Methods and metrics for fair server assessment under real‐time financial workloads

Georgakoudis, Giorgis; Gillan, Charles J.; Sayed, Ahmed Rabee; Spence, Ivor; Faloon, Richard; Nikolopoulos, Dimitrios S.

doi:10.1002/cpe.3704

Cited by 7 publications

(8 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This Section defines our metrics, describes the platforms used and presents salient details of our methodology used to obtain the power readings and calculate the energy consumption. A complete description of our methodology is available in [1].…”

Section: Experimental Setup and Measurement Methodologymentioning

confidence: 99%

“…We conducted experiments only with the highest voltage-frequency settings on each platform, to which we refer as performance mode. Previous work shows that performance mode is the most energy efficient too [1]. The details of the platforms are as follows:…”

Section: Hardware Platformsmentioning

confidence: 99%

“…We mine data presented in previous papers [1,2] to rank three servers in terms of iso-QoS under option pricing workloads: a scale-out microserver based on Calxeda SoCs; a dual-socket Intel Sandy Bridge server; and an Intel Xeon Phi server. Our experimental campaign uses option pricing workloads for which we invested identical effort to optimize on each server.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Iso-Quality of Service: Fairly Ranking Servers for Real-Time Data Analytics

Georgakoudis¹,

Gillan

Sayed

et al. 2015

Parallel Process. Lett.

Self Cite

View full text Add to dashboard Cite

We present a mathematically rigorous Quality-of-Service (QoS) metric which relates the achievable quality of service metric (QoS) for a real-time analytics service to the server energy cost of offering the service. Using a new iso-QoS evaluation methodology, we scale server resources to meet QoS targets and directly rank the servers in terms of their energy-efficiency and by extension cost of ownership. Our metric and method are platform-independent and enable fair comparison of datacenter compute servers with significant architectural diversity, including micro-servers. We deploy our metric and methodology to compare three servers running financial option pricing workloads on real-life market data. We find that server ranking is sensitive to data inputs and desired QoS level and that although scale-out micro-servers can be up to two times more energy-efficient than conventional heavyweight servers for the same target QoS, they are still six times less energy efficient than high-performance computational accelerators.

show abstract

Section: Experimental Setup and Measurement Methodologymentioning

confidence: 99%

Section: Hardware Platformsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Iso-Quality of Service: Fairly Ranking Servers for Real-Time Data Analytics

Georgakoudis¹,

Gillan

Sayed

et al. 2015

Parallel Process. Lett.

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, the current price is valid only for a limited interval, which implies that a decision about buying or selling the instrument is time critical. Binomial option pricing is a popular pricing model that enables accurate prediction with low computational requirements [35]. We use a commercial binomial option pricing kernel and a realistic setup that delivers requests to a worker thread running the kernel.…”

Section: Energy Optimization Of a Latency-critical Workloadmentioning

confidence: 99%

Alea

Mukhanov

Petoumenos

Wang

et al. 2017

ACM Trans. Archit. Code Optim.

Self Cite

View full text Add to dashboard Cite

Energy efficiency is becoming increasingly important, yet few developers understand how source code changes affect the energy and power consumption of their programs. To enable them to achieve energy savings, we must associate energy consumption with software structures, especially at the fine-grained level of functions and loops. Most research in the field relies on direct power/energy measurements taken from on-board sensors or performance counters. However, this coarse granularity does not directly provide the needed fine-grained measurements. This article presents ALEA, a novel fine-grained energy profiling tool based on probabilistic analysis for fine-grained energy accounting. ALEA overcomes the limitations of coarse-grained power-sensing instruments to associate energy information effectively with source code at a fine-grained level. We demonstrate and validate that ALEA can perform accurate energy profiling at various granularity levels on two different architectures: Intel Sandy Bridge and ARM big.LITTLE. ALEA achieves a worst-case error of only 2% for coarse-grained code structures and 6% for fine-grained ones, with less than 1% runtime overhead. Our use cases demonstrate that ALEA supports energy optimizations, with energy savings of up to 2.87 times for a latency-critical option pricing workload under a given power budget.

show abstract

“…We encourage the reader to review to gain insight into the breadth and depth of problems and innovative solutions in the multi‐disciplinary field of high‐performance computational finance.…”

Section: Introductionmentioning

confidence: 99%

Workshop on high‐performance computational finance

Dixon

Moreira

Daly

2015

Concurrency and Computation

View full text Add to dashboard Cite

EDITORIALWorkshop on high-performance computational finance BACKGROUNDThe purpose of this special issue is to collate a selection of representative research articles that were primarily presented at the Seventh Workshop on High-Performance Computational Finance, held in conjunction with SC'14. This annual workshop brings together practitioners, researchers, vendors, and scholars from the complementary fields of computational finance and high-performance computing in order to promote exchange of ideas, discuss future collaborations and develop new research directions. Financial companies increasingly rely on high performance computers to analyze high volumes of financial data, automatically execute trades, and manage risk.As financial market data continue to grow in volume and complexity, computational capabilities of emerging hardware also increases. Extracting high performance from emerging architectures requires a combination of domain knowledge and specialized technical skills. This special issues presents many examples of how researchers, scholars, vendors, and practitioners are collaborating to address high-performance computing research challenges.The scope of this special issue is broad and is representative of the multi-disciplinary nature of computational finance. In addition to submissions that deal with performance and programmability challenges, theoretical analysis, algorithms, and practical experience in computational finance, this issue also includes articles that address practical challenges with running financial applications in the cloud.The efficient deployment of applications on multi and many-core architectures is a central topic in This issue would be incomplete without coverage of the highly impactful STAC A2 benchmark, which is designed to measure the performance of market risk analysis workloads. A. Nikolaev, B. Ilya, and S. Sania [4] describe how Intel is actively contributing to the specification and provides updated benchmark implementations optimized for the Intel Xeon processor E5 and E7 product families and Intel Xeon Phi coprocessor.Performance versus accuracy of computational methods is a pervasive topic in the financial modeling. C. Brugger, G. Liu, C. De Schryver, and N. Wehn [5] consider the calibration of various models with closed form solutions and propose a hybrid solution.We encourage the reader to review [6][7][8] to gain insight into the breadth and depth of problems and innovative solutions in the multi-disciplinary field of high-performance computational finance.

show abstract

Methods and metrics for fair server assessment under real‐time financial workloads

Cited by 7 publications

References 24 publications

Iso-Quality of Service: Fairly Ranking Servers for Real-Time Data Analytics

Iso-Quality of Service: Fairly Ranking Servers for Real-Time Data Analytics

Alea

Workshop on high‐performance computational finance

Contact Info

Product

Resources

About