Recently, the clouds seem to have become a hot topic.
Many people have been teased by the myth of Musk's bull "X" company going down to the cloud to save money.
So, does Shimomo really save money?I believe you're a ghost!
In order to settle this account, you need to clarify a concept first
Going to the cloud later, on-premisesSelf-built cloud servicescost, andBuy your own hardwareThe cost is not Apple-to-Apple's relationship.
And many people who call for the cloud to fool people, when calculating **, often confuse the two, and the result is that it is far lower than renting a public cloud.
In fact, the cost of local self-built cloud services includes not only hardware procurement such as computing, storage, and networking, but also software licensing, bandwidth, O&M, and technical support. as well as the cost of renting and even building land in the computer room.
Moreover, the situation is different for enterprises of different sizes.
In order to survive in the start-up period, enterprises need to achieve high-speed innovation with minimal investment.
If you put your early energy into infrastructure procurement, operation and maintenance, and management, it will slow down the efficiency of R&D and die at the starting line.
The cloud is the best vehicle for high-speed innovation.
Startups can have the latest technology, the highest deployment efficiency, and shorten the development cycle on the cloud.
Baichuan Intelligence, which was just established in April, relies on the computing power and rapid delivery of the cloud to liberate its energy and focus on algorithm iteration. Fast verification, low-cost trial and error, and six industry-leading open source models were continuously released within half a year.
In a person's life, there are only a few opportunities to change their fate, and the same is true for enterprises.
Seize the opportunity to go further, and you can take a leap forward.
But opportunities often come in the form of traffic spikes.
It could be a sudden explosion of a game, or it could be an influx of users.
The technical capabilities and resource reserves on the cloud can help enterprises better support their foundations.
There's a long list of things to consider when going to the cloud.
Just buying a server is a technical job.
These problems are easy for even cloud vendors to step on:
Buy cost-effective resources, such as new-generation, high-performance processors
Parts fluctuate frequently, preferably when the price is low** memory and SSD;
It is necessary to adjust the scale of procurement according to the changes in demand to ensure a high utilization rate of resources
Moreover, what you buy is not the same as what you use.
Behind every resource actually used, there is inevitably a corresponding cost of empty consumption.
If you choose to build your own IDC, you will have to bear this part of the cost.
Software research and development is also an indispensable homework.
Although there is an open source version in the industry, if you want to make the performance stronger, you still need to do a lot of self-research.
After all, even if the performance of a single item is just a little worse, the final performance and cost will not be ideal.
In order to create a competitive architecture, cloud vendors often gather top technology experts and work hard for years of research and development.
With these efforts, there are continuous iterations of technologies and products on the cloud to help enterprises improve their competitiveness.
On the cloud, the time and cost of new services can be reduced by at least 20%.
I don't think about innovation, I just want to do a good job in O&M, okay?
Even for patching, system upgrades, and security management, cloud vendors have pulled a large team.
Now, the cloud not only has a more reliable SLA, but also has high availability + high stability, while the self-built cloud may have to "put out fires" every day.
So, do you want to go to the cloud or build your own machine?
Let's take the goose factory as an example.
- WeChat** account, based on the cloud family bucket technology stack.
This year, the number of users who watched the Spring Festival Gala on the ** number reached 1900 million. Nearly 100 tons of Chinese New Year's Eve traffic peak, public cloud steadily withheld.
- If there is no Tencent Cloud, would there be any success of Tencent Meeting?
In the early days, relying on technologies such as audio on the cloud, Tencent Meeting was quickly launched
In the face of the peak traffic of the epidemic, Tencent Meeting put all nodes on the cloud, scaling up to 1 million cores in 8 days, and increasing the daily activity by 200 times in 2 months
Today, with more than 400 million users, Tencent Meeting continues to reduce cloud operating costs through cloud-native transformation.
Of course, not only goose factories, but also large companies have countless cases of innovation and success on the cloud. Cloud computing is a typical idea that relies on large-scale industrial production to maintain quality and reduce marginal costs.
In other words, cloud computing has been promoted in China for more than 10 years, and the call to action of "empowering intelligence with data" has been put forward for 3 years
So, if today, you send a message to the "Three-Body Man".
Then the reply must be
Don't go down to the clouds!Don't go down to the clouds!Don't go down to the clouds!