software (SW) method of processing is emerging.

In particular, as shading requires a complex process according to various lighting models as the quality increases, it is difficult to implement all options in hardware, so it is easy to process with software. In this case, performance improvement can be expected because two independent devices can be performed in parallel, but large-scale data transmission is required between the two devices, and the performance can be degraded due to increased latency between devices. .

Korean Patent Registration No. 10-1076807 (2011.10.19)

An embodiment of the present invention is to provide a ray tracing apparatus and method for reducing bandwidth between devices for improving ray tracing performance by reducing large-scale data transmission between devices.

An embodiment of the present invention reduces the bandwidth between devices capable of minimizing data transmission between a hardware device and a software device by independently processing the collision point calculation and ray setup in dedicated hardware separately from the operation of software in the ray tracing process. An object of the present invention is to provide a ray tracing apparatus and method.

An embodiment of the present invention is a ray tracing apparatus and method for reducing bandwidth between devices that can efficiently perform interworking with a software device in charge of software processing by including a reorder buffer unit as a dedicated hardware device for ray tracing would like to provide

Among the embodiments, the ray tracing device for reducing the bandwidth between devices is through software, (a) a ray information construction operation for building ray information for ray generation, and (b) a ray using the ray information A ray generation operation for generating a first ray for ray tracing, (c) a collision point calculation operation for calculating a collision point based on collision triangle information that collides with a second ray, and (d) a collision point calculation operation on the collision point A ray generating unit that generates the second ray for a visit intersection test based on the ray information as a software unit unit that processes a shading operation for calculating a color value, respectively, and dedicated hardware interworking with the software unit unit and a visited intersection test unit that performs a visited intersection test operation based on the second ray, and an out-of-order result output from the visited intersection test operation is changed to an in-order result A reorder buffer unit and a collision point calculation unit that calculates a collision point based on the second ray and collision triangle information and a shadow ray or secondary ray to generate information about the ray to generate the ray It includes a dedicated hardware unit unit composed of a lay-up unit that transmits to the unit.

The software unit unit generates ray information about the shadow ray or the secondary ray when the occurrence of a shadow ray or a secondary ray is detected in the shading operation step and transmits the ray information construction operation step to the ray setup ray setup (Set-up) operation can be further processed.

The software unit builds information for ray generation for each pixel in the frame in the ray information construction operation step as the ray information and transmits it one-way to the dedicated hardware unit unit while passing it to the ray generation operation step, and the ray information is If it is associated with the shadow ray or the secondary ray, the one-way transmission may be omitted.

The software unit may generate a primary ray as the first ray from a viewpoint based on the ray information in the ray generation operation step and transfer the data value of the primary ray to the collision point calculation operation step. have.

The software unit may calculate the coordinates of the collision point between the collision triangle and the first ray based on the collision triangle information and the first ray in the collision point calculation operation step.

The visited intersection test unit searches an Acceleration Structure (AS) based on the second ray to determine a leaf node associated with the second ray, and a visited operation for all triangles included in the leaf node. A test may be performed to each process an intersection test operation to determine a collision triangle that collides with the second ray.

The reorder buffer unit includes a first step of storing the second ray output from the ray generation unit, a second step of receiving the collision triangle information output from the visited intersection test unit, and a second step associated with the collision triangle information By sequentially performing the third step of outputting the corresponding collision triangle information according to the priority of the ray, the collision triangle information can be transmitted one way in accordance with the order of the ray information transmitted from the software unit unit.

The collision point calculation unit calculates the coordinates of the collision point between the collision triangle and the second ray based on the collision triangle information and the second ray, and the ray setup unit calculates the coordinates of the collision point between the shadow ray or the It is possible to determine whether the secondary ray is generated or not to generate information about the corresponding ray.

