Google’s in-house Tensor chips from the beginning have faced criticism for not offering solid performance. While they are excellent for everyday tasks, the performance gap is pretty significant when compared with the latest flagships from Qualcomm and MediaTek.

The latest Tensor G5, despite Google claiming a “major performance boost,” couldn’t get anywhere near the Snapdragon 8 Elite’s benchmark figures. The story remained unchanged when compared against Snapdragon 8 Gen 3 — again, a big performance gap in Snapdragon’s favor. Now, we’re interested in finding out if the Tensor G5 can match or outperform a nearly three-year-old Snapdragon 8 Gen 2.

Let’s get straight to the benchmark numbers.

Tensor G5 vs Snapdragon 8 Gen 2: Benchmark score

AnTuTu comparison

The Snapdragon 8 Gen 2 still holds a slight edge in overall AnTuTu with a score of 1,512,682, compared to the Tensor G5’s 1,429,557. But the story gets more interesting when you look at the breakdown.

The Tensor G5 pulls ahead in the CPU department, scoring 483,989 vs 373,687 — a 29% lead that shows Google’s new custom cores are much stronger for raw processing. On the other hand, the Snapdragon 8 Gen 2 still holds a higher GPU score, with 599,957 vs 416,313 — a massive 44% advantage. That means Qualcomm’s older chips are still superior to the latest Tensor G5 for graphics-intensive workloads, such as gaming and 3D rendering.

Memory and UX scores are closer, with Tensor G5 slightly ahead in memory (293,467 vs 273,714) and Snapdragon 8 Gen 2 leading in UX (265,325 vs 235,788). So, according to AnTuTu scores, both chips should feel smooth for day-to-day use, but Snapdragon’s stronger GPU still gives it the gaming edge.

Geekbench (CPU) comparison

Google claims a 30% faster average CPU uplift for the Tensor G5 compared to last year’s G4, and this improvement is also reflected in benchmarks. On AnTuTu, Tensor G5 scored higher in CPU, and Geekbench also paints a similar picture.

In single-core performance, the Tensor G5 scores 2,316 vs 1,912, giving it a 21% lead. This means the latest Tensor should feel noticeably snappier in everyday tasks, such as app launches, UI responsiveness, and web browsing.

The gap widens even further in multi-core performance, with the G5 reaching 6,452 vs 5,204 — about a 24% advantage. This should result in better handling of multitasking, heavier apps, and background processes.

Overall, while the Snapdragon 8 Gen 2 still has the edge in GPU and gaming (as seen on AnTuTu), both benchmarks (AnTuTu and Geekbench) show stronger CPU performance for the Tensor G5. So, except for gaming and other graphics-intensive tasks, the new Tensor should feel snappier and more responsive.

Tensor G5 vs Snapdragon 8 Gen 2: Spec sheet

The spec sheet clearly shows why the Tensor G5 has a higher CPU score than the 8 Gen 2 on benchmarks. It uses newer CPU cores, which are more performance-friendly and power-efficient. Plus, they also run at a higher speed. Additionally, TSMC’s 3nm advanced manufacturing technology further enables the Tensor G5 to achieve better CPU performance and efficiency.

As for the GPU, the Tensor G5 features PowerVR DXT-48-1536, whereas the 8 Gen 2 comes with Adreno 740, which supports ray tracing and a full suite of Snapdragon Elite Gaming features. For on-device AI capabilities, the G5 has a new Google Edge TPU, which is likely more capable than the Qualcomm Hexagon NPU on the 8 Gen 2. Another notable difference is the newer Bluetooth v6.0 standard on the Tensor chip, whereas the 8 Gen 2 has Bluetooth 5.3. 

Tensor G5 phones:

• Google Pixel 10
• Google Pixel 10 Pro
• Google Pixel 10 Pro XL
• Google Pixel 10 Pro Fold

Snapdragon 8 Gen 2 phones

• Galaxy S23 / S23 Plus / S23 Ultra
• Galaxy Z Fold 5 / Z Flip 5
• OnePlus Open
• iQOO 11
• Motorola Edge 40 Pro
• OnePlus 12R

The post Tensor G5 vs Snapdragon 8 Gen 2: can the latest Tensor outperform this 3-year-old Snapdragon flagship? appeared first on Gizmochina.

Google’s in-house Tensor chips have steadily improved over the years to become a reliable daily driver with solid battery efficiency. Still, performance has often been a sticking point. While previous Tensor chips have handled everyday tasks well, they fell short when it came to gaming and other demanding workloads. The latest Tensor G5 could change that as it promises significant performance upgrades.

The tech giant says “Tensor G5 is our biggest upgrade yet and delivers a major performance boost.” The latest Tensor chip, debuted last month with the Pixel 10 series, is claimed to have 60% more powerful TPU and 30% faster average CPU. This is also Google’s first mobile chipset manufactured at TSMC’s factory (previous Tensor chips were made by Samsung). The Tensor G5 utilizes a more advanced 3nm process node, which promises improved performance and enhanced battery efficiency.

