MU5IN160 – Parallel Programming

Azure OpenAI异步编程实战:高并发任务并行化与限流优化 异步编程是提升云AI服务吞吐能力的基础技术,其核心在于利用协程高效处理I/O密集型网络请求,避免线程阻塞与资源浪费。在Azure OpenAI等托管大模型服务中,网络往返延迟、配额限流(RPM/TPM)和连接池管理构成关键约束,需结合asyncio、动态限流与有序结果聚合实现稳定高性能。该技术显著降低平均响应延迟、提升token级资源利用率,并支撑批量摘要、多路意图识别、RAG流水线等典型企业AI场景。本文聚焦Azure OpenAI环境下的Python异步落地实践,涵盖AsyncAzureOpenAI初始 阅读详情


Sorbonne Université – SESI M2
——–
MU5IN160 – Parallel Programming
Hands-on Session 6 – Dataflow for Motion Application
Very important, about the submission of your work At the end of this session you will have to
upload the following files on Moodle: 1) a zip of the src folder and 2) a zip of the include folder. After
that you will have 2 weeks to complete your work and to update your first submission. You have to work
in group of two people but each of you will have to upload the file on Moodle. Finally, please write your
name plus the name of your pair at the top of all these files.
Short introduction In this session, we will work on a streaming application that detects and tracks
moving objects from a video sequence. Contrary to the previous sessions, we will not use EasyPAP this
time. The later is not adapted for streaming applications. A working streaming application will be given
to you and you will have to use StreamPU to implement the Motion application through an explicit
dataflow representation.
1 Appetizer
First you need to clone the repository of the Motion project:
git clone --recursive https://gitlab.lip6.fr/parallel-programming/motion-sesi.git
The Motion project uses CMake in order to generate a Makefile: follow the README instructions to
compile the code.
grayscale image
(t − 1) motion detection Σ∆ (per pixel) mathematical morphology
opening-closing
connected components
labeling (CCL)
connected components
analysis (CCA)
surface filtering
grayscale image
(t) motion detection Σ∆ (per pixel) mathematical morphology
opening-closing
connected components
labeling (CCL)
connected components
analysis (CCA)
surface filtering
k-nearest neighboor
matching (k-NN)
temporal
tracking
grayscale pixels
p ∈ [0; 255]
blob of binary pixels
p ∈ {0, 1}, 0 → stationary, 1 → moving
image of labels
l ∈ [1; 232 − 1]
CCs = list of regions,
surface S & centroid (xG, yG)
sub-list of CCs with S ∈ [Smin, Smax]
list of (t − 1, t)
associations
final list of
moving objects
Figure 1: Motion detection and tracking processing graph. In gray and italic: the output of each
processing.
Fig. 1 presents the different algorithms used to detect moving objects and to track them over time. To
make it work, two strong assumptions are made: 1) the camera is fixed, 2) the light intensity is constant
over time. First, an image is read from a camera (or a video sequence) and then it is converted in a
grayscale image. Then, the Σ∆ algorithm is triggered. This algorithm is able to detect if a pixel is
moving over time. It returns a binary image, if a pixel value is 0, then it means that it is not moving.
Otherwise, if a pixel value is 1, then it means that it is moving. After that, morphology algorithms are
applied1
. This is a pre-processing to regroup moving pixels together and eliminate isolated pixels. Then,
from a binary image, a connected components labeling (CCL) algorithm is performed. The later, gives
the same label to a group of pixel that are connected to each other. CCL returns an image of labels where
l = 0 means no object and l > 0 means a moving object. From this image of label, some features are
extracted (CCA): for each object the center of mass (xG, yG), the bounding box ([xmin, xmax, ymin, ymax])
and the surface S are extracted. Depending on their surface, the objects are filtered (Smin < S < Smax).
From two images at t − 1 and t, a matching algorithm determines which objects are the same in the two
different images (mainly according to their distance). At the end, the identified objects are tracked to
have a constant identifier over time.
This graph of tasks is then repeated until the video sequence is over. It is not mandatory to understand
perfectly each algorithm. The purpose of this session is to work on a streaming application, representative
of a real application, and to perform optimizations at the task graph level.
1Mathematical morphology: https://en.wikipedia.org/wiki/Mathematical_morphology
In this graph, two tasks cannot be replicated. The per pixel motion algorithm requires its previous
output to compute the current binary image. It detects intensity variations over time. It is almost the
same for the tracking algorithm that maintains a list of tracks that are updated according to the last
