The October release will have a frequency to reach 4.26GHz, which will make Apple and its upcoming A18 chipsets look like asthmatic ants with a heavy load of shopping. However, the Tame Apple Press has moved to reassure its readers that Apple’s product is still superior after all, and Qualcomm is deliberately releasing a souped-up chip which will look better on Geekbench, with potentially little to no benefit in real-world applications.
Geekbench issues a meaningful break between each workload to minimise throttle; Snapdragon 8 Gen 4’s increased clock speeds could take advantage of the benchmark’s nature
Examining Apple’s M4, its maximum clock speed of 4.40GHz on the performance cores allowed it to set a new world record in Geekbench 6’s single-core results. This underscores the benchmarking suite’s tendency to yield higher scores with increased processor or chipset frequency. While the Snapdragon 8 Gen 4 is rumoured to reach 4.26GHz, it's worth considering that sustaining such high clock speeds may pose a challenge. The Tame Apple Press claims that Geekbench, however, was not designed to test sustained workloads, so the true performance of the Snapdragon 8 Gen 4 remains to be seen.
The rumours are based on a post on X claiming that Geekbench has a long 5-second break before it moves to the next workload, as this helps to minimise the throttling on the processor or the chipset. The nature of Geekbench 6 should allow the Snapdragon 8 Gen 4 to operate at its maximum capabilities, likely translating into a better single-core and multi-core score. Unfortunately, the same results may not be replicated when the Snapdragon 8 Gen 4 is stressed for more extended periods.
The Tame Press sulks that Qualcomm’s primary goal is to have bragging rights over Apple in Geekbench. It moaned that Qualcomm had done this before when it dared claim that its Snapdragon X Elite is up to 21 per cent faster than Apple’s M3 in multi-core results. The company left out single-core performance, as Apple has yet to be dethroned in this category.