Google promises significant upgrades for Tensor G5, but will that bring a substantial boost to benchmark numbers where previous Tensor chips have not performed well? We’ll look at the benchmark figures. Before that, here’s the spec sheet to see the differences on paper:

Google Tensor G5 vs Tensor G4: AnTuTu score

AnTuTu (version 10) benchmark shows a modest overall gain for the new Tensor G5 chip with a score of 1,429,557 compared to Tensor G4’s 1,348,226 — that’s a 6% gain. The CPU, however, shows a nice 20% increase (483,989 vs 401,746). The memory and UX scores also favor the newer chip with an increase of around 10%.

The most surprising of all is the GPU score, which is actually lower than last year’s G4 — the difference is about 10%. This suggests Google may have focused more on efficiency rather than pushing raw graphics performance.

Google Tensor G5 vs Tensor G4: Geekbench score

On Geekbench 6, the Tensor G5 shows clear improvements over the Tensor G4. In the single-core test, it scores 2,316, about 20% higher than the G4’s 1,923, meaning faster responsiveness in everyday tasks like app launches and UI navigation.

The multi-core score jumps to 6,452, a massive 40% gain over the G4’s 4,602, highlighting stronger performance in multitasking, productivity apps, and heavier workloads.

Overall, Geekbench results suggest promising improvements for Tensor G5 in raw CPU performance, especially in multi-core efficiency, which should translate into much smoother performance.

Google Tensor G5 vs Tensor G4: 3DMark score

The Tensor G5 shows a clear improvement over the Tensor G4 in the 3DMark Wild Life Extreme Stress Test, which measures sustained GPU performance under heavy load. The G5 achieves a best loop score of 3,319, compared to the G4’s 2,603, marking an impressive 27% uplift in peak graphics performance.

More importantly, the G5 also delivers better consistency. Its lowest loop score is 2,954, while the G4 drops to 2,071. That’s not only higher performance, but also much greater stability — 89% vs 79.6%. This means the Tensor G5 is far less prone to thermal throttling and should provide smoother, more reliable gaming performance over longer sessions.

Tensor G5 vs Tensor G4: Verdict

The Tensor G5 shows promising upgrades to CPU performance, thanks to more advanced TSMC manufacturing and improved CPU cores. The 3DMark test shows that GPU performance and consistency have also improved noticeably. This should result in snappier everyday performance and a better gaming experience. The new chip also addresses most of the heating and throttling issues reported by users with previous Tensor chips.

While the newer Tensor chip brings notable upgrades, Google still has a lot to catch up to compete with the best from Qualcomm and MediaTek. We’ve compared the Tensor G5 with Snapdragon 8 Elite and Snapdragon 8 Gen 3. You can explore those comparisons to find out how Google’s latest chip stacks up against the industry best.

The post Tensor G5 vs Tensor G4: Benchmarks and specs appeared first on Gizmochina.

The Google Pixel 10 series is finally here with the all-new Tensor G5 chipset, promising significant upgrades over last year’s G4 chip. This is the biggest upgrade to Tensor chips and promises a significant performance increase, but despite all that, it couldn’t stand strong against the Snapdragon 8 Elite (see the comparison here). The latter is much more powerful than the Tensor G5.

So now, we’re more interested in finding out how the latest Tensor chip fares against the 2023’s Snapdragon 8 Gen 3. Can it beat the Snapdragon’s two-year-old flagship chipset? Let’s see what the benchmarks have to say.

Tensor G5 vs Snapdragon 8 Gen 3: Benchmark score

Note: The Tensor G5 benchmark scores are obtained from leaked sources (1, 2). We’ll update them once we test the device. The reference devices used for benchmark testing are Pixel 10 Pro XL (Tensor G5) and Galaxy S24 Ultra (Snapdragon 8 Gen 3).

Tensor G5 vs Snapdragon 8 Gen 3 AnTuTu score

The Snapdragon 8 Gen 3 pulls well ahead of the Tensor G5 in AnTuTu, scoring 1,768,372 compared to the Tensor’s 1,140,286 — a gap of nearly 55% in overall performance.

The Qualcomm chip leads across the board. Its CPU score of 441,650 is about 41% higher than Tensor’s 313,500. The GPU score widens the gap even further, with the Snapdragon chip posting nearly a 66% boost, which should translate into smoother gaming and stronger graphics performance.

The memory (371,609 vs 246,571) and UX (302,008 vs 185,520) also lean in favor of the Snapdragon. So, while the Tensor G5 promises a significant performance boost, it still can’t beat the Snapdragon 8 Gen 3 in raw performance.

Tensor G5 vs Snapdragon 8 Gen 3 Geekbench score

On Geekbench, the results are much closer than on AnTuTu. In fact, the Tensor G5 achieves a higher single-core score than the Snapdragon 8 Gen 3 (2,296 vs. 2,216), but the gap is narrow. The Snapdragon chip outperforms in multi-core performance, achieving a 9.5% higher score (6,791 vs. 6,203).