frame.
If you’d like to better understand the algorithms used in this project, some of them are described in more
detail in the document’s appendix. In any case, it’s worth noting that you don’t need to understand
exactly what these algorithms do to complete this lab.
1.1 Run Motion
To run the code you will need some input videos. You can download a videos collection on Moodle (see the
“Artifacts” section) or from this web link: http://www.potionmagic.eu/~adrien/data/traffic.zip.
First, unzip the traffic.zip and from the build directory run the code with the following command:
./bin/motion2 --vid-in-path ./traffic/1080p_day_street_top_view_snow.mp4 \
--flt-s-min 2000 --knn-d 50 --trk-obj-min 5 --vid-out-play --vid-out-id
You should see a window with a top view of a highway and some moving cars (see Fig. 2) and you should
see green bounding boxes around the cars.
Figure 2: Motion screenshot (with –-vid-out-play –-vid-out-id parameters).
1.2 Architecture of the Project
Motion is mainly a C-style project but it is compiled in C++ to use StreamPU. The sources are
located in the src folder, and there are 3 sub-folders:
• common: contains implementations of the processing tasks,
• main: contains source files that correspond to a final binary executable,
• wrapper: contains C++ files to wrap the C-style processing functions into StreamPU modules
and tasks.
The headers are located in the include folder. Inside there are two sub-folders: c/motion for the C-style
headers and cpp/motion for the C++ headers.
Page 2
2 From Imperative to Dataflow Programming
We will convert the motion2 main into a dataflow description (= StreamPU modules and tasks). The
motion2 is located here src/main/motion2.c. This implementation is very close to the task graph
presented in Fig. 1.
Task #1 Understand the code, run the motion2 executable and play with the parameters (-h shows
and describes the available parameters).
To help you in the task, we created an other main based on motion2.c and we converted some C functions
into StreamPU modules for you. See the motion2_spu.cpp file.
Task #2 Understand the code, run the motion2-spu executable and play with the parameters (-h
shows and describes the available parameters). Understand the code of motion2_spu.cpp by comparing
it with the C-style motion2.c code.
Task #3 Create new StreamPU stateful modules, each time you will create new .cpp and .hpp
files in the wrapper folders. You will only declare input and output sockets (DON’T use forward
sockets at this time):
1. Sigma_delta: Add a StreamPU compute task that will call the sigma_delta_compute function,
2. Morpho: Add a StreamPU compute task that will call the C morpho_compute_opening3 and
morpho_compute_closing3 functions,
3. CCL: Add a StreamPU apply task that will call the C CCL_LSL_apply function,
4. Features_CCA: Add a StreamPU extract task that will call the C features_extract function,
5. Features_filter: Add a StreamPU filter task that will call the C features_filter_surface
and features_shrink_basic functions (note that the maximum input size of the features differs
from the maximum size of the output features: indeed, the main purpose of the shrink function is
to reduce the maximum number of features and to save memory space),
6. KNN: Add a StreamPU match task that will call the C kNN_match function,
7. Tracking: Add a StreamPU perform task that will call the C tracking_perform function.
Add the StreamPU modules and tasks incrementally in the motion2_spu.cpp file and you will test
if their integration is working (you can compare the logs with a diff, see Note #2 below). Have a look
on how we did this for the other StreamPU tasks that are given to you. You will follow the same
philosophy: 1) bind the sockets to the buffers allocated in the main file and 2) call the exec() method
explicitly.
Note #1 It is NOT possible to create sockets of RoI_t structure. Only the basic C types are supported.
To get around this limitation you can count the number of bytes in the structure. For instance, you can
do something like:
auto si_RoIs = this->template create_socket_in<uint8_t>(t, "in_RoIs", max_size * sizeof(RoI_t));
Note #2 motion2 is our golden model. To compare the results of motion2 and motion2-spu you need
to generate the logs of motion2 executable first (we do it for only 20 frames to execute faster):
./bin/motion2 --vid-in-path ./traffic/1080p_day_street_top_view_snow.mp4 \
--vid-in-stop 20 --flt-s-min 2000 --knn-d 50 --trk-obj-min 5 --log-path logs_refs
