The black technology empire started with a debt of one billion

Chapter 181 177 New Memory Technology and Chip Superposition
"I just sent them back to the airport, I will send you my link later." Comrade Xiao Wang hung up the phone after finishing speaking.

[Crazy Life: You didn’t tell me when you left? 】

After waiting for 3 minutes, the other party did not reply with any message.

"Drip drop~"

Just as Cao Mang was about to go back to the lounge to change clothes, he received a message from another person.

The other party asked him to wait in the industrial park.

Tonight, the other party wants to try the feeling of doing that kind of thing leaning on the floor-to-ceiling windows.

Cao Mang shook his head helplessly after seeing this message.

It seems that the other party has a big addiction to food.

Before Cao Mang changed his clothes, he sent a message to Li Yujing asking him to troubleshoot the foundry.

Now that everyone has signed the contract.

Compensation should still be compensated for leaked information.

Huaxia Android Alliance chat group.

Rebs: "What do you think about the new Deep Blue model exposed online?"

Shen Wei: "From the design point of view, it really looks like the design of Deep Blue, but I think things are not that simple."

Chen Yongming: "Let's wait for Shenlan to say something about this matter. I also think this matter is not that simple."

Both of them felt that Cao Mang must have hidden his backhand somewhere.

If the new machine of Deep Blue is like this.

There is no need for Cao Mang to tell the outside world that he is more than 20 years ahead.

Cao Mang sent a Weibo post after returning to the front office from the lounge.

“The mobile phone forwarded online is indeed a mobile phone that is about to be released by Deep Blue Technology.

I have already expected that the new phone will be stolen before it is released.

After all, money can turn ghosts around.

So I kept a small hand and hid the key things.

As for what the new dark blue mobile phone will look like.

Everyone will know it on the day of the press conference.

I believe that the new product of Deep Blue No. 1 will satisfy everyone. "

After Cao Mang finished posting on Weibo, he checked the system panel.

[Host: Cao Mang]

[Reputation value: [-] million (rounded to an integer of [-] million)]

The leak of the dark blue mobile phone this time is a good thing for Cao Mang, because this leak has given him a lot of prestige.

This incident not only attracted the attention of domestic netizens, but also attracted the attention of countless foreign netizens who care about Deep Blue.

This gave Cao Mang another reputation of nearly [-] million.

Cao Mang purchased two or three products on the system mall.

The first product is HMBSR memory technology. HMBSR is an evolved version of HBM memory technology.

When HBM (High Bandwidth Memory) exists as a type of GPU memory, it seems to be not uncommon now.

Many people probably know that HBM is expensive, so it is only seen on high-end products, if not rare.

One of the characteristics of HBM is that it achieves higher transmission bandwidth with smaller size and higher efficiency (partially) than DDR/GDDR.

From the perspective of transmission bit width, each layer of DRAM die has two 2bit channels, and the HBM memory with 128 layers of DRAM die height has a total bit width of 4bit.

Many GPUs and CPUs have four such HBM memories around them, so the total bit width is 4 bits.

It can be seen that HBM has ultra-high throughput in terms of bandwidth.

If this kind of memory is used on mobile phones, it will definitely bring better efficiency to customers.

Just why isn't anyone using it on phones?
Is it because it is expensive that it is not distributed to the consumer market?
No matter how expensive HBM memory is, it cannot be so expensive. Consumers can afford the cost of hundreds of dollars.

Consumer: Are we the ones who are short of those few hundred dollars?
Actually it is not.

There are three problems in it.

The first one is the cost issue. Although this is a small issue, manufacturers have to pay attention to it.

The second problem is that the CPU of a technical mobile phone does not use such a high-bandwidth memory at all.

The third problem is that the latency of HBM memory is too high.

Because of these three factors, manufacturers have to give up HBM memory.

The HMBSR memory purchased by Cao Mang adopts a brand-new design, which solves all the current problems of HBM memory.

Lower manufacturing costs and lower read and write delays.

The second product Cao Mang purchased was kos storage memory technology.

Currently, the two types of memory used on the market are emmc and usf.

eMMC is currently the most popular portable mobile product solution, the purpose is to simplify the design of terminal product memory.

Since different brands of NAND Flash chips include Samsung, Toshiba, Hynix, Micron, etc., when the design manufacturers import them, they need to redesign according to the products and technical characteristics of each company. In the past, there was no one technology that could be used universally. The brand's NAND Flash chip.

The only advantage of this kind of memory is that it is cheap, and its disadvantage is that the reading and writing efficiency is relatively low.

The highest reading efficiency of this kind of memory so far is 400mb/s, and the writing efficiency is 70mb/s. (Specifically referring to 2015.)
usf memory is the joint committee of electronic equipment engineering in 2011, which released the first generation of universal flash memory storage standard, the predecessor of UFS 2.0.

However, the first generation of UFS was not popular and did not have a significant impact on the eMMC standard.

In 2013, JEDEC released UFS 9, a new-generation flash memory storage standard, in September of that year.

The reading and writing speed of UFS 2.0 flash memory can theoretically reach 1400MB/s, not only has a huge advantage over eMMC, but it can even dwarf the SSD used on computers.

As a result, it gradually replaced eMMC in the high-end device market and became the mainstream standard for mobile devices.

In fact, there are two versions of UFS 2.0, one of which is HS-G2, which is the current UFS 2.0.

However, the other version is HS-G3, which can be called UFS 2.1, and its data reading speed will soar to 1.5G/s, which is twice that of UFS 2.0.

The best read and write efficiency of kos storage memory can reach 30G/s, and it also has higher lifespan and capacity.

(The efficiency depends on the design. Cao Mang purchased the first-generation kos memory technology, and the read-write efficiency of the first-generation technology is 5.0g/s)

The last technology Cao Mang purchased was chip stacking technology.

This technology is actually not advanced technology.

Warwick is not the only one that has researched this technology. Intel and AMD had this technology as early as five or six years ago.

It's just that the cpus produced by these two superpositions had to give up this technology due to high power consumption.

Can the chip superposition technology purchased by Cao Mang solve the problem of high power consumption?
You may say you can or you may say you cannot.

Because energy conservation is an immutable thing, the power consumption of two chips is bound to increase.

The technology Cao Mang purchased only reduced the power consumption of some chips after stacking.

This technology reduces power consumption by about 48%.

With this technology, 14nm beating 7nm is not a big deal.

The only downside is power consumption.

Cao Mang's solution to this problem is to stack batteries.

Significantly increases battery capacity.

It is much simpler to increase battery capacity than developing a lithography machine.

This is also the simplest and fastest way to solve chip bans.

It is also the cheapest solution.

(End of this chapter)