It seems the world of mobile processors is about to undergo a major transformation. Qualcomm is preparing its heavy artillery with its new series of chips, where the Snapdragon 8 Elite Gen 6 Pro (also mentioned in some leaks as Extreme Gen 6) promises to be a game-changer, although the road ahead won't be entirely smooth.
The most striking thing is that we won't have a single flagship chip; instead, the American company has decided to segment its high-end range . The idea is to clearly differentiate the standard model from the Pro, reserving the latter for devices that spare no expense and, above all, have cooling systems capable of handling the demands of high performance.
Raw performance and the first fire tests
The first results have started appearing on Geekbench, although they should be taken with a grain of salt since they are engineering samples. The Extreme Gen 6 model scored 3.434 points in single-core and 9.334 in multi-core . At first glance, these figures are somewhat disappointing compared to the Gen 5, but it's normal for performance to be inconsistent when the firmware is not yet finalized.
One technical detail that has left many scratching their heads is that the device identified as QTI SM8975 failed to maintain its maximum frequency during testing. Although it boasts a top speed of 5.01 GHz , telemetry data indicates it hovered around 4.60 GHz, which explains why the results weren't as explosive as expected.
The architecture: The secret of the Pro model
To achieve such power, Qualcomm has opted for a 2+3+3 CPU design. In the case of the Pro model, the configuration is ambitious: three cores at 3.74 GHz, another three at 4.03 GHz, and the two fastest cores reaching 5.01 GHz , thus surpassing the psychological barrier of 5 GHz in smartphones. This comparison of architectures and performance helps to understand Qualcomm's ambition.
- Adreno 850 GPU: Exclusive to the Pro model, while the standard would use the Adreno 845.
- Graphic Memory: The Pro version has 18 MB compared to the 12 MB of the base version.
- L2 cache: Both versions share a 16 MB cache to optimize efficiency.
- Memory Support: Only the Pro model will be compatible with the cutting-edge Lleaving LPDDR5X for the standard chip.
This strategy aims to make the Pro model the true technological showcase, also compatible with UFS 5.0 storage and manufactured using TSMC's 2nm process. By limiting LPDDR6 support to the more expensive model, Qualcomm justifies the premium price and targets the chip towards phones with large vapor chambers and bulkier designs.
The shadow of Samsung and the Exynos 2700
But beware, Qualcomm doesn't have a clear path. Rumors indicate that Samsung's Exynos 2700 could give it a run for its money in performance. According to internal leaks, the Korean chip has outperformed the Snapdragon 8 Elite Gen 6 Pro in multi-core performance by 9.5% and in GPU performance by 5%, using Samsung's SF2P technology (its second generation of 2nm GAA).
The most worrying aspect for Qualcomm is energy efficiency. In tests with a 2.5W limit, the Exynos was reportedly 22% more efficient than the Pro. Furthermore, in AI tasks using Llama 3.1 8B , Samsung's processor was said to be 18% faster. This, combined with the fact that Exynos chips are cheaper for Samsung, makes it highly likely that the upcoming Samsung Galaxy S27 will once again use its own processors in various regions.
In short, we are facing a pitched battle where Qualcomm is betting on aggressive segmentation and record frequencies, while Samsung seeks to win the game through efficiency and vertical integration, leaving the end user in a privileged position to choose between raw power or energy optimization.