Secondly, you need to generate the logs of the motion2-spu executable:
Page 3
./bin/motion2-spu --vid-in-path ./traffic/1080p_day_street_top_view_snow.mp4 \
--vid-in-stop 20 --flt-s-min 2000 --knn-d 50 --trk-obj-min 5 --log-path logs_spu
Finally you need to compare the logs together:
diff logs_refs logs_spu
If the later command returns nothing, it means that motion2 and motion2-spu are equivalent (in term
of features). This is good, your new implementation is correct! If not... it is time to debug :’-(.
Task #4 At this point, you should only have StreamPU tasks that call their exec() method explicitly
(no more C style function calls). However, the code is still using the data allocated in the main function.
This can be improved because StreamPU performs the data allocation and deallocation automatically
for you. In order to remove most of these allocations you have to perform partial “output to input socket”
bindings. For instance, if we only consider to eliminate the IB0 buffer, it is possible to remove “pointer
to output socket” bindings and to add “output to input socket” bindings instead, as shown in Code 1.
Do it for all the buffers, EXCEPT for IG0 and IG1. It is strongly advised to do it step by step and to
check if the code is giving exactly the same results after each modification (please refer to Note #2).
// [...]
// step 1: motion detection (per pixel) with Sigma-Delta algorithm
sd0["compute::in_img"].bind(IG0[0]);
// sd0["compute::out_img"].bind(IB0[0]); // this line can be removed
sd0("compute").exec();
// step 2: mathematical morphology
// mrp0["compute::in_img"].bind(IB0[0]); // this line can be removed
mrp0["compute::in_img"] = sd0["compute::out_img"]; // <-- [NEW] output to input socket binding
// mrp0["compute::out_img"].bind(IB0[0]); // this line can be removed
mrp0("compute").exec();
// step 3: connected components labeling (CCL)
uint32_t n_RoIs_tmp0;
// ccl0["apply::in_img"].bind(IB0[0]); // this line can be removed
ccl0["apply::in_img"] = mrp0["compute::out_img"]; // <-- [NEW] output to input socket binding
ccl0["apply::out_labels"].bind(L10[0]);
ccl0["apply::out_n_RoIs"].bind(&n_RoIs_tmp0);
ccl0("apply").exec();
// [...]
Source code 1: Example of partial socket binding to eliminate IB0 buffer allocation/deallocation in the
main function.
Task #5 Now, replace IG0 and IG1 buffers by the binding of the video["generate::out_img_gray8"]
socket. For this, you will need to use a Delayer module in order to keep the t − 1 image in memory
(previously kept in the IG0 buffer). If you don’t use it, the t − 1 image will always be overwritten when
executing the video("generate") task.
Note #3 In the motion2 executable, some tasks are not executed in the first stream (see the following
condition in the motion2.c file: “if (n_processed_frames > 0)”). To manage it you have two possible
options:
• Always execute the tasks (no control flow) but in this case you need to carefully initialize the
Delayer module to the first frame with the Delayer::set_data() method (this solution is
simpler to implement),
Page 4
• Use a Switcher and a Controller_limit module to implement the control flow (= if condition).
To simplify, you will only put the Sigma_delta.compute() task in the condition. In other terms,
the CCL, the CCA and the filtering will be executed anyway.
Task #6 At this point you should not have memory allocations and deallocations anymore in the main
function. Next objective is to get rid of the multiple exec() calls over the tasks. You will separate the
binding from the execution. To do this, the socket bindings need to be moved outside of the while(1)
loop and the while(1) loop needs to be replaced by a StreamPU Sequence. Once it is done, only one
exec() call should remain: the one over the newly created Sequence object. Of course, you will check
if it works correctly (please refer to Note #2).
Sub-sequence 0
Delayer
exec order: [13]
addr: 0x16b1660a8
memorize (id = 13)
Tracking
exec order: [14]
addr: 0x16b166468
perform (id = 14)
Sigma_delta
Sigma_delta1
exec order: [7]
addr: 0x16b166cc8
compute (id = 7)
Morpho
Morpho1
exec order: [8]
addr: 0x16b166b38
compute (id = 8)
CCL
CCL1
exec order: [9]
addr: 0x16b1669a8
apply (id = 9)
Features_CCA
CCA1
exec order: [10]
addr: 0x16b166808
extract (id = 10)
Features_filter
Ftr_filter1
exec order: [11]
addr: 0x16b166618
filter (id = 11)
KNN
exec order: [12]
addr: 0x16b166548
match (id = 12)
Delayer
exec order: [0]
addr: 0x16b1660a8
produce (id = 0)
Sigma_delta
Sigma_delta0
exec order: [1]
addr: 0x16b166d98
compute (id = 1)
Morpho
Morpho0
exec order: [2]
