저자: LisaSu · 외국인
요약 보기
AMD 2분기 실적 호조: 매출 전년비 50%↑, 데이터센터 사업 급성장. 데이터센터 사업은 2027년까지 100% 이상 성장 전망, 범용·적응형 고성능 컴퓨팅 시장은 2030년 2조 달러 이상으로 제시. OpenAI·Meta·Anthropic과 파트너십 강조. 차세대 Venice 프로세서 출시. 기업들은 AI를 도구로 활용하되 AI 비용 통제에 집중하며 다양한 모델 수요가 중요하다고 주장.
다듬은 글
So, AMD Chairman and CEO Lisa Su joins us now in an interview with CNBC. Lisa, first I want to congratulate you. I think people don't understand that your stock has risen so significantly that it's logical to expect a little profit-taking. They have significantly outperformed Nvidia's stock price this year.
Therefore, I think it's important to start by saying that I didn't see any disappointment here. Jim, it's nice to be with you this morning. It's nice to see you, David and Carl, too. And yes, it was a very strong quarter for us. You know, we grew revenue by over 50% year-over- year, and we're seeing incredible momentum, especially in our data center business, where people need a lot more AI power, and we're making great products.
So we are very excited about the trajectory of our business and the momentum we are seeing heading into the second half of the year. Let's discuss some of what was said during the conference call. It seems to me that the analysts on the call were mostly overly pessimistic, and at one point you said, "Look, I think you're hearing from us that the number for AI in data centers is probably an underestimate," and then you mentioned the analyst's question without giving the exact numbers.
I think that a figure of over 100% should be considered as " significantly more than 100%" . In other words, it's entirely possible that the analyst didn't hear how strong your business would actually be. Jeremy, I would actually say that we are in a time of extraordinary AI computing right now. We have just updated our overall market indicators.
We see the market for general computing, as well as adaptive high-performance computing, growing to over $2 trillion by 2030. And with all this information, we wanted to rethink the state of our business. The fact is that our data center business is accelerating. We are extremely excited about our great strategic partners in fundamental models like OpenAI and Meta, and we just added Anthropic to that list.
In this case, we see our data center business growing by well over 100% by 2027, and these are very large numbers considering the base we are on. I also want to talk about processors. You, as you taught me, are just a processor company, and it seems like some analysts are just working with analysts, because the stock is down today.
Some of these analysts for some reason believe that your processor business will hurt you because of lower margins. I don't see how that's possible. I think it's just 100% a supplement. Am I being too optimistic about AMD ? Well, Jim, I think you're right in a few respects. You know, what we've seen in the AI market over the last six to seven months is a real disruption in demand, especially in the processor market.
So we see that as AI becomes more useful to humans, as we use more agents, and as enterprises move their workloads from the experimental stage to real production, humans need significantly more processing power, as well as graphics computing. And we are leaders in the field of processor computing. We are very proud of our technology stack .
We just announced the launch of our next generation Venice, and as you know, you know our Italian cities well. Venice is leadership in all aspects of computing, whether it's peak performance or overall performance per core. These are the areas where we are simply ahead. So we're excited about the state of the market because it needs a lot more computing power, and we're even more excited about our overall positioning in the market because our product portfolio is so broad and strong.
Lisa, as IT budgets are being formed for next year, how should people understand the mood of businesses? That is, is it about a certain spending discipline, trying not to spend too much and being cautious about potential profitability, or is it a game where you can "break things" a little and press the gas until failure?
Well, Carl, that's a great question. I think it's a little bit of both , and from what I see, spending time with a lot of CEOs and CIOs right now, the desire is to really use AI as a tool. AI is a tool that helps all of our businesses become more productive, and there are a lot of experiments underway to figure out which model is right and will deliver the best business results.
But we are also at a point where the costs of AI are becoming so large that everyone has to control them. Just as you manage personnel costs, you must manage AI costs. And I am deeply convinced that a wide variety of models will be important in the future. This includes using both the largest base models and open models, as well as, say, smaller and medium-sized models for all your tasks.
So, in my opinion, businesses today are focused on how to take advantage of AI while ensuring it is managed effectively for significant growth in the future.
Comments 0