Nvidia revised a "disclosure notice," which immediately irritated users in the circle. However, although this matter has little to do with the average general user, the most important is still some commercial users.
However, with the exposure of some corporate users, media forwarding has really become a hot issue in these two days. Nvidia did not expect himself to think that he was merely a “little trick†and it caused an uproar.
Recently, Qing Shuiliang, president and CEO of Ubiquitous Entertainment in Japan, published an article saying that Nvidia updated its End User License Agreement (EULA) to prohibit users from deploying GeForce companion software in data centers.
The GeForce companion software is necessary for GeForce to use its own hardware capabilities. Many people in the industry use the graphics card provided by Nvidia, but also because its supporting software is very good.
Qing Shuiliang proposed in the article that in today's practice, Nvidia can be said to be the only semiconductor company in the world that provides APIs and enough computing functions for deep learning. The updated EULA of Nvidia still allows you to purchase GeForce graphics cards for use in data centers. However, you cannot deploy GeForce software for deep learning in the data center.
This new ban means that you bought Nvidia's GeForce graphics card, but its software has been forced to "castrate" and you cannot use deep learning in the data center.
For a time, some companies and individuals who have conducted deep learning studies have come forward with their responsibilities, especially for those who are limited to cost and want to use deep learning to conduct research. This practice even limits the “deep learning depthâ€. .
However, according to several media reports. Although Nvidia's amendment to the user agreement has just started because of this article, in fact, Nvidia has long been directly under the "ban" in the circle of corporate users, and has issued written agreements to major corporate users.
For example, S&T understands that this operation was launched almost two months ago. All NVIDIA partners have received notifications from NVIDIA and are ready to start operations.
For example, on December 21, the famous Japanese cloud service provider SAKURA announced that it would stop providing its GeForce Titan X-based host because it received a written notice from Nvidia (Japan) requesting that it stop using the Geforce series.
Of course, the issue is not simply that "business users do not use deep learning." Now that public opinion has gradually taken shape - isn't it because now the entire market has been dominated by itself, and it has begun to adopt various measures? Possible gains from more profits, even at the expense of monopoly. Does it deprive some people of their "deep learning democracy"?
After all, it's still a question of profit
Nvidia through a "user agreement" consent button, I am afraid to think of the direct purpose of achieving, or related to profits.
For example, the Forcedly Modified User License Agreement (EULA) in the Nvidia Group prohibits the use of the consumer-grade graphics card GeForce in the data center for deep learning. Instead, use the high-end processor Tesla series.
According to media exposure, between consumer GeForce and recommended Tesla. In terms of performance, GeForce graphics cards have more performance cards than Tesla, and GeForce positioning is lower than Tesla, but the features it provides are sufficient for deep learning and cost-effective.
In the field of deep learning, GeForce is a god card and street card. It is also said that the two products are actually similar in performance, and even in some of the data GeForce outstanding performance, but the only problem is that Tesla price is ten times that of the GeForce graphics card... So, it is estimated that no one can afford the Tesla graphics card, especially It is in the area just getting started.
For example, the price of GeForce 1080 Ti is actually around 699 US dollars; while the price of Tesla K80 is directly to 3399 US dollars, Tesla P100 is up to 5150 US dollars.
But the new "user agreement" is such a regulation - you ah, you are the data center, you have the money, you can not save money, you have to get more expensive, and now you have no choice.
In short, in order to achieve the goal of selling more Tesla, the ban began to limit the restrictions on server companies and system integrators could no longer launch server platform products built on low-cost GeForce graphics cards. Nvidia uses this to increase the profit of Tesla's business.
However, despite this, Nvidia is not so radical - Nvidia is currently only restricting the use of GeForce for deep learning in the data center. It does not restrict the sale of graphics entities, but the general personal, research non-commercial users, and Will not be forcibly deactivated.
However, the agreement has already been issued. If you agree to it, you can't use it. If you don't agree, you can't use it. Don't worry about any specific order. No matter whether it's 3721 or Nvidia, is it compulsory for you to pay? Do you feel that you are leading the market?
Separate product lines to achieve a sustainable profitability tradition
Nvidia's exposure in the past two years has focused on areas such as AI, autonomous driving, and deep learning. But in fact, as early as the video card period, Nvidia also emphasized the introduction of different levels of products, emphasizing its exclusive positioning in their respective fields in order to achieve more profits.
Although it is a GPU computing technology company, it offers different product lines based on different performance and features. With different technical trends, there are "eating chicken cards," mining cards, server cards and even deep learning cards on the market.
However, in the previous product performance segmentation, it was feared that there should be no such products with a central location and a vague positioning.
It is precisely because of this that more and more people choose the price and performance of the GeForce graphics card for a large number of deep-learning development work. The entanglement is that, for its customers, shipping a large number of GeForce graphics cards is actually very good for Nvidia and its partners, such as MSI, Colorful, and so on.
The ban on Nvidia is actually more of a "hidden teammate," which is highly likely to reduce the shipments of its partners.
Of course, blocking the shipments of the board manufacturers, customers who provide cloud computing system integrators have even been washed. Nvidia’s ban may have hurt many customers in the field, such as board manufacturers, cloud computing service providers, and even infiltrating the entire deep learning industry.
Nvidia does not hesitate to offend so many customers, but also enforce this measure, just for money? Is it monopoly?
Nvidia's concerns
Even though Nvidia is mentioned today, its symbols are more of the infrastructure providers for cutting-edge technologies such as “mine miningâ€, “autopilot†and “deep learningâ€.
It is precisely because of this that Nvidia can go higher in front of today’s demanding Wall Street. However, in reality, the business that drives Nvidia’s major profitability is still based on PC graphics.
In the past two years, bitcoin mining has continued to be hot, and the game of chicken consumption has also led to an increase in the graphics card business. For AI and autopilot business that are suitable for long-term investment, although Huang Renxun frequently appears in these technologies, it is still insufficient. In order to constitute the main cause of financial results in the red.
In Nvidia's financial results, the game business is still the largest source of revenue for Nvidia. Despite many times in the industry slamming the PC market, Huang Renxun also made an early remark about the transformation of artificial intelligence, but PC games are still a strong sector.
According to the Q3 quarter of 2018, Nvidia’s game chip revenue was US$1.56 billion, an increase of 25.5%, accounting for 60% of overall revenue. The data center business reached US$501 million, and the automotive business revenue was US$1.44, which increased to 13.3%.
However, although the game business is gratifying, the industry has also questioned Nvidia for the soaring valuation and Huang Renxun’s active reduction.
For Nvidia, securing the old profitable business and supporting the growth of the new AI business is now a new discipline. After all, blocking the shipment of GeForce graphics cards actually blocked the revenue of the game business and ensured the support of the new AI business. The increase in new business revenues was also blocking the benefits of the old business.
Nvidia estimates that he has long wanted to understand that it does not hesitate to use this method of undermining deep learning democracy to maximize the actual income of the AI ​​business. After all, this is only visible.
Some media also mentioned a lot of grounding gas. For example, using a consumer-grade graphics card with a design life of about 5 years to use it in a data center can easily result in rework. This is in fact a side-point increase in the operating costs of the consumer-grade graphics card business.
Bang xxl is a popular Disposable Vape. It's 2000 puffs with 850mah battery build in, 3.5ml e-liquid and 50mg nic salt strength, the package is 10pcs per pack. Each pack include one same flavors, with a power battery and big capacity, bang xxk can last for several days to meet your daily vaping.
Bang Xxl,Bang Xxl Vape,Bang Xxl Disposable,Bang Xxl 2000
Shenzhen Kate Technology Co., Ltd. , https://www.katevape.com