addr: 0x16b166c00
compute (id = 2)
CCL
CCL0
exec order: [3]
addr: 0x16b166a70
apply (id = 3)
Features_CCA
CCA0
exec order: [4]
addr: 0x16b1668d8
extract (id = 4)
Features_filter
Ftr_filter0
exec order: [5]
addr: 0x16b166710
filter (id = 5)
Video
exec order: [6]
addr: 0x16b166e98
generate (id = 6)
out[0]:out
in[0]:in_img
out[1]:status
out[1]:out_img
in[0]:in_img
out[2]:status
out[1]:out_img
in[0]:in_img
out[2]:status
out[1]:out_labels
in[0]:in_labels
0
in[0]:in_labels
1
out[2]:out_n_RoIs
in[1]:in_n_RoIs
0
in[1]:in_n_RoIs
1
out[3]:status
out[2]:out_RoIs
in[2]:in_RoIs
out[3]:status
out[3]:out_labels out[4]:out_n_RoIs
in[1]:in_n_RoIs0
out[5]:out_RoIs
in[0]:in_RoIs0
out[6]:status
out[0]:out_img
in[0]:in_img
0
in[0]:in
1
out[1]:out_frame
in[0]:in_frame
out[2]:status
out[1]:out_img
in[0]:in_img
out[2]:status
out[1]:out_img
in[0]:in_img
out[2]:status
out[1]:out_labels
in[0]:in_labels
0
in[0]:in_labels
1
out[2]:out_n_RoIs
in[1]:in_n_RoIs
0
in[1]:in_n_RoIs
1
out[3]:status
out[2]:out_RoIs
in[2]:in_RoIs
out[3]:status
out[3]:out_labels out[4]:out_n_RoIs
in[3]:in_n_RoIs1
0
in[2]:in_n_RoIs
1
out[5]:out_RoIs
in[2]:in_RoIs1
out[6]:status
out[4]:out_RoIs0 out[5]:out_RoIs1
in[1]:in_RoIs
out[6]:out_nearest out[7]:out_distances out[8]:status
out[1]:status
out[3]:status
Figure 3: Expected StreamPU task graph without logs, without visualization and without control flow.
Note #4 To help you in the debugging, you can print the sequence graph with the export_dot method.
Enable/disable the logs, enable/disable the visualization and observe the impact on the task graph. If
you chose to do not implement control flow, the output graph should looks like in Fig. 3. Note that you
can personalize the name of a module with the set_custom_name(std::string custom_name) method.
Task #7 Before the sequence execution, you will enable the statistics of the task (call the get_modules
method on a sequence object). And after the sequence execution you will print them at the end
(tools::Stats::show function). The application will display the statistics only if there is the --stats
parameter. What do you see? Is it different than from the motion2 executable? Explain.
[Bonus] Task #8 When you think it’s necessary, create new tasks, postfixed with a f, that use forward
socket instead of input/output sockets combination. For instance, if we consider a task named compute
without forward socket, the task that uses forward socket will be named computef. You will NOT
replace the former compute task. Using forward sockets should help you to remove useless copies.
Do it incrementally to validate that the application is still working (see Note #2). Can you see an
improvement in the statistics of the tasks?
Page 5
Appendix
2.1 Sigma-Delta Algorithm (Σ∆)
The motion detection problem consists in separating moving and static areas in each frame. At each
instant, each pixel must be tagged with a fixed/moving binary identifier. When the camera is fixed, such
detection can be performed using the time differences computed for each pixel.
The following notations apply:
• t : current instant of time, used to identify the frames,
• It: grayscale source image at time t,
• It−1: grayscale source image at time t − 1,
• Mt: background image (mean image),
• Ot: grayscale difference image,
• Vt: image of variance (standard deviation) computed for each pixel,
• Lt: binary label image (motion/background), Lt(x) = {0, 1} or Lt(x) = {0, 255} to encode
{background, movement},
• x: the current pixel with (i, j) coordinates.
Most of motion detection techniques in an image sequence It(x) are based on an estimate of the modulus
of the temporal gradient |
∂I
∂t |. If the light intensity of the scene vary slowly (= is constant between two
consecutive images), then a significant variation in the pixel grayscale (above a threshold) between two
images will imply that there is movement at that point.
The Σ∆ algorithm assumes that the noise level can vary at any point. To achieve this, the pixel grayscale
is modeled by a mean Mt(x) and a variance (standard deviation) Vt(x). If the difference between the
current image and the background image is greater than N times the standard deviation, then movement
occurs. The value of N is a parameter. In this project, N is always set to 2.
This is a motion detection system based on the estimation of static background statistics using Σ∆
modulation: an iterative analog/digital conversion method that increments or decrements the digitized
value by one unit according to the result of the comparison between the analog value and the current