In embodiments, the ray tracing method for reducing the bandwidth between devices includes (a) constructing ray information for ray generation through the software unit and one-way transmitting to the dedicated hardware unit, (b) ) through the dedicated hardware unit, generating a second ray for a visited intersection test based on the ray information, (c) through the dedicated hardware unit, performing a visited intersection test operation based on the second ray (d) changing the out-of-order result output from the visited intersection test operation to an in-order result through the dedicated hardware unit unit to give the software unit one-way transmitting, (e) through the software unit unit, generating a first ray using the ray information and calculating a collision point based on the first ray and collision triangle information, and (f) the software unit unit through, calculating a color value on the collision point.

The ray tracing method includes the steps of (d') calculating a collision point between the second ray and the collision triangle information output from the visited intersection test operation independently of the operation of the software unit through the dedicated hardware unit unit, and (e) ') through the dedicated hardware unit unit, generating information about a shadow ray or secondary ray based on the collision point independently of the operation of the software unit unit and transferring the information to the ray generation unit further may include

The step (a) includes building information for ray generation as the ray information for each pixel in a frame, transferring the ray information to the ray generation operation step, and transmitting the ray information to the dedicated hardware unit unit It may include the step of transmitting.

The step (c) may include transmitting the result of the visited intersection test operation to the reorder buffer unit and the collision point calculation unit, respectively.

The step (d) stores the second ray input to the visited intersection test unit, receives the collision triangle information output from the visited intersection test unit, and the priority of the second ray associated with the collision triangle information It may include the step of transmitting the collision triangle information one way in response to the order of the ray information transmitted from the software unit by one side by outputting the corresponding collision triangle information according to the.

The step (f) may further include generating ray information about the shadow ray or the secondary ray when the occurrence of a shadow ray or a secondary ray is detected and transferring the generated ray information to the step (a). have.

The ray tracing method may further include (g) outputting, through the software unit, an image of the frame when color values for all pixels in the frame are determined.

The disclosed technology may have the following effects. However, this does not mean that a specific embodiment should include all of the following effects or only the following effects, so the scope of the disclosed technology should not be construed as being limited thereby.

A ray tracing apparatus and method for reducing bandwidth between devices according to an embodiment of the present invention is a ray tracing process between a hardware device and a software device by independently processing the collision point calculation and ray setup in dedicated hardware separately from the operation of the software. Data transmission can be minimized.

A ray tracing apparatus and method for reducing bandwidth between devices according to an embodiment of the present invention include a reorder buffer unit as a dedicated hardware device for ray tracing to efficiently perform interworking with a software device in charge of software processing. can

1 is a view for explaining an embodiment of a ray tracing process performed by a ray tracing apparatus.
2 is a diagram for explaining an embodiment of a KD tree as an acceleration structure used in a ray tracing process.
3 is a view for explaining a general example of a ray tracing process performed in a ray tracing apparatus.
4 is a view for explaining an embodiment of processing ray tracing through interworking of SW operation and dedicated HW.
5 is a diagram illustrating a ray tracing apparatus for reducing bandwidth between apparatuses according to an embodiment of the present invention.
6 and 7 are flowcharts illustrating an embodiment of a ray tracing process for reducing bandwidth between devices performed in the ray tracing apparatus of FIG. 5 .

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiment described in the text. That is, since the embodiment may have various changes and may have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea. In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such effects, it should not be understood that the scope of the present invention is limited thereby.

On the other hand, the meaning of the terms described in the present application should be understood as follows.

Terms such as “first” and “second” are for distinguishing one component from another, and the scope of rights should not be limited by these terms. For example, a first component may be termed a second component, and similarly, a second component may also be termed a first component.

When a component is referred to as being “connected” to another component, it should be understood that the component may be directly connected to the other component, but other components may exist in between. On the other hand, when it is mentioned that a certain element is "directly connected" to another element, it should be understood that the other element does not exist in the middle. On the other hand, other expressions describing the relationship between elements, that is, "between" and "immediately between" or "neighboring to" and "directly adjacent to", etc., should be interpreted similarly.

