技术领域Technical field
本申请属于3D-IC热分析领域，具体是一种基于先进工艺下芯片动态调节分辨率的数据可视化方法。This application belongs to the field of 3D-IC thermal analysis, specifically a data visualization method that dynamically adjusts the resolution of the chip based on advanced technology.
背景技术Background technique
随着芯片制作工艺步入3nm，集成电路技术已经逐渐接近了摩尔定律物理极限的边缘。后摩尔时代的到来让半导体行业面临着严峻的挑战，三维集成电路(3D-IC)技术被视为延续摩尔定律的有效技术之一。3D-IC在芯片设计与模拟阶段的广泛应用产生了越来越多的工程数据，如何理解这些重要的数据成为解决问题的一个重要环节。计算机计算能力和交互性能的大幅提高，为数据处理和分析提供了媒介。As the chip manufacturing process enters 3nm, integrated circuit technology has gradually approached the edge of the physical limit of Moore's Law. The arrival of the post-Moore era has caused the semiconductor industry to face severe challenges. Three-dimensional integrated circuit (3D-IC) technology is regarded as one of the effective technologies to continue Moore's Law. The widespread application of 3D-IC in the chip design and simulation stages has generated more and more engineering data. How to understand these important data has become an important part of solving problems. The substantial improvement in computer computing power and interactive performance provides a medium for data processing and analysis.
在3D-IC热分析领域，常使用有限元分析法或有限体积分析法对芯片结构进行划分，为了在不降低精度的前提下加快运算速度，通常在温度变化较大的区域加密网格，而在其他温度变化平稳的区域降低网格密度。因此，获得的三维空间温度场通常是分布不均匀的数据。In the field of 3D-IC thermal analysis, finite element analysis or finite volume analysis is often used to divide the chip structure. In order to speed up the calculation without reducing accuracy, the mesh is usually densified in areas with large temperature changes. Reduce mesh density in other areas where temperature changes are smooth. Therefore, the obtained three-dimensional spatial temperature field is usually unevenly distributed data.
与此同时，当前大多数的研究都集中在单一变量场可视化领域，而在科学研究和工程实践中，多物理场分析和可视化场景应用广泛存在。尤其是随着大规模科学和工程计算领域应用的兴起，科学数据的复杂性呈现出空前的、爆炸式的增长，这些科学数据不仅数据量巨大，而且通常都包含多变量、时变以及高分辨率的特点。At the same time, most current research focuses on the field of single variable field visualization, while in scientific research and engineering practice, multi-physics analysis and visualization scenario applications are widespread. Especially with the rise of large-scale scientific and engineering computing applications, the complexity of scientific data has shown unprecedented and explosive growth. These scientific data not only have huge amounts of data, but also usually contain multi-variable, time-varying and high-resolution data. rate characteristics.
目前，对于多变量空间数据场可视化的研究还相对较少，而与之对应的需求却在不断加大，现有技术无法满足特定领域的高频率、大数据量的数据可视化。随着多变量空间数据场问题的应用领域和涵盖学科逐渐拓宽，提出了更多对多物理量数据计算、结果求解和可视化显示的要求。除了对多变量空间数据场可视化技术本身进行创新改进，研究适用于该呈现方式的交互环境非常关键。At present, there are relatively few studies on the visualization of multi-variable spatial data fields, but the corresponding needs are constantly increasing. Existing technologies cannot meet the high-frequency and large-volume data visualization in specific fields. As the application fields and covered disciplines of multi-variable spatial data field problems gradually broaden, more requirements for multi-physical quantity data calculation, result solution and visual display are put forward. In addition to innovative improvements in the multivariate spatial data field visualization technology itself, it is critical to study the interactive environment suitable for this presentation method.
当前，大多数可视化分析功能面向的是通用性软件，例如paraview等，缺少一个针对3D-IC的专用型可视化分析程序。而对于有些适用于有限体积法的可视化软件，在显示多达几十GB甚至上百GB的数据时，在保留一定精度的条件下，渲染速度也明显出现了瓶颈。Currently, most visual analysis functions are oriented to general-purpose software, such as Paraview, etc., and there is a lack of a dedicated visual analysis program for 3D-IC. For some visualization software that is suitable for the finite volume method, when displaying tens of GB or even hundreds of GB of data, there is an obvious bottleneck in rendering speed while retaining a certain degree of accuracy.
