| 241. |
- Sander, Ingo, et al.
(författare)
-
High-Level Estimation and Trade-Off Analysis for Adaptive Real-Time Systems
- 2009
-
Ingår i: 2009 IEEE INTERNATIONAL SYMPOSIUM ON PARALLEL & DISTRIBUTED PROCESSING. - 9781424437511 ; , s. 2985-2988
-
Konferensbidrag (refereegranskat)abstract
- We propose a novel design estimation method for adaptive streaming applications to be implemented on a partially reconfigurable FPGA. Based on experimental results we enable accurate design cost estimates at an early design stage. Given the size and computation time of a set of configurations, which can be derived through logic synthesis, our method gives estimates for configuration parameters, such as bitstream sizes, computation mid reconfiguration times. To fulfil the system's throughput requirements, the required FIFO buffer sizes are then calculated using a hybrid analysis approach based on integer linear programming and simulation. Finally, we are able to calculate the total design cost as the sum of the costs for the FPGA area, the required configuration memory and the FIFO buffers. We demonstrate our method by analysing non-obvious trade-offs for a static and dynamic implementation of adaptivity.
|
|
| 242. |
- Sander, Ingo, et al.
(författare)
-
Modelling Adaptive Systems in ForSyDe
- 2008
-
Ingår i: Electronical Notes in Theoretical Computer Science. - : Elsevier BV. - 1571-0661 .- 1571-0661. ; 200:2, s. 39-54
-
Tidskriftsartikel (refereegranskat)abstract
- Emerging architectures such as partially reconfigurable FPGAs provide a huge potential for adaptivity in the area of embedded systems. Since many system functions are only executed at particular points of time they can share an adaptive component with other system functions, which can significantly reduce the design costs. However, adaptivity adds another dimension of complexity into system design since the system behaviour changes during the course of adaptation. This imposes additional requirements on the design process, in particular system verification. In this paper we illustrate how adaptivity is treated as first-class citizen inside the ForSyDe design framework. ForSyDe is a transformational system design methodology, where an initial abstract system model is refined by the application of semantic-preserving and non-semantic preserving design transformations into a detailed model that can be mapped to an implementation. Since ForSyDe is based on the functional paradigm we can model adaptivity by using functions as signal values, which we use as the base for our concept of adaptive processes. Depending on the level of adaptivity we categorise four classes of adaptive process, spanning from parameter adaptive to interface adaptive process. We illustrate our concepts by two typical examples for adaptivity, where we also show the application of design transformations.
|
|
| 243. |
- Sander, Ingo, et al.
(författare)
-
System modeling and transformational design refinement in ForSyDe
- 2004
-
Ingår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems. - : Institute of Electrical and Electronics Engineers (IEEE). - 0278-0070 .- 1937-4151. ; 23:1, s. 17-32
-
Tidskriftsartikel (refereegranskat)abstract
- The scope, of the Formal System Design (ForSyDe) methodology is high-level modeling and refinement of systems-on-a-chip and embedded systems. Starting with a formal specification model, that captures the functionality of the system at a high abstraction level, it provides formal design-transformation methods for a transparent refinement process of the system model into an implementation model that is optimized for synthesis. The main contribution of this paper is the ForSyDe modeling technique and the formal treatment of transformational design refinement. We introduce process constructors, that cleanly separate the computation part of a process from the synchronization and communication part. We develop the characteristic function for each process type and use it to define semantic preserving and design decision transformations. These transformations are characterized by name, the format of the original process network, the transformed process network, and a design implication. The implication expresses the relation between original and transformed process network by means of the characteristic function. The objective of the refinement process is a model that can be implemented cost efficiently. To this end, process constructors and processes have a hardware and software interpretation which shall facilitate accurate performance and cost estimations. In a study of a digital equalizer example, we illustrate the modeling and refinement process and focus in particular on refinement of the clock domain, communication refinement, and resource sharing.
|
|
| 244. |
|
|
| 245. |
|
|
| 246. |
|
|
| 247. |
|
|
| 248. |
|
|
| 249. |
- Saqib, Eiraj, et al.
(författare)
-
Optimizing the IoT Performance : A Case Study on Pruning a Distributed CNN
- 2023
-
Ingår i: 2023 IEEE Sensors Applications Symposium (SAS). - 9798350323078
-
Konferensbidrag (refereegranskat)abstract
- Implementing Convolutional Neural Networks (CNN) based computer vision algorithms in Internet of Things (IoT) sensor nodes can be difficult due to strict computational, memory, and latency constraints. To address these challenges, researchers have utilized techniques such as quantization, pruning, and model partitioning. Partitioning the CNN reduces the computational burden on an individual node, but the overall system computational load remains constant. Additionally, communication energy is also incurred. To understand the effect of partitioning and pruning on energy and latency, we conducted a case study using a feet detection application realized with Tiny Yolo-v3 on a 12th Gen Intel CPU with NVIDIA GeForce RTX 3090 GPU. After partitioning the CNN between the sequential layers, we apply quantization, pruning, and compression and study the effects on energy and latency. We analyze the extent to which computational tasks, data, and latency can be reduced while maintaining a high level of accuracy. After achieving this reduction, we offloaded the remaining partitioned model to the edge node. We found that over 90% computation reduction and over 99% data transmission reduction are possible while maintaining mean average precision above 95%. This results in up to 17x energy savings and up to 5.2x performance speed-up.
|
|
| 250. |
|
|