The singular expression is to be understood as including the plural expression unless the context clearly dictates otherwise, and terms such as "comprises" or "have" refer to the embodied feature, number, step, action, component, part or these It is intended to indicate that a combination exists, and it should be understood that it does not preclude the possibility of the existence or addition of one or more other features or numbers, steps, operations, components, parts, or combinations thereof.

Identifiers (eg, a, b, c, etc.) in each step are used for convenience of description, and the identification code does not describe the order of each step, and each step clearly indicates a specific order in context. Unless otherwise specified, it may occur in a different order from the specified order. That is, each step may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be embodied as computer-readable codes on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices in which data readable by a computer system is stored. . Examples of the computer-readable recording medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, HDD, SSD, and the like. In addition, the computer-readable recording medium may be distributed in a network-connected computer system, and the computer-readable code may be stored and executed in a distributed manner.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Terms defined in the dictionary should be interpreted as being consistent with the meaning of the context of the related art, and cannot be interpreted as having an ideal or excessively formal meaning unless explicitly defined in the present application.

1 is a view for explaining an embodiment of a ray tracing process performed by a ray tracing apparatus.

Referring to FIG. 1 , the ray tracing method may correspond to a rendering method according to global illumination. This may mean that light reflected or refracted from another object also affects the image of the current object. As a result, reflection, refraction, and shadow effects are naturally provided, so that realistic 3D images can be created.

In FIG. 1 , the ray tracing apparatus may first generate a primary ray (P) from a camera position for each pixel and perform a calculation to find an object that meets the corresponding ray. The ray tracing device generates a reflection ray (R) for a reflection effect or a refraction ray (F) for a refraction effect at the location where the ray and the object meet if the object that meets the ray has a property of reflection or refraction. You can create a shadow ray (S) in the light direction for a shadow effect.

At this time, when a shadow ray directed to the corresponding light position and an object meet, a shadow is created, otherwise a shadow is not created. Reflected ray and refractive ray are called secondary ray, and the ray tracing device can perform a calculation for each ray to find an object that intersects that ray. This process may be performed recursively by the ray tracing device.

2 is a diagram for explaining an embodiment of a KD tree as an acceleration structure used in a ray tracing process.

Referring to FIG. 2 , acceleration such as a k-depth tree or BVH (Bounding Volume Hierarchy) generated with full geometry data (consisting of triangle coordinates) in order to perform ray tracing An Acceleration Structure (AS) is essential. Therefore, it is necessary to build the AS before performing ray tracing. Building such an acceleration structure (AS build) may take a lot of time because a large amount of computation is required.

In FIG. 2, the overall configuration diagram of the kd-tree is described. The kd-tree may correspond to a binary tree having a hierarchical structure with respect to the divided space. A kd-tree can be composed of an inner node (including a top node) and a leaf node, and a leaf node corresponds to a space containing objects that intersect with the corresponding node. can That is, the kd-tree may correspond to a kind of spatial partitioning tree.

On the other hand, an internal node may have a spatial region based on a bounding box, and the spatial region may be divided into two regions and allocated to two lower nodes. As a result, an inner node may be composed of a sub-tree of a division plane and two regions divided through it, and a leaf node may include only a series of triangles.

On the other hand, the position p that divides the space may correspond to a point where the cost (number of node visits, the number of times to calculate whether a triangle intersects, etc.) for finding a triangle that hits a random ray is minimized, , the currently most used method to find the corresponding position p may correspond to a surface area heuristic (SAH).

3 is a view for explaining a general example of a ray tracing process performed in a ray tracing apparatus.

Referring to FIG. 3 , the ray tracing process may be performed recursively, (i) Ray Generation step (S310), (ii) Visit (Traversal) & Intersection Test step (S330) , (iii) a collision point calculation (Hit Point Calculation) step (S350) and (iv) shading (Shading) & ray setup (Next Ray Set-up) step (S370) may include.