In terms of raw performance, the Snapdragon 8 Gen 3 outperforms Google’s latest Tensor G5 mobile chipset significantly in benchmarks. Let’s now look at the spec sheet and major differences for better clarity about these two chipsets.

Tensor G5 vs Snapdragon 8 Gen 3: Spec sheet

For the first time, Google chose TSMC to manufacture the Tensor G5, while the previous Tensor chips were manufactured by Samsung. The former is well-reputed for making high-performance and highly efficient smartphone chips.

The Snapdragon 8 Gen 3 is also manufactured using TSMC’s manufacturing facility. The difference here is that the Tensor G5 utilizes a more advanced 3nm processing node, compared to the Snapdragon 8 Gen 3’s 4nm, resulting in improved power efficiency.

The two chips also feature different CPU architectures for mid-cores, with the Tensor G5 equipped with five Cortex-A725 cores, while the Snapdragon chip utilizes five Cortex-A720 cores. The remaining cores, however, are the same.

For visual rendering, the Tensor G5 features a PowerVR DXT-48-1536 GPU, whereas the Snapdragon 8 Gen 3 has an Adreno 750 GPU with higher floating-point operations per second (FLOPS).

Tensor G5 vs Snapdragon 8 Gen 3: Conclusion

The Tensor G5 brings many solid upgrades, and for this reason, Google calls it “a game-changer for Pixel”. However, when compared to the Snapdragon 8 Gen 3, it’s clear that Google still has some catching up to do. The Snapdragon chip dominates in almost every category on AnTuTu and Geekbench.

Please note that the comparison uses leaked benchmark data for Tensor G5. So, there can still be some differences when we’ve the data from our testing. We’ll update the article accordingly.

The post Tensor G5 vs Snapdragon 8 Gen 3: Benchmarks and specs appeared first on Gizmochina.

The Tensor G5 is the latest and greatest smartphone chip from Google, debuted this week with the Pixel 10 series. Google describes it as its biggest upgrade yet, promising a significant boost in performance — a key area where previous Tensor chips have often been criticized.

The Tensor G5 chip is a game-changer for Pixel, says Google. Besides a major performance leap, the new chip promises better efficiency, greater on-device AI capabilities, and excellent photography, thanks to a new ISP. But how does Google’s latest Tensor chip stack up against Qualcomm’s most powerful Snapdragon 8 Elite? Let’s dive into benchmark scores and real-world differences to find out.

Tensor G5 vs Snapdragon 8 Elite: Benchmark score

[Note: The Tensor G5 benchmark scores are obtained from leaked sources (1, 2). We’ll update them once we test the device. The devices in question are Pixel 10 Pro XL (Tensor G5) and Galaxy S25 Ultra (Snapdragon 8 Elite).]

AnTuTu comparison

The Snapdragon 8 Elite easily outperforms the Tensor G4, scoring 2,209,476 on AnTuTu compared to 1,140,286 — nearly double the score. It delivers stronger CPU (574,518 vs 313,500) and GPU (842,351 vs 394,695) performance on AnTuTu. Even in memory and UX, the Snapdragon chip maintains a clear advantage, resulting in smoother multitasking, faster app performance, and better gaming.

Geekbench comparison

The Snapdragon 8 Elite delivers a commanding lead over the Tensor G4 on Geekbench as well. Its 3,179 single-core score easily outperforms Tensor’s 2,296, giving Qualcomm a 38% advantage.

In multi-core performance, the gap widens further — 10,114 vs 6,203 — meaning the Snapdragon handles multitasking, productivity apps, and demanding workloads far more efficiently.

So, while Google claims a significant performance upgrade for the Tensor G5, it’s still a lot behind Qualcomm’s Snapdragon 8 Elite. This gap could widen even further with the Snapdragon 8 Elite 2, set to be announced next month.

Tensor G5 vs Snapdragon 8 Elite: Spec sheet

Tensor G5 vs Snapdragon 8 Elite: Conclusion

The Tensor G5 marks a big leap forward for Google, bringing much better performance, efficiency, and AI capabilities than any previous Tensor chip. However, when compared to Qualcomm’s Snapdragon 8 Elite, it’s clear that Google still has some catching up to do. The Snapdragon chip dominates in almost every benchmark category — from CPU and GPU to memory and overall system performance — delivering faster gaming, smoother multitasking, and a more responsive experience.

With the Tensor G5, the focus is also on advanced AI processing, smarter on-device intelligence, and camera-driven optimizations that aren’t reflected in raw benchmark scores. If your priority is sheer power, the Snapdragon 8 Elite still leads the pack. But if you’re after Google’s unique AI features and Pixel’s software-first experience, the Tensor G5 could still be the better fit.

