Last year in March, Samsung was found guilty of thermally throttling around 10,000 apps on its Galaxy smartphones. The culprit was Samsung's new Game Optimizing Service (GOS) which was "optimizing gameplay" by essentially throttling or limiting the performance of other applications and tasks such that gaming is prioritized. This is meant to reduce heat output as well.
Interestingly, it was found that GOS did not include synthetic benchmark tests like 3DMark and Geekbench. Following this, the latter delisted several Galaxy devices, including the S22, the S21, S20, and, the S10 series of phones, as the Geekbench benchmark scores obtained by these phones wouldn't be representative of real-world scenarios.
Primate Labs, the makers of Geekbench, provided the following statement to Android Police:
Earlier this week, we were made aware of Samsung's Game Optimizing Service (GOS) and how it throttles the performance of games and applications. GOS decides to throttle (or not to throttle) applications using application identifiers and not application behavior. We view this as a form of benchmark manipulation as major benchmark applications, including Geekbench, are not throttled by this service.
Regardless, due to over-heating, the high-end Galaxy devices were probably thermal throttling a bit either way during benchmarks, even if GOS wasn't intentionally doing it.
The latest version of Geekbench, version 6.1, has made an interesting change regarding this issue as it has increased the workload gap between its various tests from two seconds up to five seconds. Geekbench explains that this has been done to reduce run-to-run variability.
The full changelog is given below:
Geekbench 6.1 features the following changes:
- Upgrade to Clang 16 Geekbench 6.1 is built with Clang 16 on all platforms. Geekbench 6.1 also improves the optimization switches used when building Geekbench.
- Increase workload gap Geekbench 6.1 increases the workload gap (the pause between workloads) from two seconds to five seconds. The increased workload gap minimizes thermal throttling and reduces run-to-run variability on newer smartphones such as the Samsung Galaxy S23.
- Introduce support for SVE instructions Geekbench 6.1 includes SVE implementations of several image processing and machine learning functions.
- Introduce support for AVX512-FP16 instructions Geekbench 6.1 includes AVX512-FP16 implementations of several image processing functions.
- Introduce support for fixed-point math Geekbench 6.1 introduces fixed-point implementations of several image processing functions. Geekbench uses fixed-point math to implement some image processing functions on systems without FP16 instructions.
- Improve Multi-Core Performance Geekbench 6.1 improves the multi-core implementations of the Background Blur and Horizon Detection workloads, especially on high-end desktop processors such as 12- and 16-core AMD Ryzens, AMD Threadrippers, and Intel Xeons.
Thanks to these changes, Geekbench 6.1 single-core scores are up to 5% higher, and multi-core scores are up to 10% higher than Geekbench 6.0 scores.
As you may have noticed, Geekbench 6.1 is also expanding support for AVX-512 and it is going to greatly benefit the newer AMD Ryzen chips extend their dominance over Intel.