Summary

An Inside View of Mobile Video Gaming over 5G

SRG just completed its 41st 5G benchmark study. For this endeavor we collaborated with GameBench, along with Accuver Americas and Spirent Communications to conduct an independent benchmark study on how mobile video gaming performs over 5G, including 5G FWA, SA, NSA and LTE, mmWave and Open RAN. We also include traditional broadband cable for comparison purposes.

Highlights of the Report include the following:

Our Thanks. We did this study primarily with the use of the GameBench Network Analysis Tool (NAT) to evaluate a host of performance metrics that pertain to mobile video gaming. We also continued our close relationship with Accuver Americas (XCAL-M and XCAP) and Spirent Communications (Umetrix Data), whose tools and platforms we used during the study. SRG is responsible for the data collection and all analysis and commentary provided in this report.

All Throughput Isn’t Created Equal. If we were to summarize our findings in one phrase, it would be “all throughput isn’t created equal.” In all test scenarios, the network was capable of delivering far more bandwidth than our most bandwidth intensive mobile gaming session, yet it was the “little things” like the choice of the air interface (5G, LTE, or mixed) used to deliver the video packets, the variability of the throughput, and the core network architecture that played a significant role.

Our Methodology. We used the GameBench NAT, which is a client/server platform that is designed to play cloud gaming test streams at various resolutions, frame rates, and corresponding data rates from the public cloud (in our case a geographically close Google cloud server) to the smartphone. The application on the phone also interacts with the video streaming session by requesting character movements or actions, like a trigger request, which can then be used to measure latency-related metrics along with the video streaming metrics.

The Scope. We used a range of mobile gaming scenarios, from 720p 60 fps (12 Mbps) to 1080p 120 fps (75 Mbps) on each network that we tested. We ran the sequence of tests for at least one hour on each network configuration to get a large sampling of results. Some networks survived the most challenging test cases while other networks fell short with the most basic mobile video gaming session.

Other Fun Snippets of Info. We didn’t officially rank results since it wasn’t an apples-apples study, but overall the Mediacom 1 Gbps broadband cable service had the best KPIs, with one KPI that lagged its 5G ppers. 5G mmWave was surprisingly good when it avoided using LTE, while 5G FWA performance was disappointing. SA was generally better than NSA, especially during handovers, while LTE-only was better than expected. DISH Open RAN more than held its own, other than latency which was higher than most other networks.

Next Steps. For a follow-on study, we plan to flip the narrative and look at a range of devices / chipsets against a common 5G network scenario (likely 5G mid-band SA). These tests will help us identify performance differences between phones (high tier to low tier), and 5G modems/application processors. Stay tuned.