A Survey on Different Overclocking Methods
01 Mar 2018-
TL;DR: This paper goes through different kind of overclocking methods that can be used to improve the performance of computer such as BCLK overclocking, FSB Overclocking etc.
Abstract: Improving computer performance is always been an area of intensive research. Overclocking is used to gain additional performance from a given component by increasing its operating speed. Through overclocking a computer is forced to operate faster than the manufactured clock frequency. Overclocking increases the operating frequency of a given hardware such as CPU, graphics card, memory chip etc. Increasing a component's clock rate causes it to perform more operations per second. Overclocking is a complex process. There are different overclocking methods that can be used to improve the performance of computer such as BCLK overclocking, FSB Overclocking etc. This paper goes through different kind of overclocking methods.
Citations
More filters
TL;DR: An intelligent overclocking mechanism for the MEC server that operates for an intelligently calculated period to allow it to leverage more computing power without introducing additional hardware resources for a certain period of time is suggested.
Abstract: With the rapid growth of user equipment (UE), the amount of data transmitted over networks has become enormous, exerting immense pressure on backbone networks and central cloud infrastructures. Simultaneously, corresponding applications requiring high energy consumption and low latency have multiplied the requirements for UE. Mobile edge computing (MEC) has been proposed to support the offloading of UE tasks to edge clouds for execution. The implementation of MEC requires fast data transmission between UE and edge servers, and the emerging 5G network appears to render this technology possible. In this paper, considering a large number of UE, a fixed MEC server, and an advanced intelligent network, we suggest an intelligent overclocking mechanism for the MEC server that operates for an intelligently calculated period to allow it to leverage more computing power without introducing additional hardware resources for a certain period of time. We jointly manage task offloading, server resource allocation, and overclocking to minimize the system-wide computation overhead and other risks. The proposed optimization problem is a mixed-integer nonlinear programming problem that is divided into three subproblems: offloading decision, resource allocation, and overclocking decision. We solve these subproblems using non-convex techniques and provide an iterative algorithm to obtain a heuristic solution for the original problem. Finally, simulation results show that the overclocked MEC server has lower system-wide computation overhead, faster task processing, and more offloaded UE as compared with the case without overclocking.
12 citations
DOI•
23 Jul 2021TL;DR: In this article, the authors analyze the performance, power and energy behaviors of four DNN models (i.e., LeNet, GoogLeNet, AlexNet, and CaffeNet) with various parameter settings (both hardware and software) on both CPU and GPU platforms.
Abstract: Deep learning has achieved accuracy and fast training speed and has been successfully applied to many fields, including speech recognition, text processing, image processing and video processing. However, the cost of high power and energy comes together with the high accuracy and training speed of Deep Neural Network (DNN). This inspires researchers to perform characterization in terms of performance, power and energy for guiding the architecture design of DNN models. There are three critical issues to solve for designing a both accurate and energy-efficient DNN model: i) how the software parameters affect the DNN models; ii) how the hardware parameters affect the DNN models; and iii) how to choose the best energy-efficient DNN model. To answer the three issues above, we capture and analyze the performance, power and energy behaviors for multiple experiment settings. We evaluate four DNN models (i.e., LeNet, GoogLeNet, AlexNet, and CaffeNet) with various parameter settings (both hardware and software) on both CPU and GPU platforms. Evaluation results provide detailed DNN characterization and some key insights to facilitate the design of energy-efficient deep learning solutions.
5 citations
TL;DR: In this paper , an adaptive overclocking algorithm (AOA) is proposed to achieve performance improvement while the total energy remains roughly unchanged, where the dynamic power upper bound is set to reflect the real-time behavior characteristics of the program so that algorithm can better meet the total power unchanging constraints.
