In today’s smartphone application processor market, the head manufacturers are MediaTek, Qualcomm, and Apple triad. Counting the entry-level machines that love to use the ZiGuang ZhanRui, the declining Samsung Exynos, is barely together with the five giants. If not for some complex external factors, Huawei Heidi should have been here to reign supreme. A few years further back, there are still many old players doing cell phone processors, such as Nvidia, Intel, Texas Instruments, etc. That was the boom period for smartphones.
During this period, there was also a big factory that also devoted itself to developing its own chips and wanted to get a share of the pie from the big factories, that is LG.
At the end of 2014, LG launched the first-generation self-developed chip LG NUCLUN, which was based on TSMC’s 28nm process technology and used the popular big.LITTLE size core architecture at that time. The CPU adopts an eight-core architecture, including four 1.5GHz Cortex-A15 large cores and four 1.2GHz Cortex-A7 small cores; the GPU is Imagination’s PowerVR G6430, the same as the Apple A7 the previous year.
LG NUCLUN’s on-paper parameters are close to Huawei Kirin’s, and comparable to the Exynos 5410 on Samsung S4 the year before, making it a qualified mid-range, mid-to-high-end chip.
The LG NUCLUN debuted with the LG G3 Screen phone, which was sold only in the Korean market as a test, similar to the strategy of Samsung Exynos chips in recent years.
Jong-Seok Park, head of LG Mobile, is confident about its own chip: “NUCLUN opens a new chapter in LG’s history of innovation in the mobile industry. With this in-house solution, we will be able to achieve better vertical integration and further diversify our product strategy to address fiercer competition.”
The LG G3 Screen didn’t make much of a splash in the market and soon disappeared into obscurity.
A year has passed, and rumors of the LG NUCLUN second generation finally surfaced on the internet. Compared with its predecessor, LG NUCLUN 2 specifications have been significantly improved, with the processor architecture upgraded from 32-bit to 64-bit and the core cluster upgraded to A72 large core with A53 small core. It will also use the mainstream TSMC 16nm process, performance, and energy efficiency than the previous generation of 28nm greatly improved.
Not only that, but the industry has also spread the news that LG is working with Intel to create another version of NUCLUN 2, based on 14nm FinFET process development – at that time Intel’s 14nm is still a positive term, no less than Samsung, TSMC’s 12nm, basically the symbol of the industry’s most advanced process.
Is LG to turn the tide? Fate and LG played a joke.
First, at the end of 2015, there is news that LG canceled the Intel version of the chip plan because Intel 14nm can not give enough capacity.
Then in mid-2016, LG NUCLUN 2 scores were revealed online, and the performance was eye-opening. Its Geekbench single-core score of 1302 and multi-core score of 3278 are only comparable to Qualcomm’s mid-range processor Snapdragon 625.
There is also a bit more incredible, earlier rumors that NUCLUN 2 will use an eight-core processor, and run scores show that this processor only has six cores.
At least the chip was made, so the next logical step would be to release it to the market, right?
The LG V20 was released in September 2016, and instead of using its own chip, it was on the last train of the Snapdragon 820. Perhaps LG itself also realized that with such an immature chip on the flagship phone, the consequences may be unimaginable.
After falling twice, LG still seems to be upset and ready to do a big job, directly over the current flagship chip.
Before and after the release of the LG V20, new information pointed out that LG’s future NUCLUN chip will use Intel’s 10nm process, equipped with the latest A75 and A55 CPU cores. You know, Qualcomm’s next-generation flagship Snapdragon 835 is prepared using Samsung’s 10nm process, but Intel’s 10nm process is more advanced, and the density is comparable to TSMC and Samsung’s 7nm process.
However, this time, Intel “pigeon” LG, but also “pigeon” themselves. Because the process is difficult to produce, Intel 10nm process jumped continuously.
Originally planned for the 2015 debut of Intel 10nm, first pigeon to 2019, and then once jumped to the end of 2021 – now the 12th generation Core. During this period, Intel had launched the initial generation of 10nm process processor i3-8121U, but the performance was bad, and the plan to launch 10nm had to be delayed again and again.
The third generation of NUCLUN was sunk in the sea.
So, LG’s faltering self-developed chips are all thanks to Benze, no, all thanks to Intel? From the perspective of hindsight, in addition to Intel’s external conditions, LG’s own factors can not be ignored.
First of all, LG started late in the field of chips, technical strength is too much behind. The company’s talent base comes from the two old chip companies PA Semi and Intrinsity. it can be said that LG is too far behind in terms of technology, talent, and patent accumulation in its own chip development.
Second, LG is not fully committed to chip development. LG’s in-house chip has only appeared in one mass-production model since the beginning, and it has no competition with the rest of the world.
Huawei’s Kirin 920 series chip from 2014 is closer to LG’s initial NUCLUN in the same year. But Huawei has finally caught up with and even surpassed Qualcomm through continuous design improvements. If not for the sanctions, Huawei phones and Kirin chips would have been able to make a big splash.
LG’s uncontested self-made chip also ended up in the same way as its smartphone business, leaving a sigh of relief.