digitized value.
Algorithm 1: Sigma-Delta (Σ∆).
1 [Part #1: mean computation]
2 foreach pixel x do // Step #1: Mt estimation
3 if Mt−1(x) < It(x) then Mt(x) ← Mt−1(x) + 1
4 if Mt−1(x) > It(x) then Mt(x) ← Mt−1(x) − 1
5 otherwise do Mt(x) ← Mt−1(x)
6 [Part #2: difference computation]
7 foreach pixel x do // Step #2: Ot computation
8 Ot(x) = |Mt(x) − It(x)|
9 foreach pixel x do // Step #3: Vt update and clamping
10 if Vt−1(x) < N × Ot(x) then Vt(x) ← Vt−1(x) + 1
11 if Vt−1(x) > N × Ot(x) then Vt(x) ← Vt−1(x) − 1
12 otherwise do Vt(x) ← Vt−1(x)
13 Vt(x) ← max(min(Vt(x), Vmax), Vmin)
14 foreach pixel x do // Step #4: Lˆt estimation
15 if Ot(x) < Vt(x) then Lˆt(x) ← 0
16 else Lˆt(x) ← 1
The algorithm initialization for t = 0 is the following: M0(x) ← I0(x) and V0(x) ← Vmin. Then, the
algorithm is applied to the images from t = 1. The Vmin and Vmax constants are used to restrict the
possible values of Vt. Typically, Vmin = 1 and Vmax = 254. The complete algorithm after initialization is
shown in Alg. 1.
In the Motion project, a naive Σ∆ implementation is given to you:
Page 6
• Header: in the include/c/motion/sigma_delta/sigma_delta_compute.h file,
• Source: in the src/common/sigma_delta/sigma_delta_compute.c file.
See the sigma_delta_compute function.
2.2 Mathematical Morphology
In this project, we consider squared elements B of size 3 × 3. Let X be the set of pixels associated with
the B element. There are two basic operations: the dilation of X noted δB(X) and the erosion of X
noted ϵB(X). The application of mathematical morphology operators is similar to filtering operators
(stencils or convolutions), but with non-linear operations.
For binary images, dilation consists in computing a OR on the B neighborhood in the source image and
writing it to the destination image. Conversely, erosion consists in computing a AND on the neighborhood.
So, if a point in the neighborhood is 1, the dilation produces a 1 (since x OR 1 == 1), thus dilating the
binary connected component. Conversely, if only one pixel is 0 in the B neighborhood, the erosion will
produce a 0 (since x AND 0 == 0), thus eroding the connected component.
Erosion is used to reduce noise in images: if we consider that a small group of pixels is the noise that
we’re trying to remove, then applying erosion with a B element of size 3 × 3 will make any group of
pixels with a radius smaller than its size disappear.
Figure 4: Left: the initial binary image. Center: eroded image with a 3 × 3 squared element: the gray pixels are
removed. Right: dilated image with a 3 × 3 squared element: the gray pixels are added. Source: Wikipedia.
Let r be the radius and d = 2r + 1 the diameter of a squared element B, then an erosion of radius r
removes, to any connected component, a thickness of r pixels of contour while a dilation of radius r adds
a thickness of r pixels to the contour (see Fig. 4, note that in the figure the logic is reversed: pixels at 1
are black while pixels at 0 are white).
Figure 5: Left: the initial binary image. Center: opened image with a 3 × 3 squared element: the gray pixels
are removed. Right: closed image with a 3 × 3 squared element: the gray pixels are added. Source: Wikipedia.
From these two operators, two others can be defined: the closing ϕB(X) = ϵB(δB(X)) and the opening
γB(X) = δB(ϵB(X)). Closing reduces (or even completely close) holes in connected components, while
opening does the opposite, enlarging these same holes (see Fig. 5, note that in the figure the logic is
reversed: pixels at 1 are black, while pixels at 0 are white).
One of the advantages of opening and closing is that they preserve the (discrete) size of the regions,
unlike erosion, which reduces it, or dilation, which increases it. Depending on requirements, either a
closing or an opening can be chosen. As these operators are idempotent, applying them several times
does not change the result (which will be identical to that obtained after a single application). On the
other hand, they can be chained (opening and then closing or closing and then opening) to improve the
result image (noise reduction, filling holes, ...). By gradually increasing their radius, we obtain sequential
alternating filters, which are particularly effective for removing noise.
In the Motion project, naive 3 × 3 mathematical morphology implementations are given to you:
• Header: in the include/c/motion/morpho/morpho_compute.h file,
Page 7
• Source: in the src/common/morpho/morpho_compute.c file.
See the morpho_compute_opening3 and morpho_compute_closing3 functions.
 