First, in the ray generation step, a primary ray may be generated from a viewpoint position for each pixel (step S310 ). In the next step, a leaf node that meets the ray can be found by searching an acceleration structure (AS) such as a kd-tree and a bounding volume hierarchy (BVH) (step S330). Here, triangle information is stored in the leaf node.

In the next intersection test step, it is possible to test whether all triangles in the met leaf nodes meet the ray (step S330). This process can be repeated until a triangle that intersects the ray is found. After that, it is possible to calculate a hit point in the hit point calculation step for the triangle met with the ray (step S350).

In the next step, shading, a color value on a ray-triangle hit point may be calculated (step S370). If it is necessary to generate a shadow ray due to lighting or a secondary ray due to the material of the collided triangle, this information is determined in the next ray set-up stage and the ray It may be transmitted to the generation (ray generation) step. In the ray generation step, a shadow ray and a secondary ray may be generated based on this information (step S310).

4 is a view for explaining an embodiment of processing ray tracing through interworking of SW operation and dedicated HW.

Referring to FIG. 4 , the ray tracing process processes the tree traversal & intersection test and hit point calculation, which require the most amount of computation, in dedicated hardware (HW), and the rest Part ray generation and shading may be performed in a software (SW) manner.

First, in the ray generation step, which is a part of the software, a primary ray may be generated from a position of a view point for each pixel. Data values for these lads may be transmitted to hardware (HW). This value may include a vector value of the ray, a ray type value for distinguishing the type of ray (eg, primary ray, secondary ray, shadow ray, etc.).

In the hardware part, tree traversal & intersection test and hit point calculation can be performed. Hit point information and collision triangle information, which are results performed in the hardware part, may be transmitted to the software part. This is useful under the assumption that hit triangles are determined in the order of the input ray, otherwise additional information (e.g. ray id to distinguish which ray is input) is more useful. may be needed

In the software shading step, it is possible to calculate a color value on a ray-triangle hit point. If it is necessary to generate a shadow ray due to lighting or to generate a secondary ray due to the material of the collided triangle, this information is determined in the next ray set-up stage. It can be sent to the ray generation stage. In the ray generation step, a shadow ray and a secondary ray may be generated based on this information. The data values for the generated ray can be sent back to the hardware part.

In FIG. 4 , performance improvement can be expected because the hardware (HW) part and the software (SW) part can be performed in parallel, but large-scale data transmission is required between the two devices, so that the latency between the devices is reduced. Stretching can degrade performance.

5 is a diagram illustrating a ray tracing apparatus for reducing bandwidth between apparatuses according to an embodiment of the present invention.

Referring to FIG. 5 , the ray tracing apparatus 500 may include a software unit unit 510 and a dedicated hardware unit unit 530 . The software unit unit 510 may include a ray information construction module 511 , a ray generation module 513 , a collision point calculation module 515 , and a shading & ray setup module 516 . The dedicated hardware unit unit 530 may include a ray generation unit 531 , a visited intersection test unit 533 , a reorder buffer unit 535 , a collision point calculation unit 537 , and a ray setup unit 539 . In an embodiment, the software unit 510 may be implemented including a central processing unit (CPU), and the dedicated hardware unit 530 may be implemented including a graphics processing unit (GPU).

The software unit unit 510, through software, (a) a ray information construction operation for building ray information for ray generation and (b) a first for ray tracing using ray information A ray generation operation that generates a ray, (c) a collision point calculation operation that calculates a collision point based on collision triangle information that collides with the second ray, and (d) a shading operation that calculates a color value on the collision point can be processed individually. That is, each operation may be independently processed by each module included in the software unit unit 510 .

