130: New M chip

Building an android tablet ecosystem is no less difficult than establishing an ecosystem for a self-developed OS, and it cannot be rushed; it must be done step by step.

Currently, Red Star is the company with the most smart devices in the country, arguably unrivaled, and the ecological interaction in this area can be perfectly ported to tablets.

Secondly, after the release of the red star bracelet, the Starlink Interconnection technology provided by the system will allow for a perfect realization of the interconnection between phones and tablets once the laboratory fully masters the technology.

Of course, the first step now is to refine the system and make the performance powerful enough.

Chen Chen is very confident in the performance of the processor chip provided by the system. At the same time, due to improvements and breakthroughs in cooling technology, the performance scheduling for the new tablet will be tuned quite aggressively.

The stronger the performance of a good tablet product, the more popular it will be with consumers, because tablets aren't replaced as frequently as phones.

Users prefer tablets that can last a long time rather than ones that become obsolete just a year or two after purchase.

As a leader in the global semiconductor industry, TSMC was already capable of producing chips with a 20nm process in 2014. As long as a processor using the new 20nm process doesn't suffer from forced core stacking or design issues, its performance will naturally see a significant boost compared to 28nm chips.

In the 20nm era, Qualcomm's aggressive designs, forced core stacking, and the introduction of 64-bit threads caused the Qualcomm snapdragon series of that period to become "fire dragons" one after another. Can you say the performance of the snapdragon 810 wasn't strong?

With water cooling and semiconductor cooling, the performance of this processor is quite formidable when fully unleashed.

But no matter how powerful a processor is, it has no room to perform within the cramped internal space of a mobile phone.

Other manufacturers cannot obtain the smaller Yangtze River Battery and thus cannot save space for cooling modules like Red Star does, which dooms them to poor thermal management in their phone designs.

Furthermore, the most important point is that other manufacturers wanting to use the same cooling modules as Red Star have no other suppliers; currently, Telong is the only domestic supplier of copper tube cooling modules, and it is a supplier Red Star invested in.

Although called a jointly developed cooling technology, it's actually Red Star's laboratory providing the technology, with the supplier merely improving the manufacturing process. If other phone manufacturers want to purchase these modules, they have to ask if Red Star is even willing to sell!

All this is to say that because tablets have more space, they allow processors to unleash more performance. Therefore, Red Star is particularly aggressive with tablet processors, pursuing even greater performance.

"It's time to spin off the semiconductor department."

Red Star's semiconductor department now has over 800 Engineers and can design various types of chips, such as the control IC on Red Star's fast chargers, which was designed by the Hongxing Semiconductor department.

Spinning off the semiconductor department like Huawei's HiSilicon to establish a specialized subsidiary would make flat management easier, rather than relying entirely on Red Star Company as it does now.

Of course, even if spun off, selling lingxiao chips to other manufacturers is unrealistic; like Huawei, they can only be produced for internal use.

The main reason is communication patents. Other manufacturers using Qualcomm chips only need to pay Qualcomm. If they used Huawei or Red Star chips, they would have to pay for the chips plus an additional patent fee, making it less cost-effective than buying directly from Qualcomm.

"Dr. Lu, you're going to release three models of Lingxiao processors all at once this year?"

Arriving at the semiconductor department with Shen Rou, Chen Chen asked impatiently as soon as they sat down in Dr. Lu's office.

It's not that Chen Chen doesn't care about the semiconductor department; in fact, apart from providing funding and resources, he doesn't manage it at all. For researchers, no management is the best management.

"Yes, the processor and soaring c1 designs are complete. Both processors can be used in our phones. Little Li is currently at TSMC overseeing the tape-out process, and I believe there will be good news soon."

It was clear that Dr. Lu was in high spirits. After Chen Chen provided the 20nm chip blueprints from the system last year, Dr. Lu's team created some variant designs based on them, including the M-series large chip that is currently unfinished.

As for where his boss got the chip design blueprints, Dr. Lu tacitly remained silent. Whether the boss stole them from a USA lab or picked them up in Zhongguancun or Huaqiangbei had nothing to do with him.

Dr. Lu knew not to ask too many questions; knowing too much wouldn't do any good.

"And the other one?"

Speaking of this, Dr. Lu grew even more excited, gesturing with his hands: "I plan to name this chip the M-series. Unlike the high-performance S-series and the mid-to-low performance C-series, I call this one 'Super Performance'."

"First, its size is nearly double that of the S-series, allowing for more transistors. We've also integrated the Kaitian baseband for the first time. It has 8 main cores—6 large and 2 small—and 48 GPU cores. Its graphics performance is a 130% improvement over our soc level!"

"Most importantly, we've also integrated a separate IPS chip into this processor for the first time—not just an IPS module, but a brand-new chip."

"This way, our processor will have more computing power to handle photography and images!"

Chen Chen nodded. Anything from the system was bound to be high quality; it was only natural for this M-class chip, built on the system's designs, to be more powerful.

Compared to the extreme-performance M-series, Chen Chen was more concerned with the mid-to-low-end C-series. This was the bulk of Red Star's shipments. For the foreseeable future, there would be only one choice for mid-range chips: Qualcomm's 6-series processors.

Among them, the legendary Qualcomm snapdragon 625 is a household name. No one knows how many models used this chip, nor how many Qualcomm actually produced, but Chen Chen knew that major phone manufacturers used it for three whole years!

Currently, Red Star's mid-range phones use the Lingxiao S2 processor. This outdated flagship processor is merely adequate. Unlike Qualcomm's strict hierarchy—where Qualcomm wouldn't allow manufacturers to use flagship processors in mid-to-low-end phones, even outdated ones—Red Star can trickle down its own processors as it pleases.

The C-series processor project began early last year. After finishing simulation tests by the end of the year, it was sent to TSMC for tape-out in January.

Mass production will have to wait until mid-year.

"The soaring c1 processor targets Qualcomm's 6-series. Compared to Qualcomm's snapdragon 600, we've improved performance by 17%, but our graphics processing performance still doesn't match Qualcomm's level."

"How big is the gap?" Chen Chen asked.