저자: 미상 · 한국인
요약 보기
엔비디아는 게이밍·프로페셔널 그래픽에서 데이터센터 중심으로 수익구조가 안정화되었고, 2020년 발표된 Ampere 아키텍처로 AI 추론·학습 성능과 효율이 크게 향상되어 데이터센터 점유율(90%+)을 유지할 것으로 전망. 자동차·물류 등 AI 적용처도 확대 중. 하반기 신형 GeForce 기대감과 게이밍 시장 성장도 긍정 요인. 리스크로는 단기 주가 급등 부담과 코로나19로 인한 수요 감소 가능성을 언급하며, 데이터센터는 중장기 접근이 필요하다고 조언.
다듬은 글
222 Hello, this is [Name] lying down and doll when an agreement was reached. Today, I would like to talk about NVIDIA. I will discuss this through four major works. First, let me talk about the first one. NVIDIA has continued to grow focusing on gaming and professional graphics since releasing "Look" following the GeForce in 1999.
In the mid-2000s, there were high expectations as GPUs began to be used in computation, but for quite some time, GPUs have continued to be used in gaming and professional graphics. Now, there was a significant event. In 2016, the second generation and Alpha A-Gu faced off. At that time, the professional graphics and gaming sectors accounted for more than half of the total revenue.
And regarding the data center, which is what we call the 23, the GPU used in the battle against Alpha also became the pillar of this data center. In the case of the data center, it was 2015 and 2000 carefully both apps the entire It was not deployed to occupy a share comparable to a chip pro. However, when it actually suffered and competed against others, the revenue share of the data center sector increased to 22% from that point on.
Furthermore, we are predicting that for the first time this year, the revenue share of the data center sector will rise to number one in NVIDIA's total revenue. In the past, the gaming industry relied too heavily on RAD, so the cost was very high. However, as data centers and gaming have now become comprehensive sources, I can say that overall stability has been achieved.
Let me explain the second investment point. I can mention the competitiveness of the data center sector and the Ampere architecture announced at GTC 2020. Specifically, at PC 2920, NVIDIA introduced the latest announced Ampere architecture. While the Ampere architecture boasted better performance than previous products, the Ampere architecture showed an effect of increasing general computational performance by 1.5 to 2.1 times compared to the existing Volta architecture, but regarding AI...
I can tell you that yesterday, there was a dramatic effect of increasing performance by about 7 times from 3b, and the AR side was further strengthened. I can describe it as an architecture that fills the gaps enabling both inference and learning, unlike the past. With this enhanced AI, the question is what kind of competition NVIDIA will face in data centers.
Currently, NVIDIA holds an overwhelming market share in the accelerator server market for data centers. Therefore, it continues to maintain a market share of over 90% compared to other competitors. As for the newly released Ampere, compared to the previous Volta, it can be said that it has captured all aspects, including efficiency and price.
In the case of the previous Digix model, almost 50 units or a P600 size were required, but with Ampere, about 5 units are sufficient to handle a similar amount. Therefore, from the perspective of consumers and companies, using this product offers advantages in both fundamental and performance aspects, so naturally, they will have no choice but to prefer it.
Additionally, this Mela We have now merged with Knox, a QR code company. Therefore, while in the past we were responsible only for the physical aspects of accelerators, I can say that we are now reborn as a comprehensive solution company that also handles networking software. With this policy, I can say that NVIDIA, possessing such solutions, is bound to continue staying ahead of its competitors within the data center market.
Now, let me mention the third investment point. I can say that the AI market has come a little closer than we anticipated. NVIDIA has invested heavily in AI since the past and has fostered a complex ecosystem. Therefore, we have built an ecosystem in various areas, not just the data center accelerators we mentioned earlier.
Recently, AI is being used in areas beyond just data. Representative examples include Walmart, which uses it to analyze consumer patterns; USPS, which uses it to efficiently sort parcels and letters at the post office; and BMW, which has signed a contract to handle factory automation and sorting in their manufacturing logistics centers.
Regarding Five, however, we have started performing tasks such as separating the lamps using AR. Therefore, practically speaking, compared to the past, I can say that the scope is expanding not only in AI and data learning but also in the patent application sector. If you look at the chip frame side here, regarding learning like 1, it is a market that has consistently shown overall growth since the past.
However, regarding the 2 plus inference sector shown here, I can say that it is a market that has been growing rapidly recently. Since NVIDIA possesses unrivaled competitiveness in this market as well, I expect it to maintain its dominant second-place position for the time being, and I can say that NVIDIA's growth is expected along with this patent application market.
The final investment point is that I have some expectations for the gaming market in the second half of the year. The reason I say this is that we are anticipating new GeForce algorithms in the second half. As for the sports market, we are expecting weekend increases in viewership numbers this year.
We can say that the gaming market is continuing to grow to that extent; while we may not buy massive amounts of computers like in the past, we can expect to at least achieve a certain level. We expect sales to remain steady, and I can say that we can also anticipate the effect of this new product in the second half of this year.
Finally, to briefly explain NVIDIA's investment points and touch upon them briefly: the key investment point is that revenue stability is being achieved not only through the gaming business, which has long been a favorite, but also through the rapid growth of the data center division. Secondly, the company continues to strengthen its competitiveness in the data center sector; with the recent acquisition of Telex Kim and Earth, in addition to the new product, it has comprehensively developed the ability to apply even self-contained data center solutions.
Thirdly, the EL (Electronic Network) business is developing, and since we can continue to see NVIDIA's unrivaled competitiveness within that sector, I can say that NVIDIA's position remains promising alongside the expansion of the EL market. Lastly, there is a high expectation that new GeForce products will be released in the second half of this year; accordingly, compared to the first half, momentum in the gaming sector is definitely expected in the second half.
A brief point to consider regarding risk points is that recently, other Compared to semiconductor companies, M's stock price has risen sharply. Therefore, while this price may be a burden in the short term, since the data centers and JI mentioned earlier require a mid-to-long-term approach, I would advise you to consider this aspect when selecting your investment strategy.
Furthermore, as the COVID-19 pandemic has not completely ended yet, we still face the possibility of reduced demand. Regarding competitor products, which I briefly mentioned when discussing the third and fourth investment points, while there are concerns about new products, I anticipate that NVIDIA will continue to maintain its competitive edge.
That is all I have prepared for today. I will greet you again with good content from Gundam. Thank you.
Comments 0