In one embodiment, the software unit 510 generates ray information about a shadow ray or a secondary ray when the occurrence of a shadow ray or a secondary ray is detected in the shading operation step to the ray information construction operation step. It can further process the ray set-up operation to be transmitted. That is, the shading operation and the ray setup operation may be performed by the shading & ray setup module 517 .

On the other hand, the entire process of ray tracing can be performed recursively. In this process, if shadow ray generation is required due to lighting or secondary ray generation is required due to the material of the collided triangle, the information is stored in software. It may be processed in the ray setup operation step by the unit unit 510 . The information processed in the ray setup operation step may be transferred to the ray information construction operation step, and in the ray information establishment operation step, the information may be transmitted to the external dedicated hardware unit unit 530 or transferred to another internal operation step. have.

In one embodiment, the software unit unit 510 builds the information for ray generation as ray information for each pixel in the frame in the ray information construction operation step and transmits it to the ray generation operation step to the dedicated hardware unit unit 530 . One-way transmission is possible (S551). The ray information construction operation may be performed by the ray information construction module 511 . That is, the data movement from the software unit unit 510 to the dedicated hardware unit unit 530 may be limited to the ray information generated by the software unit unit 510 . In this case, the ray information may include information for generating a primary ray.

In one embodiment, the software unit 510 may omit the one-way transmission operation to the dedicated hardware unit 530 when the ray information is associated with a shadow ray or a secondary ray in the ray information construction operation step. That is, the ray tracing apparatus 500 is implemented to perform collision point calculation operation and ray setup operation by itself in the dedicated hardware unit unit 530 , thereby reducing the data transmission amount from the software unit unit 510 to the dedicated hardware unit unit 530 . can be significantly reduced. The software unit unit 510 transmits only ray information related to the primary ray to the external dedicated hardware unit unit 530, and ray information related to the shadow ray or the secondary ray may be transmitted only to the internal operation step.

In an embodiment, the software unit 510 generates a primary ray as a first ray from a viewpoint position based on ray information in the ray generation operation step and calculates a collision point calculation operation for a data value of the corresponding primary ray can be passed in stages. The ray generation operation may be performed by the ray generation module 513 .

At this time, the first ray may correspond to a ray generated according to a software operation by the software unit 510 as a ray generated based on ray information, and a data value related to the ray is a vector (vector) value of the ray , may include a ray type value for distinguishing the type of ray. On the other hand, the order of the first ray generated by the ray generation module 513 of the software unit 510 is the same as the order of the second ray generated by the ray generation unit 531 of the dedicated hardware unit 530 . can do.

In one embodiment, the software unit 510 may calculate the coordinates of the collision point between the collision triangle and the first ray based on the collision triangle information and the first ray in the collision point calculation operation step. The collision point calculation operation may be performed by the collision point calculation module 515 . The collision point calculation operation may not affect the overall performance of ray tracing even if it is processed in software by the software unit 510 because the amount of calculation is very small compared to the tree visit & intersection test.

The dedicated hardware unit unit 530 is dedicated hardware that interoperates with the software unit unit 510, and based on the ray generation unit 531 and the second ray that generates the second ray for the visiting cross test based on the ray information. A visited cross test unit 533 that performs a visited cross test operation, and a reorder buffer that changes an out-of-order result output from the visited cross test operation into an in-order result. ) unit 535, a collision point calculation unit 537 that calculates a collision point based on the second ray and collision triangle information, and a ray generation unit by generating information about a shadow ray or secondary ray ( It may be configured with a ray setup unit 539 that forwards to 531 .

That is, the dedicated hardware unit unit 530 processes each of the tree visit & cross test operations that require a lot of ray generation operation and calculation amount in the ray tracing process through dedicated hardware, and at the same time calculates the collision point separately from the operation of the software. By independently processing the ray setup operation and the ray setup operation, the amount of data transmission with the software unit unit 510 can be significantly reduced. At this time, the data transmitted from the dedicated hardware unit 530 to the software unit 510 corresponds only to information on the collided triangles, and there is no need to transmit information on the collision point ( S553 ).