Google Tensor G5 phones:

• Google Pixel 10
• Google Pixel 10 Pro
• Google Pixel 10 Pro XL
• Google Pixel 10 Pro Fold

Snapdragon 8 Elite phones:

• Galaxy S25 / S25+ / S25 Ultra
• Xiaomi 15 Ultra
• ASUS ROG Phone 9 Pro
• OnePlus 13
• iQOO 13
• Honor Magic 7 Pro
• POCO F7 Ultra
• Vivo X200 Ultra

The post Tensor G5 vs Snapdragon 8 Elite: Flagship showdown appeared first on Gizmochina.

Google has launched the Pixel 10 series with its most advanced processor yet, the Tensor G5. Built on TSMC’s 3 nm process, the chip delivers faster performance, more efficient AI, stronger security, and longer battery life.

Performance Upgrades

Tensor G5 includes a TPU that is up to 60 percent more powerful and a CPU that is 34 percent faster on average compared to the previous generation. The 3 nm process allows more transistors to fit on the chip, improving speed and efficiency. New security hardware is built into Tensor G5 to protect Pixel devices from the manufacturing stage through everyday use.

On-Device AI

Tensor G5 is the first mobile chip to run Google’s Gemini Nano model fully on-device. This allows Pixel 10 to handle generative AI features without relying on the cloud. Gemini Nano runs 2.6 times faster and twice as efficiently with Tensor G5, powering features such as Pixel Screenshots, Recorder, Magic Cue, Call Notes with actions, Voice Translate, and Personal Journal. Scam Detection and Gboard’s natural voice editing also run better with the new chip.

Camera Improvements

Pixel 10 Pro uses an all-new Image Signal Processor inside Tensor G5. It improves photo and video quality with motion deblur, 10-bit video by default in 1080p and 4K30, and an updated Real Tone system for more accurate skin tones. Tensor G5 also powers Add Me, Auto Best Take, and Pro Res Zoom, which extends zoom to 100x with more detail. The Camera app now includes C2PA Content Credentials, created on-device using Tensor G5 and the Titan M2 security chip for verified metadata.

Battery Life

Despite the performance and feature upgrades, every Pixel 10 delivers more than 30 hours of battery life.

For more daily updates, please visit our News Section.

Tech enthusiast? Get the latest news first! Follow our Telegram channel and subscribe to our free newsletter for your daily tech fix!

(Source)

The post Pixel 10’s Tensor G5 chip is Google’s biggest leap in performance with 60% faster TPU, 34% faster CPU, and new security hardware appeared first on Gizmochina.

Google’s custom chips are getting a new home. According to a report from Digi Times, Google is working out a long-term deal with TSMC to manufacture its Tensor processors for future Pixel phones. 

The agreement will reportedly last anywhere from three to five years. If everything goes as planned, TSMC will be granted the rights to manufacture Tensor chips from the upcoming Pixel 10 to the Pixel 14.

This is a big move. Google has relied on Samsung Foundry since the debut of the original Tensor chip. However, as the Samsung Foundry is facing yield issues with its own Exynos 2500 SoC, Google seems to be taking a step in the right direction. 

The Pixel 10’s Tensor G5 will be the first chip to come out of the new TSMC partnership. It will be manufactured on TSMC’s N3E process, a more advanced version of the 3nm node. 

Compared to Samsung’s 4nm process used in the Tensor G4, this should bring better performance, lower power draw, and improved thermal efficiency. Basically, all the things Google’s chips have been criticized for.

Tensor G5 will be the first Tensor chip by TSMC 

The Tensor G5, internally called “Laguna,” will come with a pretty typical layout: one Cortex-X4 core for performance, five Cortex-A725 cores for mid-level tasks, and two smaller Cortex-A520 cores for efficiency. 

The main focus here is still on Google’s own TPU, its custom AI hardware that powers a lot of the on-device machine learning features, like voice typing, photo processing, and more. The goal is to keep more of that work local to the device instead of relying on the cloud.

Google is also using TSMC’s InFO-POP packaging for the chip, which should help make it thinner and better at handling heat. And it’s not just the foundry that’s changing. The Tensor G5 is dropping a number of Samsung-made components, too. 

The image signal processor is now designed by Google itself. The display controller comes from VeriSilicon, the video codec from Chips&Media, and the modem is being sourced from MediaTek. It’s a pretty clear signal that Google wants more control over its hardware stack, while slowly distancing itself from Samsung’s ecosystem.

For more daily updates, please visit our News Section.

Don’t miss a thing! Join our Telegram community for instant updates and grab our free daily newsletter for the best tech stories!

The post Google will use TSMC to build Tensor chips from Pixel 10 to Pixel 14 appeared first on Gizmochina.

Google’s upcoming Tensor G5 chipset looks set to bring some of the biggest changes to the company’s in-house silicon yet. 

Previous reports have suggested that the processor will move away from Samsung’s foundries and opt for TSMC’s advanced 3nm manufacturing process. But fabrication is not the only area seeing a shake-up. Google might also be switching GPU providers entirely.

A now-deleted document from a RISC-V Day Tokyo conference last month briefly listed Tensor G5 among chips set to use Imagination Technologies’ GPUs. A postgraduate researcher shared the document, which was later amplified by Imagination on WeChat, adding weight to the claim. 