发明内容Contents of the invention
针对当前工程数据爆炸式增长，多变量空间数据场可视化速度显著下降的问题，本申请提出了基于先工艺下自适应分辨率渲染的可视化方法，根据图像区域的重要性动态调整分辨率，以达到细节显示和性能优化的目标。In response to the current explosive growth of engineering data and the significant decrease in the visualization speed of multi-variable spatial data fields, this application proposes a visualization method based on adaptive resolution rendering under advanced technology, which dynamically adjusts the resolution according to the importance of the image area to achieve Detailed display and performance optimization goals.
本申请技术方案如下：The technical solution for this application is as follows:
基于先进工艺下芯片动态调节分辨率的数据可视化方法，包括如下步骤：A data visualization method based on dynamic chip resolution adjustment under advanced technology includes the following steps:
步骤1：数据集准备与预处理：Step 1: Dataset preparation and preprocessing:
步骤2：根据芯片结构和功耗构建模型：Step 2: Build a model based on chip structure and power consumption:
步骤3：计算分辨率内像素点的坐标；Step 3: Calculate the coordinates of the pixels within the resolution;
步骤4：将温度值转化为颜色信息，同时生成纹理；Step 4: Convert the temperature value into color information and generate texture at the same time;
步骤5：根据已经构建好的模型，使用不同的像素插值算法计算不同区域内像素点的坐标；Step 5: Based on the built model, use different pixel interpolation algorithms to calculate the coordinates of pixels in different areas;
步骤6：加载并显示图像；Step 6: Load and display the image;
步骤7：等待下一次交互操作，基于交互操作动态调整分辨率。Step 7: Wait for the next interaction and dynamically adjust the resolution based on the interaction.
与现有技术相比，本申请具有如下优点和有益效果：Compared with the existing technology, this application has the following advantages and beneficial effects:
本申请在3D-IC热分析领域，从芯片封装的复杂结构出发，对多物理场耦合的海量数据可视化问题提出了相应的解决方案，使预知的结果更为准确，三维图形渲染时间更快。本申请能够提升大规模多变量数据的可视化速度和效率，帮助研究人员更好地分析、理解原始数据。In the field of 3D-IC thermal analysis, this application proposes corresponding solutions to the massive data visualization problem of multi-physics coupling based on the complex structure of chip packaging, making the predicted results more accurate and the three-dimensional graphics rendering time faster. This application can improve the speed and efficiency of visualizing large-scale multivariate data and help researchers better analyze and understand raw data.
附图说明Description of the drawings
图1为本发明方法流程图；Figure 1 is a flow chart of the method of the present invention;
图2为本发明构建的模型网格划分及像素点示意图；Figure 2 is a schematic diagram of the model mesh division and pixel points constructed by the present invention;
图3为本发明构建的模型坐标到像素坐标转换示意图；Figure 3 is a schematic diagram of the conversion from model coordinates to pixel coordinates constructed by the present invention;
图4为本发明uml活动图；Figure 4 is a UML activity diagram of the present invention;
图5为本发明实施例模型示意图；Figure 5 is a schematic diagram of the model of the embodiment of the present invention;
图6为本发明实施例模型网格划分示意图；Figure 6 is a schematic diagram of model meshing according to the embodiment of the present invention;
图7为本发明实施例模型网格分类以进行插值计算的示意图。Figure 7 is a schematic diagram of model grid classification for interpolation calculation according to an embodiment of the present invention.
具体实施方式Detailed ways
在3D-IC热分析领域，通常使用有限元分析法或有限体积分析法对芯片结构进行划分，划分后通过求解方程组来得到温度分布。In the field of 3D-IC thermal analysis, finite element analysis or finite volume analysis is usually used to divide the chip structure. After division, the temperature distribution is obtained by solving a system of equations.
传统的三维空间数据场是带有三维空间坐标的离散数据采样集，可分为规则或不规则的网格结构。均匀划分方式简单，但是在同等求解精度下，网格数量会明显多于非均匀划分，因此求解时间也会增加。相比之下，非均匀划分方式较为困难，但能够在不降低求解精度，充分考虑芯片结构、功耗的前提下，减少网格数量，进而减少求解时间。两类划分方式的原理是，在芯片结构、材料、功耗等变量发生突变的地方，温度范围变化也较大，需要加密网格增加求解精度，而在其他平稳的地方，温度变化平稳，可以适当降低网格密度。The traditional three-dimensional spatial data field is a discrete data sampling set with three-dimensional spatial coordinates, which can be divided into regular or irregular grid structures. The uniform division method is simple, but under the same solution accuracy, the number of grids will be significantly more than that of non-uniform division, so the solution time will also increase. In contrast, the non-uniform partitioning method is more difficult, but it can reduce the number of grids and thus reduce the solution time without reducing the solution accuracy and fully considering the chip structure and power consumption. The principle of the two classification methods is that in places where variables such as chip structure, materials, and power consumption suddenly change, the temperature range changes greatly, and the grid needs to be encrypted to increase the solution accuracy. In other stable places, the temperature changes are stable and can be Reduce the grid density appropriately.