In one embodiment, the visited intersection test unit 533 searches an Acceleration Structure (AS) based on the second ray to determine a leaf node associated with the second ray and a visiting operation for all triangles included in the leaf node. Each of the intersection test operations for determining a collision triangle that collides with the second ray can be processed by performing an intersection test on the ray. That is, the visited intersection test unit 533 may independently process the visited operation and the intersection test operation using a pre-built acceleration structure.

In one embodiment, the reorder buffer unit 535 is a first step of storing the second ray output from the ray generation unit 531 and input to the visited intersection test unit 533, output from the visited intersection test unit 533 Ray transmitted from the software unit unit 510 by sequentially performing the second step of receiving the collision triangle information and the third step of outputting the corresponding collision triangle information according to the priority of the second ray associated with the collision triangle information Corresponding to the order of the information, the collision triangle information may be transmitted unilaterally.

More specifically, the reorder buffer unit 535 performs a visit cross test in the step where the dedicated hardware unit 530 receives the ray information from the software unit 510 and generates the second ray through the ray generation unit 531 . The data regarding the second ray transmitted to the unit 533 may be allocated and stored. Further, the reorder buffer unit 535 may receive the collision triangle information from the visited intersection test unit 533 to determine that the processing operation for the associated second ray is complete.

If the processed second ray corresponds to the oldest second ray in the reorder buffer (that is, the second ray having the highest priority), the reorder buffer unit 535 provides the second ray as an output result. Also, only the collision triangle information may be transmitted to the software unit 510 by the dedicated hardware unit 530 .

In general, the visited cross test performed by the visited cross test unit 533 of the dedicated hardware unit 530 may be performed out-of-order for high-performance processing. That is, the hit triangle may not be determined in the order of the input second ray. The tree visit & intersection test requested from the software unit 510 is performed out-of-order, while the dedicated hardware unit 530 has a reorder buffer to be completed in-order. It may be implemented including a unit 535 . For example, the reorder buffer unit 535 may be implemented as a reorder buffer that is frequently used in a pipelined RISC processor.

In one embodiment, the collision point calculation unit 537 is configured to perform an interpolation between the collision triangle and the second ray based on the collision triangle information output from the visited intersection test unit 533 and the second ray generated by the ray generation unit 531 . The coordinates of the collision point can be calculated. That is, the dedicated hardware unit unit 530 may be implemented by including the collision point calculation unit 537 as dedicated hardware for performing the collision point calculation operation independently of the collision point calculation operation performed by the software unit unit 510 . .

In an embodiment, the ray setup unit 539 may generate information about the corresponding ray by determining whether a shadow ray or a secondary ray is generated at the coordinates of the collision point. That is, the dedicated hardware unit unit 530 may be implemented including the ray setup unit 539 as dedicated hardware for performing the ray setup operation independently of the ray setup operation performed by the software unit unit 510 . On the other hand, the ray setup unit 539 may transmit information about the shadow ray or the secondary ray to the ray generation unit 531 .

6 and 7 are flowcharts illustrating an embodiment of a ray tracing process for reducing bandwidth between devices performed in the ray tracing apparatus of FIG. 5 .

Referring to FIG. 6 , the ray tracing apparatus 500 may generate ray information for ray generation through the software unit unit 510 and transmit it one-way to the dedicated hardware unit unit 530 (step S610). . In one embodiment, the software unit unit 510 constructs information for ray generation as ray information for each pixel in a frame in the ray information construction operation step, and transfers the ray information to the ray generation operation step while dedicating the ray information. It can be unidirectionally transmitted to the hardware unit unit 530 .

In addition, the ray tracing apparatus 500 may generate a second ray for a visiting cross test based on the ray information through the dedicated hardware unit 530 (step S620).