Abstract: Abstract Although GPUs have been used to accelerate various convolutional neural network algorithms with good performance, the demand for performance improvement is still continuously increasing. CPU/GPU overclocking technology brings opportunities for further performance improvement in CPU-GPU heterogeneous platforms. However, CPU/GPU overclocking inevitably increases the power of the CPU/GPU, which is not conducive to energy conservation, energy efficiency optimization, or even system stability. How to effectively constrain the total energy to remain roughly unchanged during the CPU/GPU overclocking is a key issue in designing adaptive overclocking algorithms. There are two key factors during solving this key issue. Firstly, the dynamic power upper bound must be set to reflect the real-time behavior characteristics of the program so that algorithm can better meet the total energy unchanging constraints; secondly, instead of independently overclocking at both CPU and GPU sides, coordinately overclocking on CPU-GPU must be considered to adapt to real-time load balance for higher performance improvement and better energy constraints. This paper proposes an Adaptive Overclocking Algorithm (AOA) on CPU-GPU heterogeneous platforms to achieve the goal of performance improvement while the total energy remains roughly unchanged. AOA uses the function $$F_k$$ F k to describe the variable power upper bound and introduces the load imbalance factor W to realize the CPU-GPU coordinated overclocking. Through the verification of several types convolutional neural network algorithms on two CPU-GPU heterogeneous platforms (Intel $$^\circledR $$ ® Xeon E5-2660 & NVIDIA $$^\circledR $$ ® Tesla K80; Intel $$^\circledR $$ ® Core™i9-10920X & NIVIDIA $$^\circledR $$ ® GeForce RTX 2080Ti), AOA achieves an average of 10.7% performance improvement and 4.4% energy savings. To verify the effectiveness of the AOA, we compare AOA with other methods including automatic boost, the highest overclocking and static optimal overclocking.
References
More filters
25 Jun 2007
TL;DR: The key objective is to see the effect of overclocking on superscalar processors for various benchmark applications, and analyze the associated overhead, in terms of extra hardware and error recovery penalty, when the clock frequency is adjusted dynamically.
Abstract: Synchronous circuits are typically clocked considering worst case timing paths so that timing errors are avoided under all circumstances. In the case of a pipelined processor, this has special implications since the operating frequency of the entire pipeline is limited by the slowest stage. Our goal, in this paper, is to achieve higher performance in superscalar processors by dynamically varying the operating frequency during run time past worst case limits. The key objective is to see the effect of overclocking on superscalar processors for various benchmark applications, and analyze the associated overhead, in terms of extra hardware and error recovery penalty, when the clock frequency is adjusted dynamically. We tolerate timing errors occurring at speeds higher than what the circuit is designed to operate at by implementing an efficient error detection and recovery mechanism. We also study the limitations imposed by minimum path constraints on our technique. Experimental results show that an average performance gain up to 57% across all benchmark applications is achievable.
51 citations
"A Survey on Different Overclocking ..." refers background in this paper
...In [7], Viswanathan Subramanian, Mikel Bezdek, Naga D....
[...]
...Viswanathan Subramanian [7] Superscalar Processor Performance Enhancement Through Reliable Dynamic Clock Frequency Tuning This paper aims at the performance improvement of superscalar processor through dynamically varying the operating frequency....
[...]
...In [7], Viswanathan Subramanian, Mikel Bezdek, Naga D. Avirneni and Arun Somani propose the mechanism of overclocking superscalar processor....
[...]
...Viswanathan Subramanian [7] Superscalar Processor Performance Enhancement Through Reliable Dynamic Clock Frequency Tuning This paper aims at the performance improvement of superscalar processor through dynamically varying the operating frequency....
[...]
TL;DR: The aim of the work is to exploit the techniques of overclocking and throttling to enhance the performance, also to achieve optimal utilization of processor resources and at the same time maintaining the system reliability.
Abstract: Traditional design methodologies of a digital system assumes the worst case operating conditions to tolerate the physical and environmental variations, which ensure the system operates correctly and conservatively. The clock frequency of the processor is generally set to operate below the maximum permissible operating frequency which achieves less than the maximum performance gains. However, in-fact higher performance gain can be achieved by dynamic overclocking mechanism, which tunes the clock rate beyond the worst case assumptions. Dynamic overclocking mechanism may eventually lead to rise in temperature if it is adopted for a longer duration of time hence this problem is circumvented by using throttling mechanism at appropriate instances. Further, the throttling technique can be used to conserve energy when the system is with low work load. The aim of the work is to exploit the techniques of overclocking and throttling to enhance the performance, also to achieve optimal utilization of processor resources and at the same time maintaining the system reliability. The goal is hence achieved by employing an Adaptive Neuro Fuzzy Inference System to self-tune its frequency based on work load characteristics and also adapting itself to environmental variations.
14 citations
11 Jul 2016
TL;DR: A dynamic speculation window to be used in double sampling schemes for timing error detection and correction in pipelined logic paths is introduced and employed in the Xilinx Virtex VC707 FPGA for various benchmarks.
Abstract: Error detection and correction based on double-sampling is used as common technique to handle timing errors while scaling Vdd for energy efficiency. An additional sampling element is inserted in the critical paths of the design, to double sample the outputs of those logic paths at different time instances that may fail while scaling the supply voltage or the clock frequency of the design. However, overclocking, and error detection and correction capabilities of the double sampling methods are limited due to the fixed speculation window which lacks adaptability for tracking variations such as temperature. In this paper, we introduce a dynamic speculation window to be used in double sampling schemes for timing error detection and correction in pipelined logic paths. The proposed method employs online slack measurement and conventional shadow flipflop approach to adaptively overclock or underclock the design and also to detect and correct timing errors due to temperature and other variability effects. We demonstrate this method in the Xilinx Virtex VC707 FPGA for various benchmarks. We achieve a maximum of 71% overclocking with a limited area overhead of 1.9% LUTs and 1.7% flip-flops.