根据上述原理，在可视化阶段，可以把算力集中在关键区域，例如结构变化大、功耗密度大，温度梯度变化明显的区域，而在其他变化平稳的区域，可以适当减少算力。Based on the above principles, during the visualization stage, computing power can be concentrated in key areas, such as areas with large structural changes, high power consumption density, and obvious temperature gradient changes. In other areas with smooth changes, computing power can be appropriately reduced.
本发明提出了基于先进工艺下芯片动态调节分辨率的数据可视化方法，适用于3D-IC的大规模多变量数据的可视化。The present invention proposes a data visualization method based on chip dynamic adjustment resolution under advanced technology, which is suitable for the visualization of large-scale multi-variable data of 3D-IC.
下面将结合具体实施例及其附图对本申请提供的技术方案作进一步说明。结合下面说明，本申请的优点和特征将更加清楚。The technical solutions provided by this application will be further described below with reference to specific embodiments and the accompanying drawings. The advantages and features of the present application will become clearer in conjunction with the following description.
如图1所示，本发明基于先进工艺下芯片动态调节分辨率的数据可视化方法的流程主要包括如下步骤：As shown in Figure 1, the process of the data visualization method for dynamically adjusting resolution of chips based on advanced technology in the present invention mainly includes the following steps:
步骤1：数据集准备与预处理：输入热分析结果，然后进行数据集准备与处理，将离散的点组织成结构块；Step 1: Data set preparation and preprocessing: input the thermal analysis results, then perform data set preparation and processing, and organize discrete points into structural blocks;
步骤2：根据芯片结构和功耗构建模型：输入芯片的结构文件和功耗文件，根据已有信息构建模型，将全时空域划分为重要性不同的多个分类，并赋予对应的优先级；Step 2: Build a model based on chip structure and power consumption: input the structure file and power consumption file of the chip, build a model based on existing information, divide the full-time and spatial domain into multiple categories of different importance, and assign corresponding priorities;
步骤3：计算分辨率内像素点的坐标；Step 3: Calculate the coordinates of the pixels within the resolution;
步骤4：将温度值转化为颜色信息，同时生成纹理；Step 4: Convert the temperature value into color information and generate texture at the same time;
步骤5：根据已经构建好的模型，使用不同的像素插值算法计算不同区域内像素点的坐标；Step 5: Based on the built model, use different pixel interpolation algorithms to calculate the coordinates of pixels in different areas;
步骤6：加载并显示图像；Step 6: Load and display the image;
步骤7：等待下一次交互操作，基于交互操作动态调整分辨率。Step 7: Wait for the next interaction and dynamically adjust the resolution based on the interaction.
步骤1：数据集准备与预处理。Step 1: Data set preparation and preprocessing.
热分析完成后所得到的原始数据，通常是三维的离散点，包含点的坐标和温度，但是丢失了点的其他属性。对此，需要将其重新组织为带有芯片结构、材料、和功耗等属性的有效信息。假设每个点包含三维空间坐标、温度值(如果是瞬态则为温度值序列)，材料属性，功耗密度。此时无法从一个点直观地找到与其有连接关系的其他点，即便是找到了也仅仅是几何意义上的连接，并不能区分物理结构。因此，本发明在这个阶段读入芯片的真实物理结构，通过芯片的结构信息把点重新组织为一个个相互关联的物理块，块由点组成，而点属于块。The original data obtained after thermal analysis is completed is usually a three-dimensional discrete point, containing the coordinates and temperature of the point, but other attributes of the point are lost. In this regard, it needs to be reorganized into effective information with attributes such as chip structure, materials, and power consumption. It is assumed that each point contains three-dimensional space coordinates, temperature value (temperature value sequence if it is transient), material properties, and power consumption density. At this time, it is impossible to intuitively find other points that are connected to it from one point. Even if it is found, it is only a geometric connection and cannot distinguish the physical structure. Therefore, the present invention reads the real physical structure of the chip at this stage, and reorganizes the points into interconnected physical blocks through the structural information of the chip. Blocks are composed of points, and points belong to blocks.