Also, the ray tracing apparatus 500 may perform a visiting cross test operation based on the second ray through the dedicated hardware unit 530 (step S630). In an embodiment, the dedicated hardware unit unit 530 may transmit the result of the visited intersection test operation to the reorder buffer unit 535 and the collision point calculation unit 537 respectively through the visited intersection test unit 533 .

In addition, the ray tracing apparatus 500 converts an out-of-order result output as an in-order cross-test operation through the dedicated hardware unit unit 530 into an in-order result to form a software unit. One-way transmission is possible to the unit 510 (step S640).

In addition, the ray tracing apparatus 500 may generate a first ray using ray information through the software unit unit 510 and calculate a collision point based on the first ray and collision triangle information (step S650). The ray tracing apparatus 500 may calculate a color value on the collision point through the software unit 510 (step S660). When the color values for all pixels in the frame are determined through the software unit unit 510, the ray tracing apparatus 500 may output an image related to the frame (step S670).

Referring to FIG. 7 , the ray tracing device 500 through the collision point calculation unit 537 of the dedicated hardware unit 530 , the collision output as a visit intersection test operation separately from the operation of the software unit 510 . It is possible to calculate the collision point between the triangle information and the second ray (step S710). The ray tracing device 500 through the ray setup unit 539 of the dedicated hardware unit 530, independent of the operation of the software unit 510, a shadow ray or secondary based on the collision point. The ray-related information may be generated and transmitted to the ray generating unit 531 (step S730). The ray tracing apparatus 500 may generate a shadow ray or a secondary ray through the ray generation unit 531 of the dedicated hardware unit 530 (step S750).

Although the above has been described with reference to the preferred embodiments of the present invention, those skilled in the art can variously modify and change the present invention within the scope without departing from the spirit and scope of the present invention as set forth in the claims below. You will understand that it can be done.

500: ray tracing device
510: software unit unit
511: ray information building module 513: ray generating module
515: Collision Point Calculation Module 517: Shading & Ray Setup Module
530: dedicated hardware unit 531: ray generation unit
533: visit cross test unit 535: reorder buffer unit
537: collision point calculation unit 539: ray setup unit

Claims (15)