If accurate, it would mark a shift away from the Arm Mali GPUs Google has used in previous Tensor chips.

Tensor G5 with Imagination GPU

The specific GPU in question is said to be the DXT-48-1536, a modern graphics processor that supports features like ray tracing—something the Mali-G715 in the Tensor G3 and G4 lacks. 

This could bring improvements in rendering performance, making Tensor-powered Pixel devices more competitive in gaming and graphics-heavy applications.

Beyond the GPU, the Tensor G5 is expected to incorporate several fully custom components, including a new image signal processor (ISP). Google’s past Tensor chips have relied on modified Samsung ISPs with custom tweaks, but the G5 could introduce a more independent solution. 

Additionally, the chipset is rumored to feature a custom memory controller, system-level cache, and power management modules. For connectivity and storage, Google will still rely on third-party solutions for USB, PCIe, display interfaces, and LPDDR5x memory.

Interestingly, the same document that mentioned the Tensor G5 also referenced a custom Xiaomi chipset called “XuanJie,” which is said to feature an Imagination GPU. 

Xiaomi has reportedly been developing its own chips as well, and this one is believed to be based on a 3nm process. Details are still scarce, though, so we’ll have to wait and see what Xiaomi has in store.

For more daily updates, please visit our News Section.

Stay ahead in tech! Join our Telegram community and sign up for our daily newsletter of top stories!

The post New Leak Accidentally Confirms Google Tensor G5’s Switch to Imagination GPU appeared first on Gizmochina.

Google‘s next-generation Tensor G5 chip, expected to power the upcoming Pixel phone lineup, has reportedly surfaced on the Geekbench benchmarking platform. The listing, while not confirmed by Google, sheds light on the chip’s potential performance and core configuration of the SoC.

Tensor G5 in Geekbench?

Listed with model ID “Google Frankel” in the Geekbench listing (previously referred to as “Laguna” in leaks), the Tensor G5 allegedly scored 1323 points in the single-core test and 4004 points in the multi-core test. 

These scores fall short of even the Tensor G4, so it’s unlikely the G5 will challenge Qualcomm’s new Snapdragon 8 Elite or MediaTek’s Dimensity 9400 flagship chips. It even seems difficult for Google to outpace last-generation flagship chips from these competitors.

The Geekbench listing aligns with previous leaks by detailing a core configuration with a single high-performance core clocked at 3.40 GHz, five mid-range cores clocked at 2.86 GHz, and two low-power cores clocked at 2.44 GHz. The listing also confirms the inclusion of an Imagination Technologies PowerVR D-Series DXT-48-1536 GPU.

Earlier reports had indicated that Google would be switching from Samsung to TSMC for the production of the Tensor G5. The chip is expected to be built on TSMC’s 3nm N3E process node, with a die size of approximately 121 mm².

Now this benchmark of Google’s upcoming SoC could mean either the chip is in its early development stage, leading to lower-than-expected scores, or Google isn’t prioritizing top-tier performance. A company executive previously stated that the Tensor G4 wasn’t designed for record-breaking performance but for improving user experience.

We hope the Tensor G5 won’t follow the same path, especially because it’s supposed to be a flagship SoC that will power phones north of $1000. 

This early version of the Tensor G5 likely needs significant optimization to improve its performance in benchmarks.

(Source)

The post Tensor G5 makes unexpected run on Geekbench, but with under-par performance appeared first on Gizmochina.

For years, Google’s Tensor chips have been manufactured in collaboration with Samsung Foundry, but that might change with the upcoming Tensor G5, expected to debut with the Pixel 10 series. A recent leak from Google’s gChips division, shared via Android Authority, indicates that Google is opting for TSMC’s advanced 3nm process for the next-gen chip, potentially leading to enhanced efficiency and performance.

Redesigned CPU, a new GPU, and slightly improved AI capabilities

One of the significant changes with the Tensor G5 is its revamped CPU layout. Unlike its predecessor, the Tensor G4, which showed only a minor boost in performance, the G5 maintains a primary Cortex-X4 core but introduces a reconfigured mid-cluster.

The new design comprises five Cortex-A725 cores, an increase from three Cortex-A720 cores, while the little cluster has been reduced to two Cortex-A520 cores.

This alteration suggests that Google is aiming for a better balance between performance and power consumption, although sticking to the Cortex-X4 instead of adopting the latest Cortex-X925 is an intriguing choice.

The Tensor G5 also breaks tradition by moving away from the ARM Mali GPU, featuring a new GPU from Imagination Technologies (IMG), the DXT-48-1536. This GPU introduces ray tracing support—a first for Tensor chips—along with GPU virtualization capabilities, which could enhance graphics performance for gaming and other intensive tasks.

While Google’s new TPU in the Tensor G5 brings a 40% increase in TOPS (Trillions of Operations Per Second), internal benchmarks suggest a more modest 14% improvement in real-world scenarios. Additionally, the TPU now supports on-device training and features embedded RISC-V cores, enabling operations that weren’t previously possible, potentially broadening its AI capabilities.