在这一阶段，输入的是热分析完成后所得的原始数据文件，输出的是相互关联的由点组成的物理块信息。At this stage, the input is the original data file obtained after the thermal analysis is completed, and the output is interrelated physical block information composed of points.
步骤2：根据芯片结构和功耗构建模型。Step 2: Build a model based on chip structure and power consumption.
需要扫描从步骤1获得的有效信息，根据网格的划分方式从中识别和分割出多个分类，然后给每个分类赋予相应的优先级属性。也就是，根据区域的重要性生成一个区域的优先队列。目前，判断区域的重要性由芯片的结构和网格划分方式得到，推而广之，可以不仅仅考虑芯片的结构，还可以综合考虑芯片的功耗密度和材料属性。It is necessary to scan the effective information obtained from step 1, identify and segment multiple categories according to the grid division method, and then assign corresponding priority attributes to each category. That is, a priority queue for a region is generated based on the importance of the region. At present, the importance of the judgment area is obtained by the structure and grid division of the chip. By extension, not only the structure of the chip can be considered, but also the power density and material properties of the chip can be comprehensively considered.
这里存在着一种特殊情况，即全局使用结构化网络进行均匀划分。在这种情况下，只会有一种分类，也就只会有一种优先级，动态调节分辨率算法退化到与未使用芯片结构、功耗一致。总体来说，均匀划分是非均匀划分的一种特例。There is a special case here, that is, the global structure is uniformly divided using a structured network. In this case, there will only be one classification, and therefore only one priority, and the dynamic adjustment resolution algorithm will degrade to be consistent with the unused chip structure and power consumption. Generally speaking, uniform partitioning is a special case of non-uniform partitioning.
步骤3：选择初始区域内的像素点，计算默认分辨率下像素点的坐标。Step 3: Select the pixels in the initial area and calculate the coordinates of the pixels at the default resolution.
在这一步，首先要确定需要的图像分辨率，即宽度和高度的像素数，通常图像中单位长度包含的像素数量是由软件内置的(例如，1280*960，800*600)。然后根据所需的分辨率，从左至右，从上到下，计算出每个像素点的坐标。In this step, you must first determine the required image resolution, that is, the number of pixels in width and height. Usually the number of pixels per unit length in the image is built-in by the software (for example, 1280*960, 800*600). Then the coordinates of each pixel are calculated from left to right and top to bottom according to the required resolution.
步骤4.生成纹理。Step 4. Generate texture.
在本发明中，纹理是一张2D的图片，包含着坐标信息和颜色信息。以热分析得到的结果温度为例，由于温度值并不包含颜色信息，为了产生直观的视觉效果，将温度值T＝(t1，t2，t3……)转化为RGB颜色值C(kr，kg，kb)，其中kr，kg，kb分别为红、绿、蓝对应的颜色值。In the present invention, texture is a 2D picture, containing coordinate information and color information. Taking the temperature result obtained by thermal analysis as an example, since the temperature value does not contain color information, in order to produce an intuitive visual effect, the temperature value T = (t1 , t2 , t3 ...) is converted into an RGB color value C ( kr , kg , kb ), where kr , kg , kb are the color values corresponding to red, green, and blue respectively.
首先，计算出包括时序和空间在内的所有温度值的最大值tmax和最小值tmin。假设现在有N种颜色ci，i∈[0，n)，那么可以把所有温度划分为N-1个区间，其中，区间值为：First, the maximum value tmax and the minimum value tmin of all temperature values including time series and space are calculated. Assume that there are N colors ci , i∈[0, n), then all temperatures can be divided into N-1 intervals, where the interval values are:
接下来，令tmin对应其中c0，tmax对应cN-1，那么就存在ti与某种设定好的颜色ci一一对应，即：Next, let tmin correspond to c0 and tmax correspond to cN-1 , then there is a one-to-one correspondence between ti and a certain set color ci , that is:
C(tmin+i*Δt)＝ci，i∈[0，n)，i∈ZC(tmin +i*Δt )=ci , i∈[0, n), i∈Z
对于不满足上述条件的点，采用线性插值的方法赋予其颜色属性。For points that do not meet the above conditions, linear interpolation is used to assign color attributes to them.
即对于温度t∈[ti，ti+1)，可以根据下列公式分别计算出kr，kg，kb：That is, for the temperature t∈[ti , ti+1 ), kr , kg , kb can be calculated respectively according to the following formulas:
综上所述，可以计算出任意温度t对应的颜色值C(kr，kg，kb)。To sum up, the color value C(kr , kg , kb ) corresponding to any temperature t can be calculated.