8 citations
"A Survey on Different Overclocking ..." refers methods in this paper
...Hence this method is considered as a mechanism to improve overall system performance [5][8][12]....
[...]
26 Apr 2016
TL;DR: Based on the calculation results overclock testing conducted on an Intel i5 4670K, increased processor performance increased by 17 %, the resulting temperature rise by 27 % and power output increased by 34 % from the default frequency.
Abstract: There processor FSB (Front Side Bus) and a multiplier that is used in setting the frequency of the processor. The FSB is the bus system is on line (bus) which physically connects the processor with the Northbridge chipset on the motherboard. This path as a path of data / information that is manifested in the form of electronic signals. The FSB also connects the processor to the main memory. Overclocking is to raise the frequency of the processor to improve computer performance. The overclocking process will use two variables above as a regulator of the size of the processor frequency. The greater the value of the FSB and multiplier, the greater the frequency or clock speed of a processor. But the greater the processor frequency, the greater the power generated. Because the greater the frequency, the greater the required current. These currents also generate heat in the processor. Thus, the greater the frequency of the processor, the greater the resulting temperature. Based on the calculation results overclock testing conducted on an Intel i5 4670K, increased processor performance increased by 17 %, the resulting temperature rise by 27 % and power output increased by 34 % from the default frequency.
7 citations
"A Survey on Different Overclocking ..." refers background or methods in this paper
...In [3], Rian Fahrizal, Rocky Alfanz and Awal Sakti analyze the durability of Processor i5 4670K while overclocking the computer....
[...]
...Rian Fahrizal [3] Analysis of Effect Overclocking Durability on Intel Processor i5 4670K FSB overclocking method is used here for analyzing the overclocking durability of Intel Processor i5 4670K....
[...]
...The main drawback to overclocking a computer is voiding the warranty provided by the manufacturer, since the system is not running at the rated specification [3][11]....
[...]
...Rian Fahrizal [3] Analysis of Effect Overclocking Durability on Intel Processor i5 4670K FSB overclocking method is used here for analyzing the overclocking durability of Intel Processor i5 4670K....
[...]
...For FSB overclocking the required amount of extra power supply is small, but even a small increase in voltage can cause processor and memory temperature to reach at a very high value [3][9]....
[...]
01 Jan 2016
TL;DR: Overclocking is not a new method for most PC users with the most cost effective perspective for all processors but it is capable of increasing the performance improvement for heavy workloads.
Abstract: Overclocking is not a new method for most PC users with the most cost effective perspective for all processors. Two main overclocking methods: Front-Size Bus (FSB) and multiplier changing are used as the popular trends. When the processor need to increase the performance improvement for heavy workloads, overclocking method increases the clock speed cycles by changing multiplier values. In fact, overclocking is the best method to implement with trial and error process which is not warranted by the manufacturers but it is capable. The detail description of types of intensive workloads and multiplier changing overclocking results are shown. The testbed of the paper is based on the specification of Intel Core i7 3770 CPU@3.40 GHz with the core speed 1596.19 MHz as well as bus speed 99.76. The most commonly tools such as CPU-Z, RealTemp, CoreTemp, FFT-z, OCCT tool and Benchmark programs testings are used to validate the system evaluation.
5 citations
"A Survey on Different Overclocking ..." refers background in this paper
...Win Win Aung presents [1] Multiplier Changing Overclocking Method for Processor Intensive Tasks, in this paper the author is indicating that under heavy work load we can improve system performance by increasing the reference clock multiplier value....
[...]
...Hence overclocking can be achieved by changing the base clock multiplier [1][4]....
[...]
...RELATED STUDIES Win Win Aung presents [1] Multiplier Changing Overclocking Method for Processor Intensive Tasks, in this paper the author is indicating that under heavy work load we can improve system performance by increasing the reference clock multiplier value....
[...]
...The basic concept of overclocking is to take advantage of the remaining unused capability out of a computer that the manufacture doesn’t want to use, so that they can improve the reliability of the system [1][10]....
[...]
...Win Win Aung[1] Multiplier Changing Overclocking Method for Processor Intensive Tasks CPU performance is improved under heavy workload by changing the multiplier values....
[...]