Good news for those waiting for the Pixel 10 series

The Pixel 10, anticipated to launch in the second half of next year, will be the first device to feature the Tensor G5. This shift to TSMC’s 3nm process could significantly enhance power efficiency, addressing past criticisms of the Tensor lineup, particularly issues related to thermal management and performance lags. Reports suggest that Google is looking to better compete with Snapdragon-powered smartphones by making this strategic switch, as TSMC’s fabrication technology has been a cornerstone for Qualcomm’s recent successes​.

Overall, these updates hint at Google’s ambitious plans to position the Pixel 10 series as a stronger contender in the flagship smartphone market.

(Via)

The post Pixel 10’s Tensor G5 finally uses TSMC’s 3nm process, but with ‘weird’ core choices appeared first on Gizmochina.

We have come across multiple rumors in the past claiming that Google will switch to TSMC for its Tensor chips. Previously leaked documents have already revealed that Google will switch to TSMC for the fabrication of the Pixel 10 series’ Tensor G5 chipset.

Pixel devices aren’t the best when it comes to power efficiency

While Pixel devices are among the smartest smartphones on the market, they are not the most power-efficient. Power efficiency is crucial for a device that relies on limited battery power. A less efficient device requires better cooling solutions or must reduce performance to operate safely. We’ve seen numerous reports in the past about Pixel devices getting hotter than expected under certain use cases.

The higher-than-usual power consumption and occasional overheating are attributed to the process nodes on which Tensor chips are fabricated. For those unfamiliar, Google’s Tensor chips are produced by Samsung Foundry.

Things could change with the Tensor G5 powered Pixel 10 series

Now a report out of Taiwan claims that Google has partnered with TSMC and has already made some progress in the development of the Tensor G5. Notably, the Tensor G5 will use TSMC’s 3nm process. Google is also reportedly investing in independent research and development of its chips.

Tensor G5 is going to be Google’s first independently designed smartphone chip. All the Tensor chips Google has used in Pixel devices so far were customized by Samsung. In contrast, Google Tensor G5 will be the first to adopt the company’s own architecture design, which is expected to significantly improve the performance of the chip. It not only holds the potential to solve some of the glaring issues of Tensor chips but also offers Google more control over its chips.

According to the report, Google will not only benefit from a higher power efficiency on the Tensor chips but will also be able to offer a more capable AI experience on Pixel devices with this development. The integration between hardware and software is also expected to improve with the use of Google‘s own architecture.

(Source)

The post Pixel 10’s Tensor G5 chip to be designed by Google and built by TSMC’s 3nm Process appeared first on Gizmochina.

According to some previous rumors, Samsung has achieved success with its 3nm process nodes and the development of its 2nm process nodes is also progressing smoothly.

However, a recent report from Businesskorea indicates that Samsung Foundry’s 3nm process remains significantly inferior to TSMC’s, both in yield rate and power efficiency. For those unfamiliar, higher yield rates mean more functional chips per wafer, lowering the cost per chip and increasing profitability.

As a result, it is expected that most fabless manufacturers including NVIDIA, AMD, Intel, Qualcomm, MediaTek, and Apple will continue to rely on TSMC for their 3nm chips. It has also been reported that Google will switch to TSMC for its 2025 flagship chipset Tensor G5. Notably, the Tensor G4, which will power the upcoming Pixel 9 series, will still be manufactured by Samsung Foundry and will be based on an updated 4nm process.

Samsung talked about the mass production of their 3nm chips 3 years ago. Back in June 2022, Samsung Foundry was the first in the industry to apply the 3nm gate-all-around (GAA) process to mass production. However, due to the limitations, the company is still struggling to secure customers for the end product. The first generation 3nm process node (SF3E) has performed below the company’s expectations when it comes to performance and power efficiency. It has reportedly been adopted by some niche scenarios like chips for cryptocurrency mining. However, a widespread adoption will still require more improvements in power efficiency and yield rates.

The relatively lower yield rates and power efficiency of Samsung Foundry’s process nodes have long been problematic and are considered the main reasons for the performance gap between Snapdragon (those that are fabricated by TSMC) and Exynos-powered Galaxy devices.

According to a previous report, Qualcomm is considering a dual-sourcing strategy by partnering with Samsung Foundry alongside TSMC. While this is aimed at securing the company’s business by not relying on a single manufacturer, the company’s CEO mentioned “the current focus must be on the foundry production at TSMC.”

(Source)

The post Samsung’s 3nm process continues to struggle with power efficiency and yield rate appeared first on Gizmochina.

We’ve been hearing rumors that Google will itself design the Tensor G5 chip for the Pixel 10 and 10 Pro. Unlike previous Tensor chips, which have relied on Samsung’s expertise and Foundry since its debut, G5 is said to make a switch to TSMC for its fabrication process. 

Now, the publication Android Authority has got its hands on the shipping manifest for a sample Tensor G5 chip imported into India, confirming that Google is switching to TSMC.

