Google is reportedly eyeing a major leap for its Tensor G6 chip, expected to power the Pixel 11 series in 2026. Rumors suggest a potential jump to TSMC’s cutting-edge 2nm manufacturing process, a move that could dramatically change Google’s position in the mobile chip race after years of playing catch-up.
A Pattern of Lagging Behind
Google’s journey with its custom Tensor chips has seen it consistently a generation behind competitors in adopting the latest silicon manufacturing nodes.
The inaugural Tensor G1, powering the Pixel 6 in 2021, was built by Samsung on a 5nm process. The following year’s Tensor G2 (Pixel 7) remained on the same 5nm node, even as rivals like Qualcomm’s Snapdragon 8 Gen 2 transitioned to 4nm.
It took Google a year to catch up, with the Tensor G3 (Pixel 8) finally adopting a 4nm process, a node the current Tensor G4 (Pixel 9) also utilizes. Meanwhile, leading chips from Qualcomm and MediaTek have moved to 3nm processes, highlighting Google’s persistent gap. The Tensor G4 was also noted for using Samsung’s older FO-PLP packaging, which some reports linked to less effective heat management compared to newer techniques.
The Critical Leap: Tensor G5 Moves to TSMC
The immediate focus, however, is Google’s significant planned shift this year. The upcoming Tensor G5, destined for the Pixel 10 series in 2025, is widely reported to mark the end of Google’s partnership with Samsung for mobile Tensor chips.
Instead, the G5 will reportedly be manufactured by TSMC, utilizing their advanced 3nm process (specifically the N3E variant, similar to nodes used by current flagship chips like Apple’s A18/M4 and top-tier Snapdragons/Dimensitys). This move, reportedly part of a multi-year deal with TSMC lasting up to five years, brings Google up to par with leading-edge nodes—though still nodes competitors adopted roughly a year prior.
The switch from Samsung is largely attributed to reports of Samsung’s struggles with low yield rates, particularly on its advanced 3nm nodes (some citing an “abysmal” 20% yield on their 3nm GAA technology), which have made manufacturing less efficient and more costly for Google. Beyond the node size, the G5 is also expected to benefit from TSMC’s advanced InFO-POP packaging, which can improve efficiency and reduce size.
The Pixel 11 Question: A Jump to TSMC 2nm?
Looking further ahead to 2026 and the expected Pixel 11 series, a tantalizing possibility has emerged. According to some reports, citing sources including analyst @dnystedt, Google is considering a dramatic leap for the Tensor G6: utilizing TSMC’s cutting-edge 2nm process node.
This would represent a fundamental shift in Google’s strategy, moving from consistently trailing competitors by a node generation to potentially leading the charge onto the next major fabrication advancement. Some speculation suggests a 2nm Tensor G6 could even arrive months before a competitor like Qualcomm launches its own 2nm Snapdragon flagship, temporarily putting Google ahead in this specific metric.
Conflicting Reports and the High Cost of 2nm
Yet, the picture for the Tensor G6 isn’t entirely clear, and conflicting reports introduce a note of caution. While some sources point to the ambitious 2nm plan, others, including reporting based on documents seen by Android Authority, suggest Google may instead opt for TSMC’s N3P process for the G6. N3P is an enhanced version within the 3nm family, offering improvements over the N3E node used for the G5, but it isn’t the full 2nm leap.
The primary reason cited for this potential pivot is cost. Manufacturing on a bleeding-edge process like 2nm is incredibly expensive, and analysts like Ming-Chi Kuo have highlighted the exceptionally high wafer costs. For a company like Google, which ships fewer phones than market leaders like Apple or Samsung, utilizing such a costly node for relatively lower volumes could be financially prohibitive. Google’s strategic decision may ultimately balance the performance and efficiency gains of the newest nodes against the significant manufacturing expenses. While Samsung also has a 2nm process in development, the reported reasons for leaving Samsung for the G5 make a return for the G6 2nm less likely, though not entirely impossible according to some speculation.
Implications for Pixel 11 Performance
Regardless of whether Google ultimately chooses the 2nm or the refined 3nm (N3P) node for the Tensor G6, the shift to TSMC is expected to yield significant benefits compared to the current Tensor G4 built on Samsung’s 4nm process. TSMC’s nodes are generally regarded as offering better performance and power efficiency.
This aligns with Google’s stated philosophy that its Tensor chips aren’t just designed to win raw benchmark scores but to enhance the overall user experience, particularly for AI tasks, computational photography, and device-specific features. A more efficient and powerful G6, built on a leading TSMC process, should translate into improved performance, better battery life, and potentially more sophisticated on-device AI capabilities for the Pixel 11 series.
The Bottom Line
In summary, while the immediate future sees Google transitioning its Tensor chip manufacturing to TSMC with the G5 on a 3nm process for the Pixel 10, the plans for the subsequent Tensor G6 in the Pixel 11 remain a subject of speculation and conflicting reports. The exciting possibility of a jump to TSMC’s 2nm node exists, potentially putting Google ahead of the curve, but the significant cost associated with such bleeding-edge technology for Google’s shipment volumes makes the alternative of utilizing a refined 3nm (N3P) node a strong, perhaps more probable, possibility. Regardless of the exact process node chosen for the G6, the move to TSMC signifies Google’s ambition to leverage top-tier manufacturing for its Pixel silicon, aiming for improved performance, efficiency, and user experience in future devices. The final decision on the Tensor G6’s node likely hinges on a careful balance of technological ambition and economic realities.
