REDEFINING CLOUD-NATIVE PERFORMANCE: A TECHNICAL EVALUATION OF MICROSOFT AZURE’S COBALT 100 ARM-BASED VIRTUAL MACHINES
Main Article Content
Abstract
Cloud computing stands to dramatically change the paradigm with the enablement
of Arm based architectures, shifting performance, power efficiency and cost
optimization possibilities. A new, customized Arm-based virtual machine (VM) called
Cobalt 100, recently announced by Microsoft Azure, and promises to rewire cloud
native computing for workloads of today. This work provides an in-depth performance
comparison between x86-based Azure VM instances and the new Cobalt 100 VMs. Our
approach is to use industry standard benchmarking suites (like Geekbench 6, SPEC
CPU 2017, and Sysbench) on Arm based Cobalt 100 and x86 based Dv5 VMs. We also
measure real-world application workloads (web servers, NGINX, and databases,
MySQL, microservices in Kubernetes cluster) as well. Performance is evaluated in
terms of CPU throughput, memory bandwidth, energy efficiency, cost/performance, and
application latency. Results show Cobalt 100 VMs providing 40% higher price performance and 60% better CPU performance on CPU-intensive workloads than Dv5
series. For webservers and database hosting Cobalt 100 showed 15–25% lower latency
and 30% less power consumption, the perfect solution for your cloud native, black
carbon free deployment! This benchmark proves the potential of Arm architecture in
enterprise cloud, and it positions Cobalt 100 as a proven alternative to traditional x86
VMs for developers and organizations seeking optimal performance and sustainability
for next generation of cloud-native applications.
Article Details
Section
How to Cite
References
L. Pons, J. Petit, J. Gómez, M. Sahuquillo, "A modular approach to build a hardware
testbed for cloud resource management research," The Journal of Supercomputing, vol.
80, no. 8, pp. 10552–10583, Dec. 2023, doi: 10.1007/s11227-023-05856-2.R. Grant, S. Hammond, J. Laros, M. Levenhagen, S. Olivier, K. Pedretti, L. Ward, and
A. Younge, "Enabling power measurement and control on Astra: The first petascale
Arm supercomputer," Concurrency and Computation: Practice and Experience, vol. 35,
no. 15, e7303, Sep. 2022, doi: 10.1002/cpe.7303.
I. Afanasyev and D. Lichmanov, "Evaluating the performance of Kunpeng 920
processors on modern HPC applications," in Parallel Computing Technologies, vol.
1336, pp. 301–321, Springer, 2021, doi: 10.1007/978-3-030-86359-3_23.
D. Xie, Y. Hu, and L. Qin, "An Evaluation of Serverless Computing on x86 and ARM
Platforms: Performance and Design Implications," in Proc. 2021 IEEE 14th Int. Conf.
Cloud Computing (CLOUD), Chicago, IL, USA, pp. 313–321, Sep. 2021, doi:
10.1109/CLOUD53861.2021.00045.
B. Brank, S. Nassyr, F. Pouyan, and D. Pleiter, "Porting Applications to Arm-based
Processors," in Proc. 2020 IEEE Int. Conf. Cluster Computing (CLUSTER), Kobe,
Japan, pp. 559–566, Sep. 2020, doi: 10.1109/CLUSTER49012.2020.00079.
D. Yokoyama, B. Schulze, F. Borges, and G. McEvoy, "The survey on ARM processors
for HPC," The Journal of Supercomputing, vol. 76, no. 10, pp. 8026–8051, Oct. 2019,
doi: 10.1007/s11227-019-02911-9.
A. Jackson, A. Turner, M. Weiland, et al., "Evaluating the Arm ecosystem for high
performance computing," Concurrency and Computation: Practice and Experience, vol.
31, no. 16, pp. e5094, Apr. 2019.
J. Maqbool, S. Oh, and G. C. Fox, “Evaluating ARM HPC clusters for scientific
workloads,” Concurrency and Computation: Practice and Experience, vol. 27, no. 17,
pp. 5390–5410, Jul. 2015, doi: https://doi.org/10.1002/cpe.3602.
J. Noor, M. B. Faysal, M. S. Amin, B. Tabassum, T. R. Khan, and T. Rahman,
“Kubernetes application performance benchmarking on heterogeneous CPU
architecture: An experimental review,” High-Confidence Computing, p. 100276, Dec.
2024, doi: https://doi.org/10.1016/j.hcc.2024.100276.
X. Xu, A. Cui, H. Deng, et al., "virtCCA: Virtualized Arm Confidential Compute
Architecture with TrustZone," arXiv preprint, arXiv:2306.00587, Jun. 2023.
S. Sridhara, A. Das, A. Paverd, et al., "ACAI: Protecting Accelerator Execution with
Arm Confidential Computing Architecture," arXiv preprint, arXiv:2305.01859, May
2023.
[12]
[13]
[14]
[15]
[16]
Dakić, V.; Mršić, L.; Kunić, Z.; Đambić, G. Evaluating ARM and RISC-V Architectures
for High-Performance Computing with Docker and Kubernetes. Electronics 2024, 13,
3494. https://doi.org/10.3390/electronics13173494
K. Gupta and T. Sharma, “Changing Trends in Computer Architecture : A
Comprehensive Analysis of ARM and x86 Processors,” International Journal of
Scientific Research in Computer Science, Engineering and Information Technology,
vol. 7, no. 3, pp. 619–631, Jun. 2021, doi: https://doi.org/10.32628/cseit2173188.
F. Mantovani et al., “Performance and energy consumption of HPC workloads on a
cluster based on Arm ThunderX2 CPU,” Future Generation Computer Systems, vol.
112, pp. 800–818, Nov. 2020, doi: https://doi.org/10.1016/j.future.2020.06.033.
Cluster
Gupta NAshiwal RBrank BPeddoju SPleiter D(2020)Performance Evaluation of
ParalleX Execution model on Arm-based Platforms2020 IEEE International Conference
on
Computing
(CLUSTER)10.1109/CLUSTER49012.2020.00080567
575Online publication date: Sep-2020
L. Xu, Z. Wang, and W. Chen, "The Study and Evaluation of ARM-Based Mobile
Virtualization," Journal of Sensors, vol. 2015, Article ID 310308, pp. 1–12, 2015.