步骤5：根据步骤2构建的模型使用不同的图像算法计算步骤3中像素坐标的值。Step 5: Use different image algorithms to calculate the values of pixel coordinates in Step 3 based on the model built in Step 2.
步骤5需要综合前四个步骤获得的所有结果，根据区域的优先级使用不同的像素插值算法计算出像素点坐标的颜色值。Step 5 requires integrating all the results obtained in the first four steps and using different pixel interpolation algorithms according to the priority of the area to calculate the color value of the pixel coordinates.
一般情况下，像素坐标和纹理坐标很难一一对应，因此要使用像素插值或重采样算法来计算那些纹理坐标对应不到的像素点值。In general, it is difficult to have a one-to-one correspondence between pixel coordinates and texture coordinates, so pixel interpolation or resampling algorithms must be used to calculate pixel values that do not correspond to texture coordinates.
如图2所示，已知模型上的某点P(x，y，z)是需要显示在屏幕上的像素点，但它不属于通过热分析根据矩阵求解计算出来的已知坐标点，由于它在可视化空间的范围内，需要通过像素插值算法获得它的温度值和颜色值。即需要找到这个像素坐标对应的纹理坐标，然后从纹理坐标中获取实际的颜色值。As shown in Figure 2, a certain point P (x, y, z) on the known model is a pixel point that needs to be displayed on the screen, but it does not belong to the known coordinate point calculated based on matrix solution through thermal analysis. Because It is within the scope of the visualization space, and its temperature and color values need to be obtained through a pixel interpolation algorithm. That is, you need to find the texture coordinates corresponding to this pixel coordinate, and then obtain the actual color value from the texture coordinates.
图像插值算法有很多种类，如最近邻插值、双线性插值、双三次插值、Lanczos插值、双重线性插值等。最近邻插值是最简单的插值方法之一，它根据距离待插值位置最近的已知像素的值来确定新像素的值。双线性插值利用周围4个已知像素的值对待插值位置进行加权平均，从而计算出新像素的值。双三次插值在双线性插值的基础上进一步扩展，利用BiCubic基函数求出周围16个像素点的权重，所求像素点的值等于16个像素点的加权叠加。There are many types of image interpolation algorithms, such as nearest neighbor interpolation, bilinear interpolation, bicubic interpolation, Lanczos interpolation, bilinear interpolation, etc. Nearest neighbor interpolation is one of the simplest interpolation methods that determines the value of a new pixel based on the value of the known pixel closest to the location to be interpolated. Bilinear interpolation uses the values of the surrounding four known pixels to perform a weighted average of the position to be interpolated to calculate the value of the new pixel. Bicubic interpolation is further expanded on the basis of bilinear interpolation, and the BiCubic basis function is used to calculate the weight of the surrounding 16 pixels. The value of the pixel is equal to the weighted superposition of the 16 pixels.
上述三种方式都可以得出f(p′)的值，在这里f(p′)既可以分别指代颜色C的R、G、B分量，也可以指代温度T，甚至可以指代其他需要计算得到的变量。但是不同算法在速度和精度上各有优劣。最近邻插值虽然计算速度较快，但可能导致图像边缘出现锯齿；双线性插值考虑了上下左右方向的4个像素点差异，可以产生更为平滑的结果；而双三次插值通过应用三次样条函数可以得到精确地估计，但是耗费的时间最长。The value of f(p′) can be obtained in the above three methods. Here f(p′) can refer to the R, G, and B components of color C, the temperature T, or even other components. The variables that need to be calculated. However, different algorithms have their own advantages and disadvantages in speed and accuracy. Although nearest neighbor interpolation is faster to calculate, it may cause jagged edges on the image; bilinear interpolation takes into account the 4 pixel differences in the up, down, left and right directions and can produce smoother results; and bicubic interpolation applies cubic splines. The function can be estimated accurately, but takes the longest time.
对现代CPU来说，在单个像素点的计算上，这三种方式耗费的资源几乎没有区别。但是在3D-IC热分析领域，通常要对数千万个点进行多次插值计算，此时，合理分配计算机的资源就显得极为重要。For modern CPUs, there is almost no difference in the resources consumed by these three methods in the calculation of a single pixel. However, in the field of 3D-IC thermal analysis, it is usually necessary to perform multiple interpolation calculations on tens of millions of points. At this time, it is extremely important to reasonably allocate computer resources.