基于JTAG的MC56F8013批量编程方案:硬件并联与软件校验实践 在嵌入式系统开发与量产中,Flash编程是确保微控制器(MCU)或数字信号控制器(DSC)正确运行的基础环节。其原理是通过调试接口(如JTAG)将编译后的固件代码写入芯片的非易失性存储器。这项技术的核心价值在于实现产品功能的固化与部署,是连接软件开发和硬件产品的关键桥梁。在工业控制、消费电子、汽车电子等应用场景中,高效、可靠的批量编程方案能极大提升生产效率和产品一致性。本文聚焦于一种针对特定DSC的优化方案,通过**JTAG接口**的并联拓扑实现高速“一对多”编程,并深入探讨了应对“盲编程”挑战的**固件完 阅读详情

相关推荐

GPT-4参数量与2%稀疏性真相:MoE架构下的计算密度与内存墙突破

大语言模型的参数量并非单纯规模指标,而是与硬件带宽、缓存效率和调度机制深度耦合的系统工程概念。MoE(Mixture of Experts)架构通过专家路由实现动态稀疏激活,使万亿级模型在单次前向传播中仅需调动约2%的参数子集,本质是计算稀疏性而非存储稀疏性。这一特性显著缓解GPU显存带宽压力,提升L2缓存命中率,降低NVLink通信开销,从而将性能瓶颈从‘内存墙’重新导向计算单元。技术价值在于支撑长上下文(如128K)、低延迟推理与边缘可部署性;典型应用场景包括大模型推理优化、MoE架构选型、vLLM/P

corg81763的博客 472

Cheatsheet: 2013 12.17 ~ 12.31

.NET Introducing ASP.NET Web API Throttling handler C# async and await: A Deeper Dive PARALLEL PROGRAMMING IN C# AND OTHER ALTERNATIVES Windows service auto-update plugin framework ...

weixin_30439131的博客 183

CPLEX 20.1.0 Windows安装与调用避坑指南

CPLEX是运筹学领域最主流的商用数学规划求解器之一,其核心能力在于高效求解线性规划(LP)、混合整数规划(MIP)等确定性优化问题。原理上依赖底层C++运行时、Java虚拟机及严格校验的许可证机制,技术价值体现在求解稳定性、数值精度与跨平台可复现性。典型应用场景涵盖高校课程设计、企业排产调度、物流路径优化等对决策确定性要求极高的工程实践。然而在Windows环境下,CPLEX 20.1.0因强制绑定Visual Studio 2019运行时、敏感的环境变量配置及学术许可验证机制,极易出现‘CPLEX li

weixin_34248023的博客 408

可伸缩系统的设计模式(译)

Ricky Ho在他的博客中分享了该文章,该文章是一个简单的概括分享,详细的可以参见他博客的其它详细文章。下面主要是意译。1、Load Balancer:负载均衡 – 由分发者来决定哪个工作者处理下一个请求,这种决定可以基于不同的策略。“In this model, there is a dispatcher that determines which worker instance will h

u012730075的专栏 882

【信息科学与工程学】【数据科学】第十八篇 数值分析 系列二 数值算法设计01

编号类型领域问题问题的数学分析及数值分析算法逐步推理思考的数学表达式参数列表及参数的边界范围及条件及常数关联知识1数值积分数值逼近计算定积分 I=∫ab​f(x)dx的近似值。数学分析:将积分区间等分,利用插值多项式近似被积函数,构造机械求积公式。数值分析:Newton-Cotes公式是等距节点的插值型求积公式。高阶公式(n≥8)因Cotes系数有正负,数值不稳定。偶阶公式(如Simpson)具有n+1次代数精度。为提高精度,常采用复化求积法(将区间细分后应用低阶公式)。

weixin_49199313的博客 161

【信息科学与工程学】【数据中心】第三十五篇 云计算数据中心的学科知识04

编号学科(课程)核心知识点在云计算/云存储/云网络/云安全/云MaaS中的作用代表教材/资料/论文 + 数学方程式列表工业界应用D1421​云原生数据库:TiDB​HTAP(混合事务/分析处理)、分布式SQL、水平扩展、强一致(Raft)、自动故障恢复、与MySQL兼容、TiFlash列式引擎提供弹性扩展的分布式数据库;支撑高并发在线交易与实时分析教材:《TiDB in Action》PingCAP(2020) 论文:《TiDB: A Raft-based HTAP Database》(2017)

weixin_49199313的博客 418

Apache Spark 2.2.0 中文文档 - Spark RDD(Resilient Distributed Datasets)

