The latest test details of Qualcomm’s next-generation 2nm flagship processor Snapdragon 8 Elite Gen 6 Pro have been officially exposed. According to current leaked information, Qualcomm is testing dual versions of this top-end chip, mainly focusing on support for memory and storage technology. The standard version will maintain the existing popular LPDDR5X memory, while the higher-positioned Pro version will be upgraded to support LPDDR6 memory with larger bandwidth and join the UFS 5.0 storage standard for the first time. The two versions are consistent in core operating clock, providing the market with flexible hardware options.
This processor is expected to be the first chip in the Android camp to use TSMC’s 2nm process, targeting Ultra-level flagship phones expected to be released in 2027. Thanks to TSMC’s advanced manufacturing process, the new generation of processors has substantial improvements in computing performance and power consumption control. Judging from the early leaked architecture circuit diagram, the Pro version will be equipped with an Adreno 850 image processor with stronger computing power and provide stable multi-core performance. The market also expects that Qualcomm will adopt a chip grading screening strategy in the future and launch specific versions such as 7-core CPU layout to cater to the market demand for models in different price ranges.
SM8975 (mmWave/sub6) block diagrams
ps: who hallucinated 6 variants of SM8975? there is 2 variants, and they both has gone though 3 revisions, so 6 engineering samples
there will only be 2 SM8975s (LP5X and LP6) at launch, with the same clockspeed pic.twitter.com/xxpN4TNWDz
— Reptalica (@Reptalicant) June 18, 2026
The main reason for introducing the dual-version strategy is the sharp increase in 2nm wafer manufacturing costs. Supply chain sources pointed out that the cost of a single chip for the top-end Pro version may exceed US$300, which will put direct material cost pressure on mobile phone brands. Therefore, Qualcomm provides a standard version compatible with LPDDR5X, allowing manufacturers to maintain flagship-level performance while launching regular flagship models with relatively stable pricing. As for the top-end version with optional LPDDR6 and UFS 5.0, it focuses on meeting extreme data transmission and high-load end-side AI computing requirements, focusing on the top product segments.