The document reveals several details about the chip, including Google’s part number (G313-09488-00), the product categories (Integrated Circuit, System on a Chip), and that it is a new product.

It also confirms that TSMC is the manufacturer of the SoC. The speculation that this is the Tensor G5 SoC is supported by the component’s codename, LGA, which stands for “Laguna Beach,” believed to be the codename for the fifth-generation Tensor chipset.

Another sign that TSMC is making the Tensor G5 can be seen on the document where InFO POP is listed as the packaging technology used on the chipset. InFO POP technology is exclusive to TSMC.

Additionally, the document confirms the chip passed a system-level test (SLT) simulating basic functionalities. It also mentions 16GB of POP (package-on-package) RAM, potentially hinting at similar memory options for the upcoming Pixel 10 series.

Considering this leak, if you’re planning to upgrade to a Pixel phone this year (Pixel 9 series), you might want to wait for the Pixel 10 in 2025. The Pixel 10 is likely to feature a more powerful and customized chip designed by Google, offering optimized performance for the Pixel 10 line.

Related:

• Gmail for Android Finally Implements a Crucial Feature It Lacked Since Day One
• Gmail Now Lets Users Unsubscribe Unwanted Mails With A Single Tap On iOS
• Lenovo Legion Y700 2024: Latest Gaming Tablet with Enhanced Display now available at Giztop
• REDMAGIC Magnetic VC Cooler 5 Pro now available at a discounted price at GeekWills
• HyperOS vs MIUI: 10 Things You Must Know

The post Google to source Tensor G5 chips from TSMC, leaked documents reveal appeared first on Gizmochina.

Google appears to be taking steps towards independent chip development as it’s said to be testing a “self-developed” Tensor chip in collaboration with Taiwanese company KYEC. 

KYEC provides testing services for new chips headed to manufacturing in Taiwan. Google is reportedly buying equipment for testing these chips at KYEC, with tests starting around the middle of this year and increasing over time (via 9to5Google).

This development could mean Google is moving away from using Samsung’s processors for its Pixel phones.

Sources suggest the chip in question is likely the Tensor G5, codenamed “Laguna.” And it is slated for a potential 2025 release. Likely to be manufactured on TSMC’s 3nm process, this chip is expected to prioritize artificial intelligence (AI) capabilities, aligning with Google’s focus on AI-powered features in its Pixel devices.

The switch would mark Google’s first fully custom chipset design, as the company has thus far been relying heavily on a foundation provided by Samsung’s Exynos chips.

Shifting production to TSMC could offer Google several advantages. Bypassing Samsung’s Exynos chip base, which currently lags behind Qualcomm’s Snapdragon series in terms of performance and efficiency, could give Google more control over its chip design and potentially lead to more powerful and efficient Pixel phones in the future.

However, in the immediate future, Google’s partnership with Samsung will continue. The Tensor G4, destined for the upcoming Pixel 9 series, will be built using Samsung’s 4nm process and will likely offer minor improvements over the G3 found in the Pixel 8 series.

The technology Google is using for its Tensor G4 is said to be similar to what Samsung employs in its Exynos 2400 SoC. So we can expect a decent amount of improvements in CPU and GPU gains. 

Related:

• Google updates Chrome’s incognito mode warning after privacy lawsuit
• Google is removing 17 features from Assistant in the next few weeks
• Lenovo Legion Y700 2023: Save $100 on this 8-inch gaming Android tablet
• Xiaomi 13 Ultra Premium Camera Phone is now only $799
• OnePlus partners with Pixelworks to Elevate Mobile Gaming Experience on the OnePlus 12

(Source)

The post Google Tensor G5 likely to be manufactured by TSMC appeared first on Gizmochina.

Googlе has strugglеd to kееp up with Qualcomm and Applе in tеrms of mobilе chip pеrformancе with its first thrее Tеnsor chips, which wеrе all built on Samsung’s procеss nodеs.

Howеvеr, a nеw rеport suggеsts that Googlе is taking full control of thе dеvеlopmеnt of thе Tеnsor G5 chip and may switch to TSMC’s 3nm procеss for manufacturing.

Thе nеws comеs from a X (formerly Twittеr) lеak by @OrеXda, who has a good track rеcord of accuratе information about Googlе’s hardwarе plans. @OrеXda claims that Googlе is dеvеloping thе CPU, GPU, and backеnd for thе Tеnsor G5 in-housе and that thе chip will bе manufacturеd by TSMC using its 3nm procеss.

Thе nеws comеs on thе hееls of a prеvious rеport that thе Tеnsor G5 chip will bе fully custom-dеsignеd. This mеans that Googlе will havе morе control ovеr thе dеsign of thе chip and can optimizе it for spеcific tasks, such as machinе lеarning and photography.

On thе othеr hand, thе Tеnsor G4 that will powеr thе Pixеl 9 sеriеs is еxpеctеd to dеlivеr slightly improvеd pеrformancе comparеd to its prеdеcеssor.

