Hello, this is Orlando Kim. Today, we will take some time to examine how far AI has come through Broadcom's performance. First, Broadcom's earnings announcement has caused its stock price to drop by about 6.73%. This has led to some speculation that NVIDIA might also be affected, and indeed, it has seen a decline of about 0.88%, but it’s not a significant hit. There must be reasons behind this, and we will explore that. 

Now, let's take a look at an economic indicator that we didn't discuss earlier after the market closed: initial jobless claims. This refers to the number of people filing for unemployment benefits for the first time after being laid off. The forecast was 230,000, but it came in slightly lower at 227,000. This indicates a positive trend, and I would describe it as a good sign. We can also visualize this with a graph, so let's take a look at the chart and discuss it further.

So, we have the initial jobless claims. You can see this as a monthly chart, and it resembles a similar chart from his experiments. It has gone up and then come down, indicating a 4.3% increase. This will determine how much it goes down. What’s more important is to find trends within the chart, whether it’s going up or down. It’s crucial to check how the trend is behaving rather than just focusing on the numbers in front or behind. Therefore, it’s essential to examine economic indicators or financial metrics, especially over a period of five years or more. Similarly, looking at trends over three or two years is important. In terms of economic and financial indicators, this is quite significant. Next, we need to look at the continuous unemployment claims.

The allowance claim amount was about 229,000, but now it has come out to 230,000, which means it is quite similar. So, there hasn't been a significant change, and here, too, there seems to be a slight downturn. If unemployment decreases to 4.2 or below from 4.3 tomorrow, and if there aren't any major fluctuations, it seems that the impact on the stock market will not be significant. Therefore, when looking at economic indicators, it is much more important to observe the trend rather than just looking at individual data points. Recently, even after the confirmed figures are released, they are often revised six months later or even later. So, as much as possible, we should focus on the trends in economic indicators and financial metrics. Now, let's take a look at Broadcom's EPS. The current estimate for EPS is 1.22 dollars.

Recently, it came out to be $1.24, which is an increase of about 1.92%. However, we shouldn't just look at this; the surprises so far have also continued to rise. There will be future surprises as well, and we see increases from 1.2 to 1.38 to 1.46. It's important to note that these numbers are generally increasing. Looking at the future trends is crucial, and rather than just checking one piece of past data, we need to consider the overall context. 

Similarly, regarding sales, the estimated figure this time was $12.93 billion, while the actual figure came out to be $13.07 billion, which is a significant difference of 0.77%. Compared to last month, which saw an increase of about 3.57%, this time the increase was somewhat smaller. The sales figures were 14.1, 14.7, 14.65, and 15.16, showing that the increase in sales is not happening at a significant rate. In this regard, the outlook does not seem very positive.

Do not look at the past sales; rather, focus on estimating future sales and observing various trends. When looking at financial and economic indicators, it's essential to understand this approach. Now, let's delve into the growth dynamics of Broadcom. Notably, there is the custom AI chip, known as ASIC, which, along with AI networking, is currently a driving force for growth. Specifically, it is expected that in 2023 and 2024, AI revenue will reach 12 billion dollars. The original guidance was 11 billion dollars, and we will look into this later. This indicates that the AI sector is expanding significantly. Therefore, the AI sector accounts for about 20% of total revenue. This is how you should view it, especially in conjunction with Broadcom and Marvel.

Technology is indeed advancing, and you can see that all these things are leading to customized chips being at the forefront of technology. Companies like Google, Microsoft, Meta, and Amazon are currently providing customized accelerators alongside hyper-scalers. In particular, Broadcom has made significant investments in 2-nanometer technology over the past five years, which has allowed them to benefit greatly from the increased demand for custom accelerators. 

Another important aspect to consider is that Broadcom is a company providing networking chips for switching manufacturers, particularly for supermicro computers. Therefore, due to the increasing demand for networking connections in the construction of data centers, Broadcom is playing a crucial role.

Broadcom is ready to rapidly grow its networking business in the near future. Major AI Ethernet networking companies are showing strong growth in the data center market. For example, Dell's networking business in the second quarter has increased by about 80% compared to the previous year. They are providing chips, and Hypercedar is also supplying chips, indicating significant potential for development. 

Now, let's take a look at the company's 2024 financial highlights. I mentioned that we should consider it on a non-GAAP basis because the sales sector is much more focused on non-GAAP metrics. If you look at it now, we are comparing Q3 of 2023 to 2024. Of course, this comparison is influenced by seasonal factors.

It is normal, but usually, when we look at semiconductors, the current time is moving quickly, so comparing with the year 2024 is even more prevalent. This is because the speed is quite fast right now. Therefore, when we compare with a year ago, everything has changed. Consequently, the data presented by the company holds little significance. There is a reason for the recent drop in stock prices. Specifically, the sales forecast for the fourth quarter of 2024, which was originally estimated at $141.1 billion, has now been reduced to only $140 billion. In other words, the sales forecast has been downgraded, which is the first reason. The second reason is that the company is currently operating in two areas: the semiconductor field and the infrastructure field. In the infrastructure software sector, the current estimate is $55.2 billion. However, for now...

Approximately 58 billion has been reported, so as a result, it has passed as a platform. However, in the case of semiconductors, the estimated figure this time is 74.2 billion, but it has only reached 72 billion, which is worse than the estimate. Because of these two factors, the stock price has fallen by about 6%. 

So, regarding the two areas I mentioned earlier—semiconductor solutions and infrastructure software—let's take a look at the year-over-year change rates. Here, we see 152%, 174%, and 199%, indicating significant progress starting from the first quarter of 2024. On the other hand, the figures are only 4%, 5%, and 4%. This clearly shows that the infrastructure software sector is advancing significantly in the AI field. Thus, we can understand that simply looking at this aspect provides valuable insight.