Through software, (a) a ray information construction operation for constructing ray information for ray generation and (b) ray generation for generating a first ray for ray tracing using the ray information Software that processes the calculation and (c) the collision point calculation calculation for calculating the collision point based on the collision triangle information that collides with the second ray, and (d) the shading calculation for calculating the color value on the collision point unit unit; and
Dedicated hardware interoperating with the software unit, a ray generation unit that generates the second ray for a visited intersection test based on the ray information, and a visited intersection test that performs a visited intersection test operation based on the second ray A reorder buffer unit that changes an out-of-order result output from the unit and the visited intersection test operation to an in-order result, and the second ray and collision triangle information A dedicated hardware unit consisting of a collision point calculation unit that calculates a collision point based on a collision point calculation unit and a ray set-up unit that generates information about a shadow ray or a secondary ray and transmits it to the ray generation unit including,
Each of the software unit unit and the dedicated hardware unit unit is implemented independently of each other, and only one-way transmission of the ray information and the collision triangle information is performed between each other.
The method of claim 1, wherein the software unit unit
When the occurrence of a shadow ray or a secondary ray is detected in the shading operation step, ray information about the shadow ray or the secondary ray is generated and transmitted to the ray information construction operation step Set-up (Set-up) ) A ray tracing device that reduces bandwidth between devices, characterized in that it further processes computation.
The method of claim 1, wherein the software unit unit
In the ray information construction operation step, information for ray generation for each pixel in the frame is constructed as the ray information and transmitted to the ray generation operation step while being transmitted to the dedicated hardware unit unit,
When the ray information is associated with the shadow ray or the secondary ray, the ray tracing device for reducing the bandwidth between devices, characterized in that the one-way transmission is omitted.
4. The method of claim 3, wherein the software unit unit
In the ray generation operation step, a primary ray is generated as the first ray from a viewpoint based on the ray information, and a data value of the corresponding primary ray is transferred to the collision point calculation operation step. A ray tracing device that reduces bandwidth between devices.
The method of claim 1, wherein the software unit unit
Ray tracing device reducing the bandwidth between devices, characterized in that calculating the coordinates of the collision point between the collision triangle and the first ray based on the collision triangle information and the first ray in the collision point calculation operation step.
The method of claim 1, wherein the visited cross-test unit is
The second ray is searched for an acceleration structure (AS) based on the second ray, and a visit operation to determine a leaf node associated with the second ray and an intersection test for all triangles included in the leaf node are performed. A ray tracing device that reduces bandwidth between devices, each processing an intersection test operation to determine a collision triangle that collides with two rays.
The method of claim 1, wherein the reorder buffer unit is
The first step of storing the second ray output from the ray generation unit, the second step of receiving the collision triangle information output from the visited intersection test unit, and the priority of the second ray associated with the collision triangle information Accordingly, by sequentially performing the third step of outputting the corresponding collision triangle information, reducing the bandwidth between devices, characterized in that the one-way transmission of the collision triangle information in response to the order of the ray information transmitted from the software unit unit ray tracing device.
According to claim 1,
the collision point calculation unit calculates the coordinates of the collision point between the collision triangle and the second ray based on the collision triangle information and the second ray,
The ray setup unit determines whether the shadow ray or the secondary ray is generated at the coordinates of the collision point to generate information about the corresponding ray.
In the method performed in a ray tracing device that is implemented independently of each other and interoperates with a software unit unit and a dedicated hardware unit unit,
(a) constructing ray information for ray generation through the software unit unit and transmitting one-way to the dedicated hardware unit unit;
(b) generating, through the dedicated hardware unit, a second ray for a visited intersection test based on the ray information;
(c) performing, through the dedicated hardware unit, a visited intersection test operation based on the second ray;
(d) changing an out-of-order result output from the visited intersection test operation to an in-order result through the dedicated hardware unit unit and one-way transmitting to the software unit unit ;
(e) through the software unit, generating a first ray using the ray information and calculating a collision point based on the first ray and collision triangle information; and
(f) calculating, through the software unit, a color value on the collision point.
10. The method of claim 9,
(d') calculating, through the dedicated hardware unit, a collision point between the second ray and the collision triangle information output from the visited intersection test operation independently of the operation of the software unit; and
(e') through the dedicated hardware unit unit, generating information about a shadow ray or secondary ray based on the collision point independently of the operation of the software unit unit and transmitting it to the ray generation unit Ray tracing method for reducing bandwidth between devices further comprising a.
10. The method of claim 9, wherein (a) step
building information for ray generation as the ray information for each pixel in a frame;
transferring the ray information to the ray generation operation step; and
Ray tracing method for reducing bandwidth between devices comprising the step of one-way transmitting the ray information to the dedicated hardware unit.
10. The method of claim 9, wherein step (c) is
and transmitting the result of the visited intersection test operation to each of the reorder buffer unit and the collision point calculation unit.
10. The method of claim 9, wherein step (d) is
Store the second ray input to the visited intersection test unit, receive the collision triangle information output from the visited intersection test unit, and the collision triangle information according to the priority of the second ray associated with the collision triangle information By outputting a ray tracing method for reducing bandwidth between devices, comprising the step of transmitting the collision triangle information one-way corresponding to the order of the ray information transmitted one-way from the software unit unit.
10. The method of claim 9, wherein step (f) is
When generation of a shadow ray or a secondary ray is detected, generating ray information about the shadow ray or the secondary ray and transferring the ray information to the step (a) between devices A ray tracing method that reduces bandwidth.
10. The method of claim 9,
(g) outputting, through the software unit, an image of the frame when the color values of all pixels in the frame are determined.

Images