Thе Tеnsor G5 chip is еxpеctеd to powеr thе Pixеl 10 sеriеs, which is schеdulеd for rеlеasе in 2025. Whilе it is still two yеars away, thе nеws that Googlе is taking full control of thе dеvеlopmеnt of thе chip and may switch to TSMC’s 3nm procеss is vеry promising for Pixеl fans.

Related:

• Google Maps receives a slew of AI-powered new features
• Google Play Protect gets a boost with real-time app scanning
• Google Pixel 8 Pro places just below the iPhone 15 Pro in DxOMark camera test
• Attention Android Users: Don’t Buy This Phone Unless You Want to Use It Forever
• Best Pixel 8 Pro accessories 2023

(Via)

The post Google’s Tensor G5 to feature custom CPU and GPU, but G4 to be a minor upgrade appeared first on Gizmochina.

Google is currently working on its next-generation Tensor G4 chip, which is intended for use in the upcoming Pixel 9 series smartphones. These smartphones are expected to launch in 2024.

Google’s Tensor G4 may have minor upgrades over Tensor G3

Google is collaborating with Samsung’s System LSI department to develop the Tensor G4 chip, codenamed Zuma Pro. According to a report by Android Authority, this chip is closely based on Samsung’s Exynos 2400 and is expected to deliver slightly improved performance compared to its predecessor, the nine-core Tensor G3 chip.

The Tensor G3 chip codenamed Zuma is being developed with input from Samsung’s Exynos 2300 chip, which has not been released yet. During the development phase, the G3 chip is being tested on a “Ripcurrent” board, while the G4 chip utilizes a “Ripcurrent Pro” board.

In 2025, Google aims to launch the Pixel 10 smartphone featuring the fully custom-designed Tensor G5 chip. A significant shift in the manufacturing process involves moving from Samsung to TSMC for producing the Tensor G5 chip. This marks Google’s ambitious step toward creating a unique chip for its future devices.

According to past reports, Google is set to release the Pixel 8 and Pixel 8 Pro smartphones, featuring the Tensor G3 chip, on October 4, 2023. These phones will boast improved designs and upgraded cameras.

Both models will include larger ISOCELL GN2 main camera sensors. Additionally, the Pixel 8 Pro will have a 64MP ultra-wide-angle camera, while the Pixel 8 will feature a 12MP ultra-wide-angle camera. The Pixel 8 Pro is reported to include a 48MP telephoto camera with 5x optical zoom capabilities.

RELATED:

• Pixel 8’s Tensor 3 chip to run a lot cooler than its predecessor: Report
• Google Pixel 8a (Akita) allegedly spotted on Geekbench with underclocked Tensor G3
• Google will Refresh the Camera App UI for the Android 14 Launch
• Google Pixel 8 and Pixel 8 Pro official European pricing leaked
• iPhone 15 Pro Max vs Samsung Galaxy S23 Ultra: Specs Comparison
• iPhone 15 Pro vs iPhone 15: Specs Comparison

(Via)

The post Google Tensor G4 will come with nominal upgrades, Tensor G5 to be entirely custom-made appeared first on Gizmochina.

Following in Apple’s footsteps, Google embarked on a journey to use its own custom chip in Pixel smartphones. The Pixel 6 series became the first to feature a custom Google processor called Tensor, albeit based on Samsung’s Exynos chips manufactured on the 5nm node process. The Pixel 7 continued to utilize a semi-custom Samsung-made processor, and so will the Pixel 8. But it seems Google is planning to drop the Korean company in favor of TSMC to launch its fully customized Tensor G5 chip in 2025.

A truly custom Tensor chip coming in 2025

A former Google executive told The Information (via 9to5Google) that the initial plan was to launch a truly custom Tensor chip as early as next year. However, the production of the chip, codenamed Redondo, got delayed due to challenges in retaining employees and coordinating development across multiple teams. According to the outlet, the chip is now expected to be ready by 2025.

Confirming previous reports, today’s source claims that Google is indeed switching from Samsung to Taiwan Semiconductor Manufacturing Company (TSMC) for the production of its Tensor G5 chip. The chip is said to be based on TSMC’s 3-nanometer manufacturing process and will utilize Integrated Fan-Out technology for reduced thickness and increased power efficiency.

We will have to wait a little longer before witnessing a fully customized Tensor chip. The upcoming Pixel 8 lineup for this year will continue to feature a processor designed and manufactured by Samsung. Similarly, the Pixel smartphones set to launch in 2024 will also incorporate a processor from the Korean company, albeit with fewer Samsung components this time.

A fully custom chip will help Google to create a tighter integration between hardware and software, making the phones perform much better than they do now.

RELATED:

• Google Pixel Watch receives July 2023 security update
• Google Pixel and Pixel Tablet receive July 2023 Android security update
• Google Pixel 8 Pro live images leaked, reveals key specs
• Best Foldables of H1 2023 (Fold & Flip)

(Source) (Via)

The post Google ditches Samsung for TSMC to produce first custom Pixel chip, Tensor G5: Report appeared first on Gizmochina.