TESLA LIVE 24.com · 24시간 반도체·SOXL 시세·속보 한국어 실시간

10년간 6,000% 상승한 반도체 3개주: 지금 사야 할 2개, 피해야 할 1개

Yahoo Finance · 2026.10.08 17:35 · 원문 사이트
🇰🇷 한글 번역
(이 부분 번역 실패) 마이크론은 지난 10년 동안 놀라운 상승세를 기록했으며, 2025년 봄 이후 주가가 포물선 형태로 치솟았다. 이러한 상승은 DRAM(동적 랜덤 액세스 메모리)과 NAND(플래시) 모두에서 공급 부족이 이어지면서 메모리 가격이 급등한 것과 직접적으로 연결되어 있다. 이 중 상당 부분은 특히 GPU와 거의 필수적으로 패키징되어야 하는 HBM(고대역폭 메모리)에 대한 AI 수요의 급증과 관련이 있다. 나스닥: MU 주요 데이터 포인트 그러나 업계의 많은 부분이 HBM에 집중하고 있는 반면, 마이크론은 오히려 차세대 메모리 분야에서 뒤처진 위치 덕분에 혜택을 누려왔다. 기존 메모리 가격이 HBM보다 더 빠르게 상승했으며, 이는 마이크론의 매출과 매출총이익률이 급증하는 데 기여했다. 하지만 여기서 반전이 있다. 기술적으로 뒤처진 위치는 장기적으로 좋은 전략이 아니며, 일반 메모리 가격이 정상화될 경우 가장 큰 타격을 입을 수 있다. 따라서 주가의 저평가된 밸류에이션에 현혹되지 말라. 이는 장기적으로 보유하기 적합한 메모리 종목이 아니다. 나는 HBM 분야 선두주자이자 엔비디아의 주요 공급업체인 SK하이닉스(SKHY -2.36%)를 훨씬 더 선호한다.
📄 원문 (English)
The semiconductor industry has been one of the most lucrative sectors to invest in over the past decade, with three stocks posting gains of over 6,000%. Nvidia (NVDA -0.74%) leads the pack with a more than 14,000% return during that stretch, followed by Advanced Micro Devices (AMD -0.55%) with a gain of over 9,000% and Micron Technology (MU +4.06%) at more than 6,000%. So, which of these chip stocks still looks like a buy for the next decade? Let's take a closer look at why Nvidia and AMD remain buys, and why I'd pass on Micron. 1. Nvidia Nvidia's decade of dominance stems from its graphics processing units (GPUs) becoming the primary chips for AI workloads, particularly for training AI models. The company created a wide moat in AI model training by seeding its CUDA software platform in places doing early AI research, leading to most foundational AI code being written on its software and optimized for its chips. That moat remains today. However, the next decade will be about Nvidia's transition from a GPU maker to a full-fledged AI infrastructure company. Networking has been one of Nvidia's fastest-growing businesses, and it has developed its own central processing units (CPUs), which are becoming more important with agentic AI. NASDAQ: NVDA Key Data Points Meanwhile, Nvidia's acquisition of "Groq" and its language processing units (LPUs) earlier this year set it up well for inference. This all allows it to offer complete end-to-end server solutions for specific AI tasks. It has also made important software-layer acquisitions with SchedMD and Hugging Face. A huge winner over the past decade, Nvidia is well-positioned for the next 10 years, making the stock a buy. 2. Advanced Micro Devices Despite ceding the AI training market to Nvidia's GPUs, AMD's performance over the past decade has nonetheless been impressive. Many of these gains have come recently, as the chipmaker has positioned itself well in three key markets: inference, agentic AI, and physical AI. Inference is more memory-bound than compute-constrained, and AMD has made important strides in capturing enterprise workloads constrained by memory availability and costs. Its chiplet designs allow its GPUs to be packaged with more high bandwidth memory (HBM), and it has recently acquired memory optimization company MEXT and inference chipmaker Taalas, which hardwire AI models into its chips. AMD has broken into the inference market with two $100 billion deals with Meta Platforms and OpenAI, as well as a large deal with Anthropic. NASDAQ: AMD Key Data Points In addition to its opportunity in inference, the company is also a leader in data center CPUs. The server CPU market is exploding due to the rise of agentic AI, with Bank of America recently projecting it will rise from $19 billion in 2025 to $180 billion in 2030. That's huge growth, and the Wall Street firm also expects AMD to continue to take market share in the space. Finally, following its announced acquisition of World Labs, AMD is also looking to tackle the physical AI and robotics markets. World Lab is a leader in the spatial AI models that will eventually play a big role in these markets, positioning AMD to become a strong player in the space. Given its opportunities in inference, agentic AI, and physical AI, this is a stock to own for the next decade. 3. Micron Technology Micron has posted spectacular gains over the past decade, with its stock going parabolic since the spring of 2025. These gains are tied directly to surging memory prices, as both DRAM (dynamic random access memory) and NAND (flash) have been in short supply. Much of this is tied to soaring AI demand, especially for HBM, which essentially needs to be packaged with GPUs. NASDAQ: MU Key Data Points But while much of the industry is focused on HBM, Micron has actually benefited from being a next-generation memory laggard. Conventional memory prices have climbed faster than HBM, which has helped Micron's revenue and gross margins skyrocket. But here's the twist: Being the technology laggard isn't a good long-term strategy, and it is most exposed to ordinary memory prices normalizing. So don't let the stock's cheap valuation fool you; this isn't the memory stock to own over the long term. I much prefer HBM leader and primary Nvidia supplier SK Hynix (SKHY -2.36%).
▶ 칩 라이브 24 홈으로 — chiplive24.com

다른 반도체 뉴스

Is Nvidia, at Bargain Levels Right Now, a Buy? Here Are the Bull and Bear Cases. · Yahoo FinanceSamsung flags 783% profit jump to record $80bn as AI chip boom also lifts TSMC sales · Yahoo FinanceWhy Micron Investors Should Pay Attention to Samsung’s Record Profit · Yahoo FinanceIntel, Marvell, Wolfspeed, Levi Strauss, and More Stocks That Explain Today’s Market · Yahoo FinanceWhy Are Nasdaq, Dow Futures Falling Premarket? MU, TSM, APLD, IONQ, HOOD, BULL, SLS In Focus · Yahoo FinanceTaiwan Semiconductor Manufacturing Could Have Big News on Oct. 15 · Yahoo FinanceShell achieves NVIDIA DSX Ready Coolant qualification for direct liquid cooling, advancing high-performance AI computing · Data Center DynamicsNous Research $90M Raise: NVIDIA’s Dual Bet Reveals the Compute Landlord Strategy · Yahoo Finance