根据步骤2构建的模型，步骤3从划分的网格中和芯片结构中提取了数个或数十个类别，并对这些类别赋予了各自的优先级。对于较低的优先级，如划分粗糙、温度变化平稳的区域，采用速度较快、精度较低的插值方式；对于重点关心的核心区域，例如网格突变区域、芯片功耗密度大的区域，采用速度慢但是精度较高的方式。Based on the model built in step 2, step 3 extracts several or dozens of categories from the divided grid and chip structure, and assigns respective priorities to these categories. For lower priorities, such as areas with rough divisions and smooth temperature changes, use faster and lower-accuracy interpolation methods; for core areas of key concern, such as grid mutation areas and areas with high chip power consumption density, Use a slower but more accurate method.
步骤6：加载并显示图像。Step 6: Load and display the image.
经过步骤1-5，已经获得了分辨率内所有像素点的坐标及其值。此时，将要绘制的计算好的数据以坐标和索引的形式传给OpenGL的着色器，让OpenGL进行绘制。OpenGL经过图形渲染管线后，能够把图形显示在屏幕上，完成最终的可视化步骤。需要说明的是，OpenGL为一个跨编程语言、跨平台的编程图形程序接口，本实施例选用OpenGL绘制图像，仅仅是举例，而非限定。After steps 1-5, the coordinates and values of all pixels within the resolution have been obtained. At this time, the calculated data to be drawn is passed to the OpenGL shader in the form of coordinates and indexes, allowing OpenGL to draw. After OpenGL passes through the graphics rendering pipeline, the graphics can be displayed on the screen to complete the final visualization step. It should be noted that OpenGL is a cross-programming language and cross-platform programming graphics program interface. This embodiment uses OpenGL to draw images, which is only an example and not a limitation.
步骤7：等待用户的交互操作，基于交互操作动态调整分辨率Step 7: Wait for user interaction and dynamically adjust the resolution based on the interaction
本发明除了对多变量空间数据场可视化技术本身进行创新改进，同样研究了适用于该呈现方式的交互环境。In addition to innovating and improving the multi-variable spatial data field visualization technology itself, the present invention also studies the interactive environment suitable for this presentation method.
根据数据分析的需求，如果需要对数据进行更为细节的显示，将返回步骤3，重复上述步骤3-步骤6，从而实现海量数据多分辨率的逐级加载。当区域放大到最大时，即一个像素的面积小于最小网格的面积时，认为此时范围内的每一个区域都是被重点关注的对象，它们被赋予相同的权重，将采用精度较高的图像插值算法计算屏幕内全部像素点的值。According to the needs of data analysis, if you need to display the data in more detail, you will return to step 3 and repeat the above steps 3 to 6, thereby realizing the step-by-step loading of massive data at multiple resolutions. When the area is enlarged to the maximum, that is, when the area of a pixel is smaller than the area of the minimum grid, each area within the range is considered to be an object of focus, and they are given the same weight, and a higher-precision method will be used. The image interpolation algorithm calculates the values of all pixels on the screen.
图4是本发明的uml活动图，活动图总共可以分为四个部分，分别是数据预处理、模型建立、纹理生成、图像显示。Figure 4 is the UML activity diagram of the present invention. The activity diagram can be divided into four parts in total, which are data preprocessing, model establishment, texture generation, and image display.
数据预处理部分，首先输入经过热分析计算后的点坐标和温度值，然后将这些离散的点组合成知道周围结构的块。In the data preprocessing part, the point coordinates and temperature values calculated by thermal analysis are first input, and then these discrete points are combined into blocks that know the surrounding structure.
在模型建立部分，需要输入芯片的结构和功耗文件，将芯片划分出不同重要性的区域，然后对不同区域赋予不同/相同的优先等级，然后把这些带有优先级的区域和预处理部分得到的块进行绑定。In the model building part, you need to input the structure and power consumption files of the chip, divide the chip into areas of different importance, and then assign different/same priorities to different areas, and then combine these prioritized areas with the preprocessing part The resulting block is bound.
在纹理生成部分，首先使用默认的解决方案计算初始化区域的点坐标与之对应的温度值，然后将温度值转换为颜色值。根据纹理的的类型不同(几何体/切片)会生成不同样式的纹理。In the texture generation part, the default solution is first used to calculate the point coordinates of the initialization area and the corresponding temperature values, and then the temperature values are converted into color values. Different styles of textures will be generated depending on the type of texture (geometry/slice).