There are various aspects to consider regarding how the standards are changing. As a result, the margins are significantly shifting. In conclusion, we are currently living off of these changes. Therefore, it is much more important that we adapt accordingly. 

Now, if you look at the chart, you can see that the infrastructure software sector is showing a very high bar graph. On the other hand, it's relatively flat here. Thus, depending on the performance of the semiconductor sector moving forward, it appears that there will be upward movement in stock prices. 

When it comes to earnings announcements, the most crucial time is the Q&A session. This is when we receive the answers regarding the future of companies and industries. Therefore, it is essential to be well-prepared for this. Typically, the Q&A follows the conference call.

First of all, there are the attendees. They are submitting their responses based on the given timeframes. The person providing the responses is currently the CFO of Draghi Co. If you look at this, there are various semiconductor companies involved. You may have heard of several periods, such as Biotech, Bank of America, and then William Stein from Truist. You know these people, right? Then there’s Stacey from Bernstein, and well-known figures like Joseph Moore from Morgan Stanley. Many people are familiar with them, including Tim O'Quinn from UBS. These are the individuals I have talked about a lot. So, when these people are asking questions and providing answers, there are also responses regarding future companies and industries. Many answers related to NVIDIA are emerging here as well. Therefore, it is essential to focus on this to fully understand the current situation.

We have familiarized ourselves with the Q&A and the earnings announcement. What we learned from this Q&A is that AI revenue continues to grow strongly. As you saw earlier, there are AI networking and customized AI, among others. All of these accelerators are being driven particularly by the strong demand from hyperscalers like Google, Meta, and Microsoft. Our hyperscale customers are continuously expanding their AI clusters. Specifically, customized AI accelerators have increased about 3.5 times compared to the previous year. This clearly indicates that AI demand remains robust. In particular, it is expected that AI revenue in the fourth quarter will grow continuously by 10%, surpassing $3.5 billion. Therefore, for 2024, AI revenue is projected to reach $12 billion, up from the previous guidance of $11 billion.

It seems that the dollar is expected to increase. The revenue from AI itself is growing, and the governance and guidance are continuously improving. You can see it this way: there is 11 billion dollars, and it is being raised to 12 billion dollars. We can understand that the demand is increasing. 

Now, there are questions coming in, particularly about semiconductor and AI revenues. They are asking whether the 3rd quarter figure of 3.1 billion dollars met expectations or was lower than expected. Additionally, they want to know where the continuous growth of 3.5% is expected to come from. It is being said that AI revenue will significantly increase next year, possibly due to additional customers from hyperscalers and consumer internet. They are asking how the growth of AI is being addressed. The 3rd quarter figures align with our expectations regarding AI.

Sales were similar, and there is a foundation that can be raised to over $12 billion. You can think of it all as hyper-scaled. In addition to cloud services, digital natives are truly creating a big tech landscape. Due to the combination of AI accelerators and networking, there is significant demand in big tech. This is what has been stated. 

Then there was another question about whether Timothy Cook thinks the growth rate of AI revenue next year will be different from or the same as the growth of GPU computing. He was asked to comment on that. Currently, there are large hyper-scalers like Cotan, which is a major platform with a massive subscriber base. When we talk about hyper-scalers, we refer to companies like Meta, Google, and Microsoft. These companies have large language models and many AI workload requirements.

We have a complete model that is being executed. Therefore, due to time constraints, it seems that we are moving towards creating custom silicon accelerators that can compute as much as possible. We are witnessing this transition, and it may take a few more years for it to fully happen. This means that we still need more time. As a result, the business model providing accelerators and networking for these large-scale hyperscale AI data centers is experiencing a good ripple effect. It still takes a considerable amount of time. It is not over yet; AI continues to evolve. Through the Q&A with these individuals, we can understand this. Now, the last question is quite important. In the second half of the year, is AI demand greater than supply? When asked this, they responded that orders keep coming in.

There is a rising trend, and this is not due to anything else but hyper-scale, meaning that it is because of big tech that orders are continuously coming in. It is expected that this will increase significantly more in 2015. This does not mean that 2024 is over; rather, it is anticipated that it will continue to grow. This is what the current president of Broadcom is indicating. Therefore, especially now that XPU is being distributed and infrastructure is becoming available, it is expected that demand will continue to come in significantly over the next 12 months. In this context, it can be said that AI is still very much alive. However, this means that while the market may not be able to grow significantly at the moment, it is not completely falling into a decline.

It doesn't seem like that is the case right now. Of course, there are some constraints on the rise, but from now on, it seems to be a decline. When viewed through the Q&A, it appears somewhat contradictory. Thank you for watching the video until the end. If you found it helpful, please subscribe and like. Thank you.

엔비디아(NVDA) 브로드컴의 Q&A를 통해 엔비디아와 AI산업의 미래를 알아보자

 # 미국 현지투자자가 30년 투자 기간동안 수없이 많은 시행착오를 통해 얻은 실전투자 매매 노하우를 솔직하게 공유하겠습니다

(유튜버 소개)
    MBA(skku), 자문사 운용역, 창투사 심사역,주식 및 외환,선물 투자 경력 30 년 차.

(유의 사항)
    본 영상은 투자권유 영상이 아닙니다.단지 참고 사항일 뿐입니다.투자의 최종 판단과 결정은 
    투자자 분께 있습니다.모든 분들의 성공 투자를 간절히 기원합니다.

(연락처 공개)
    khcapitalmg@gmail.com