Spark RDD(Resilient Distributed Datasets)论文 概要 1: 介绍 2: Resilient Distributed Datasets(RDDs) 2.1 RDD 抽象 2.2 Spark 编程接口 2.2.1 例子 – 监控日志数据挖掘 2.3 RDD 模型的优势 2.4 不适合用 RDDs 的应用 3 Spark 编程接口 3.1 ...

weixin_30670151的博客 121

【信息科学与工程学】【通信工程】计算机科学与自动化——第二百三十三篇 云NaaS服务 系列二01

国家骨干网(含核心网、IP骨干网、OTN骨干网)/城域网【含3G/4G/5G接入网/PON接入网/OTN接入网+OTN/IP城域网+其他】/中心云(多/双/单Region+多AZ/双AZ/单AZ)+边缘DC+边缘计算节点+端计算节点包括IoT传感器、移动终端、CPE设备。通过5G NR(n78频段,带宽100MHz,子载波间隔30kHz)或GPON(下行2.5Gbps,上行1.25Gbps)接入。本地缓存小于10GB,主要用于临时数据暂存。位于地市级机房,部署MEC服务器(Intel Xeon Platin

weixin_49199313的博客 1382

windows 11 搭建 TensorFlow2.6 GPU 开发环境【RTX 3060】:1 -- 本地原生方式

文章大纲本地方式主要步骤CUDA 本地安装cuDNN 本地安装anaconda 环境构建WSL 2 docker 方式主要步骤参考文献 本地方式 主要步骤 CUDA 本地安装 我点的win 11 版本,比较迷惑的是这个命名方式 CUDA Toolkit 11.5 Downloads 安装完成后: PS C:\Users\season> nvcc --version nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2021 NVID

shiter编写程序的艺术 4373

【信息科学与工程学】【数据科学】数据科学领域——第三篇 数学08 几何学00总篇

编号英文名称中文名称简要说明/归属1Foundations of Geometry几何基础公理化方法, Hilbert《几何基础》2Euclidean Geometry (Synthetic)欧几里得几何(综合法)《几何原本》传统3Analytic Geometry解析几何坐标方法, Descartes, Fermat4Plane Geometry平面几何三角形、圆、共点共线等经典问题5Solid Geometry立体几何三维空间中的图形与度量6Non-Euclidean Geometry非欧几何平行公设不

weixin_49199313的博客 220

【信息科学与工程学】计算机科学与自动化——第三百零五篇 数据中心 Scale-Up、Scale-Out、Scale-Across 18

普通CDN在遭遇50Gbps攻击时,节点瘫痪率高达70%。云化安全CDN的零信任模型通过将动态身份验证、最小权限访问与CDN的分布式加速能力深度集成,实现安全与性能的协同优化。云化的安全CDN(如腾讯云EdgeOne)与普通的安全CDN在架构、功能、性能和适用场景上存在显著差异。某视频平台接入腾讯云EdgeOne+零信任方案后,全球平均延迟降至46ms,攻击拦截率99.97%,验证了协同模型在高并发场景的普适性。​:随着边缘计算普及,云化安全CDN将逐步取代传统方案,成为高交互、高安全场景的标配基础设施。

weixin_49199313的博客 1703

AI4EDA / AI4Design 完整研究报告:全球论文、项目、机构与产业路线

AI4EDA 与 AI4Design 的核心变化,不是“让大模型写 Verilog”这么简单,而是把芯片工程中的规格、代码、脚本、日志、波形、STA report、formal collateral、layout image、PPA 指标、制造计算等对象统一纳入可学习、可检索、可验证、可迭代的闭环。NVIDIA 的论文谱系清晰呈现了这条路径:先建立可执行评测,再做领域模型和数据工程,然后进入工具反馈、多智能体、长上下文、verifier-guided skill evolution。