在图像显示部分，会根据区域的优先级选择不同的算法，根据这些算法来计算屏幕画布内需要获得的未知点的值，然后将已知点和未知点一起以图片的结构存储下来，传给OpenGL，即可实现三维图像的显示。In the image display part, different algorithms are selected according to the priority of the area. Based on these algorithms, the values of the unknown points that need to be obtained in the screen canvas are calculated, and then the known points and unknown points are stored together in the structure of the picture and passed to OpenGL can realize the display of three-dimensional images.
最后，等待用户的交互操作，根据操作进一步显示在初始化时未显示或者分辨率较低的点。Finally, wait for user interaction, and further display points that are not displayed during initialization or have low resolution according to the operation.
实施例Example
对于如图5所示的简单的模型，采用长方体网格的划分方式得到如图6所示的网络。图6为三维划分某一层的俯视图，图中存在着两种划分网络，即属于内层(BEOL)的加密网格与外层(PCB)的普通网格。For the simple model shown in Figure 5, the network shown in Figure 6 is obtained by dividing the cuboid mesh. Figure 6 is a top view of a three-dimensional division of a certain layer. There are two division networks in the figure, namely the dense mesh belonging to the inner layer (BEOL) and the ordinary mesh of the outer layer (PCB).
将从左上角开始，扫描本层的整个网络。根据扫描的结果，本层应该有如下三个分类：分类一，像素点内全是外层网格；分类二，像素点内全是内层网格；分类三，像素点处于内层网格和外层网格的交界处。Starting from the upper left corner, the entire network at this layer will be scanned. According to the scanning results, this layer should have the following three categories: Category 1, all pixels are in the outer grid; Category 2, all pixels are in the inner grid; Category 3, the pixels are in the inner grid The junction with the outer grid.
由于内层是热源，根据传热学原理，可以设置分类二、分类三有较高的优先级，而分类一有较低的优先级。又因为温度在交界面的部分梯度变化大，所以给分类三赋予最高的优先级，而分类二次之。Since the inner layer is the heat source, according to the principle of heat transfer, classification two and three can be set to have a higher priority, while classification one has a lower priority. And because the temperature gradient changes greatly at the interface, classification three is given the highest priority, and classification two is given.
综上，已经根据网格划分方式提取出了三个分类。In summary, three classifications have been extracted based on the meshing method.
那么，如果待求取像素点的坐标位于分类一所指示的区域，如普通网格区域，如图7中的方块1，将使用最近邻插值计算其值；Then, if the coordinates of the pixel to be obtained are located in the area indicated by category one, such as a normal grid area, such as square 1 in Figure 7, nearest neighbor interpolation will be used to calculate its value;
以此类推，如果待求取像素点的坐标位于分类二所指示的区域，即网格加密区域，如图7中的方块2，将使用双线性插值计算其值；By analogy, if the coordinates of the pixel to be obtained are located in the area indicated by category 2, that is, the grid refinement area, such as square 2 in Figure 7, bilinear interpolation will be used to calculate its value;
同样地，如果待求取像素点的坐标位于分类三所指示的区域，即网格突变区域，如图7中的方块3，将使用双三次插值计算其值。Similarly, if the coordinates of the pixel to be obtained are located in the area indicated by category three, that is, the grid mutation area, such as square 3 in Figure 7, bicubic interpolation will be used to calculate its value.
最后，将得到的值以坐标和索引的格式传递给顶点着色器，即可显示三维图像。Finally, the resulting values are passed to the vertex shader in the format of coordinates and indices to display the three-dimensional image.
