370: Chapter 370 The Panic of Taikun Electric

"Do you think this will work?"

Liu Jun deliberated repeatedly before finally sighing.

"Whether it works or not, it has to. We have no other path. If we use Qualcomm, in the near future, we will become the next Motorola or Nokia. Using Transsion's chips at least allows us to stay at the table. Besides, while the price of $160 is a bit expensive, we can maintain our own pricing. As long as we can hold onto the high-end market in the United States, negotiations with the channels will be easier."

"What about ZUK? Should we still proceed with our ZUK?"

This appeared to be a question, but the answer was already quite clear.

"Transsion says their production capacity is insufficient. If we launch the R&D for two new models simultaneously... Mr. Yang, the market is too volatile. Should we really be using the highest technology?"

The person on the other end of the phone nodded. This was actually a good idea; after all, it could save a significant amount in R&D costs. If Motorola's sales were good, then...

"Then let's secure the Motorola one first and discuss ZUK later. But you should come back and consult with Mr. Liu first."

After hanging up the phone, Liu Jun and the others immediately set off.

On the way, Liu Jun remained silent. He recalled a few years ago when Lenovo acquired IBM's PC business; the whole world said Lenovo was crazy. After all, it was a piece of junk.

But it was the reputation they lacked, and that deal became the most successful case of internationalization for a Chinese enterprise.

He thought acquiring Motorola would become another legend. This time, many media outlets had changed their views, saying they would win, but...

It turned into such a mess.

...

A week later, Lenovo formally replied to Transsion:

"We confirm the purchase of the first batch of 500,000 Taishan 710 chips at a unit price of $180, totaling $90 million. However, the attached condition is that Transsion must assist Motorola in completing the motherboard design and driver adaptation for the Taishan 710, and provide a full suite of development tools and technical support."

Liu Fang finished reading the fax and handed the paper to Huang Tianya.

Huang Tianya glanced at it and was instantly dumbfounded. Had they taken the wrong medicine? They wanted the full suite of development tools? How could they be so shameless?

However, his finger paused on the quantity.

"500,000 units. Our current forecast for Motorola's flagship sales next year is only 5 million. This is just one-tenth of our capacity, but it's still petty."

Motorola's sales had plummeted from 30 million units in the single United States market during the Transsion era to just one million. If they didn't figure things out, next year might really only see 5 million units... or even worse.

"Do you think we should accept?" Liu Fang's eyes suddenly darted, and she placed the report back on the table.

"Accept, why not?" Huang Tianya agreed very readily.

"At a price of $180, we have a 240% gross margin, and the cost of technical support isn't high; our Taishan 710 SDK is already very mature. Furthermore, once Motorola uses our chips and the high-end market shows improvement, they will never be able to go back to Qualcomm. We can also provide idiot-proof training; as long as we ruin their engineers, the cost of platform migration will be too high, and they will be locked in by us."

"Aren't you afraid that they'll use the Taishan 710 to turn around and negotiate lower prices with Qualcomm?"

Huang Tianya shook his head, stood up quickly, and apologized with a smile:

"Big Sister Liu, stop messing with me! You know exactly what Qualcomm is like. Besides, given Motorola's current scale, Qualcomm really isn't worth bowing down to. What they truly care about are Lenovo's competitors. When they see Motorola using 14nm chips—even if the cost is a bit higher, the market response is supportive—what will they think?"

"Alright, I won't tease you anymore. However, the priority needs to be changed. Their own products definitely come first, Honor comes second, and Motorola and other clients come third."

Huang Tianya organized the report and sent it to Liu Jun. The reply from the other side came quickly...

After all, no matter how good a partner is, they can't compare to one's own products. This was the biggest drawback of choosing Transsion, but who let them hold the technical authority... There was no comparison! They just had to accept it...

...

Meanwhile, TSMC was also very worried. Although they didn't understand why their Quadruple Patterning yield rate was so high, the 30-minute news had already reported it, and Transsion was bound to have a major move next.

However, Morris Chang comforted himself; even if the other party used a lower-end process, increasing production capacity wasn't that easy, and they still had a chance...

But on September 1st, Transsion Semiconductor directly announced the opening of its 22nm foundry services. This time, there was no official announcement; they only notified clients who had already placed orders and major mobile phone manufacturers.

Otherwise, they simply added a line of text to the foundry service page on Transsion's official website:

"22nm process, opening in Q4 2013. $1,900 per wafer. We promise a yield rate of no less than 80%; for anything below 80%, we will not charge for the damaged portion."

Although this news wasn't officially announced, the fans couldn't hold back. It was first discovered by a digital enthusiast who was looking for news on new products on Transsion's official website. He casually refreshed the Transsion Semiconductor page... and then became excited. Sure enough, 100% pure domestic 22nm had arrived.

#Transsion Opens 22nm Process#

#28nm Global Leader?#

Although 22nm couldn't compare to Transsion's own unattainable 14nm "alien technology," it was a full generation ahead of the 28nm used by the vast majority of current chips.

Power consumption was reduced by at least 30%, performance increased by 20%, and the area was reduced by 25%. This was the intergenerational gap written in textbooks.

And the most critical factor was the price. TSMC's current foundry price for a 12-inch 28nm wafer was approximately $1,200, while Transsion's 22nm 12-inch wafer foundry price was only $1,900.

Compared to the outrageous quotes for 14nm, this price seemed to have excellent cost-performance. Furthermore, if one factored in the cost per chip brought about by the yield rate and area reduction, the actual cost of each chip was almost on par with or even slightly lower than TSMC's. Because the area of a 22nm chip was smaller than a 28nm one, one wafer could yield about one-third more...

Buying a better process for a slightly higher price—it seemed... the spring of domestic semiconductors was coming? Would there be a hundred flowers blooming next?

If the Taishan 420 chip were made using 22nm, the area would likely drop from 88 square millimeters to 66 square millimeters. One wafer could yield about 880 units.

At $1,900, the cost per chip would be less than $2.16; his classic slingshot price was about the same.

After calculating, many people realized that Transsion was truly conscientious... They simply couldn't believe the result.