weixin_41468466的博客 947

【信息科学与工程学】【通信工程】第五十九篇 IP网络设计01

IP网络设计不是只写ip address;它本质是的问题,所以参数天然分布在:层级例子物理/传输(可选但你会要求光路/电路)光功率/OSNR/色散/衰减/BER/FEC;铜/微波链路预算L1/L2 链路模型速率、MTU、成帧开销、串行化、传播时延、缓冲排队/调度/QoS服务曲线、令牌桶、延迟界、丢包概率近似路由/TE/可靠性流守恒、路径容量、切换/收敛时间信息论/可达性Shannon 上限、FEC 有效码率、链路容量 after overhead计算科学/数值。

weixin_49199313的博客 513

【FDE知识库】(1):AI 编程时代开发语言选型分析,云使用go,企业用java,算法python,底层c

AI 编程时代,开发语言选型需综合考量AI适配度、编译速度、跨平台与学习成本。Python以海量语料和即时执行成首选,适用于AI/ML与原型开发;TypeScript凭借类型系统与秒级构建,成为前端及应用层主流;Go以极速编译与统一风格,契合AI代理高效迭代需求。Rust/C++虽性能强,但编译慢、学习成本高,不适于快速试错。综合推荐:优先选Python、TypeScript、Go,兼顾效率与生态,实现最优AI协作体验。

freewebsys的专栏 658

数字政府政务服务AI大模型赋能智能体平台及AI场景应用规划方案(决策评审版).pptx

数字政府政务服务AI大模型赋能智能体平台及AI场景应用规划方案(决策评审版).pptx

T18_低空经济示范区十五五通感一体化低空智联网与飞行服务保障体系设计方案_生态合作版.pptx

T18_低空经济示范区十五五通感一体化低空智联网与飞行服务保障体系设计方案_生态合作版.pptx

多目标粒子群算法求解综合能源系统最优容量配置研究(Matlab代码实现)

内容概要:本文研究了基于多目标粒子群算法(MOPSO)求解综合能源系统(IES)中最优容量配置的方法,并提供了完整的Matlab代码实现。研究旨在通过优化风电机组、光伏机组、燃气轮机、储能系统等多种能源设备的容量配置,实现系统在经济性、环保性和能源利用率等方面的多目标平衡。文中构建了综合能源系统容量配置的数学模型,明确了以最低综合成本、最小碳排放为核心的目标函数,以及能量供需平衡、设备出力能力等关键约束条件,并详细阐述了MOPSO算法的实现流程,包括种群初始化、非支配排序、拥挤距离计算和外部存档更新等核心机制。最后,通过Matlab平台的仿真实验验证了该方法的有效性,结果表明该算法能够生成分布均匀且具有良好多样性的Pareto最优解集,为决策者根据不同侧重点选择最终方案提供了科学依据。; 适合人群:具备一定电力系统、优化算法和Matlab编程基础的研究生、科研人员及工程技术人员。; 使用场景及目标:① 学习和掌握综合能源系统容量配置问题的建模方法;② 深入理解多目标粒子群算法(MOPSO)的核心原理及其在实际工程优化问题中的应用;③ 获取可直接运行和修改的Matlab代码,用于复现研究成果或进行二次开发。; 阅读建议:建议读者在阅读时,首先理解综合能源系统的背景和优化问题的数学建模过程,然后结合提供的Matlab代码,逐步跟踪算法的执行流程,重点关注目标函数和约束条件的代码实现,以及MOPSO算法中非支配排序和拥挤距离计算等核心模块。为了更好地掌握,可以尝试调整算例参数、增减优化目标或修改算法参数,观察结果的变化。

告别概念晕眩:用TC3xx SMU和TLF35584的配置实例,搞懂ASIL D功能安全落地

本文通过英飞凌TC3xx系列MCU的SMU模块与TLF35584电源管理芯片的协同配置实例,详细解析了如何实现符合ASIL D功能安全标准的刹车辅助ECU。文章涵盖SMU报警源配置、响应动作类型、TLF35584的安全状态管理,以及构建完整安全闭环的实践方法,帮助工程师将抽象的安全概念落地为具体设计方案。

weixin_30647065的博客 439
上一篇: GT CS 7280: Network Science Modeling Epidemics
下一篇: ECON5094 Applied Microeconomics
wx____codinghelp
博客等级 码龄2年 771粉丝 115原创
评论
成就一亿技术人!
拼手气红包6.0元
还能输入1000个字符
 
 条评论被折叠 查看
添加红包

请填写红包祝福语或标题

个

红包个数最小为10个

元

红包金额最低5元

当前余额3.43元 前往充值 >
需支付:10.00元
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付元
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值