上述描述仅是对本申请较佳实施例的描述，并非是对本申请范围的任何限定。任何熟悉该领域的普通技术人员根据上述揭示的技术内容做出的任何变更或修饰均应当视为等同的有效实施例，均属于本申请技术方案保护的范围。The above description is only a description of the preferred embodiments of the present application, and does not limit the scope of the present application in any way. Any changes or modifications made by a person of ordinary skill in the field based on the technical content disclosed above shall be regarded as equivalent and effective embodiments, and shall fall within the scope of protection of the technical solution of this application.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311308967.1ACN117313646B (en) | 2023-10-10 | 2023-10-10 | Data visualization method for dynamically adjusting resolution of chip based on advanced technology |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311308967.1ACN117313646B (en) | 2023-10-10 | 2023-10-10 | Data visualization method for dynamically adjusting resolution of chip based on advanced technology |
| Publication Number | Publication Date |
|---|---|
| CN117313646Atrue CN117313646A (en) | 2023-12-29 |
| CN117313646B CN117313646B (en) | 2025-05-13 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311308967.1AActiveCN117313646B (en) | 2023-10-10 | 2023-10-10 | Data visualization method for dynamically adjusting resolution of chip based on advanced technology |
| Country | Link |
|---|---|
| CN (1) | CN117313646B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030142875A1 (en)* | 1999-02-04 | 2003-07-31 | Goertzen Kenbe D. | Quality priority |
| CN105611256A (en)* | 2015-12-28 | 2016-05-25 | 南京航空航天大学 | Bayer format image interpolation method based on gradients and display device based on FPGA |
| CN113420810A (en)* | 2021-06-22 | 2021-09-21 | 中国民航大学 | Cable trench intelligent inspection system and method based on infrared and visible light |
| CN116781996A (en)* | 2023-05-06 | 2023-09-19 | 重庆邮电大学 | Infrared tracking and aiming system based on uncooled infrared detector |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030142875A1 (en)* | 1999-02-04 | 2003-07-31 | Goertzen Kenbe D. | Quality priority |
| CN105611256A (en)* | 2015-12-28 | 2016-05-25 | 南京航空航天大学 | Bayer format image interpolation method based on gradients and display device based on FPGA |
| CN113420810A (en)* | 2021-06-22 | 2021-09-21 | 中国民航大学 | Cable trench intelligent inspection system and method based on infrared and visible light |
| CN116781996A (en)* | 2023-05-06 | 2023-09-19 | 重庆邮电大学 | Infrared tracking and aiming system based on uncooled infrared detector |
| Title |
|---|
| CHIA-LIN KO ET AL.: "Multi-resolution Volume Rendering of Large Time-Varying Data using Video-based Compression", 2008 IEEE PACIFIC VISUALIZATION SYMPOSIUM, 21 March 2008 (2008-03-21)* |
| 姜定，叶茫: "面向跨模态文本到图像行人重识别的 Transformer网络", 中国图象图形学报, 31 May 2023 (2023-05-31)* |
| Publication number | Publication date |
|---|---|
| CN117313646B (en) | 2025-05-13 |
| Publication | Publication Date | Title |
|---|---|---|
| CN105608222A (en) | Rapid building method of tile pyramid for large-scale raster data set | |
| CN105701759B (en) | Map vector line feature turning reasonability method for drafting based on GPU | |
| CN105844711A (en) | Engraving a 2D image on a subdivision surface | |
| CN102937896B (en) | The method of color mapping techniques Dynamic Display two-dimensional space data is utilized in SVG | |
| CN104750937B (en) | Integrated circuit diagram conversion method based on Hadoop | |
| CN104537125B (en) | A kind of remote sensing image pyramid parallel constructing method based on message passing interface | |
| CN110214338A (en) | Application of the increment color compressed to video | |
| CN109544668A (en) | Texture coordinate processing method, terminal device and computer readable storage medium | |
| CN105741335B (en) | A kind of multi-level war game map fast drawing method based on blocking organization | |
| CN113094536B (en) | Power transmission tower lightweight display method and device based on power grid GIM | |
| CN117313646A (en) | Data visualization method for dynamically adjusting resolution of chip based on advanced technology | |
| TWI826696B (en) | Methods and apparatus for efficient interpolation | |
| Tanaka et al. | Treecube: Visualization tool for browsing 3d multimedia data | |
| Wu et al. | MLSEB: Edge bundling using moving least squares approximation | |
| Jiang et al. | Research on time-series data visualization method based on parameterized parallel coordinates and color mapping function | |
| Joy | Massive data visualization: A survey | |
| US20230038647A1 (en) | Anti-aliasing two-dimensional vector graphics using a compressed vertex buffer | |
| Zhao et al. | Fast variational design of multiresolution curves and surfaces with B-spline wavelets | |
| Doraiswamy et al. | GPU-powered spatial database engine for commodity hardware: Extended version | |
| Yin et al. | Multi-screen Tiled Displayed, Parallel Rendering System for a Large Terrain Dataset. | |
| JP2021101399A (en) | Device for editing three-dimensional shape data and program for editing three-dimensional shape data | |
| CN109410296B (en) | Method and device for realizing data display | |
| Yang et al. | Spline R-function and applications in FEM | |
| Bruijns | Quadratic Bezier triangles as drawing primitives | |
| Zhang et al. | Interactive rendering for large-scale mesh based on MapReduce |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |