Nvidia’s market value has been growing so fast — up more than five-fold since 2011, according to CNBC — that the AI chip designer’s stock price was poised to plummet.
But Nvidia defied the pessimists when it reported results for the quarter ending January 2024. After posting faster than expected growth in the last quarter and raising guidance for the current quarter, investors sent the stock up nearly 9% after-hours.
Since Nvidia’s role in the widespread adoption of Generative AI is still in its early stages, the stock remains a great investment.
Nvidia’s Q4 Performance And Prospects
By the standards of almost any company, Nvidia’s fourth quarter 2023 performance was outstanding. And fortunately for investors, Nvidia beat expectations and raised guidance.
Here are the key numbers from CNBC:
- Q4 2023 revenue up 265% to $22.1 billion — $1.5 billion more than analysts’ consensus of $20.6 billion for the period ending Jan. 28, according to London Stock Exchange Group.
- Q4 2023 adjusted earnings per share up 804% to $5.15 per share — 56 cents per share more than the analyst consensus of $4.59, according to Barron’s.
- Q1 2024 revenue guidance of 300% growth to $24 billion — $2,4 billion more than the $21.6 billion analysts’ consensus, according to the Wall Street Journal
Nvidia CEO Jensen Huang described AI as hitting “the tipping point” and indicated demand for the computing power that underlies it remained astronomical. “Demand is surging worldwide across companies, industries and nations,” the Journal reported.
Why Nvidia Leads The AI Chip Race
Nvidia dominates the AI chip market — with 80% to 95% of the AI chip market, according to The Globe and Mail— and has been building competitive advantages for more than 15 years to sustain its leadership position.
Nvidia’s Differentiation Strategy
If a company wants to lead its industry, it must choose one of two strategies:
- Differentiation — where the company delivers more value to customers than competing products and charges a price premium; and
- Low-Cost Producer — where a competitor sets its price well-below the industry average and reaps scale economies enabling it to reduce its costs below its price.
As I wrote last August, Nvidia is a differentiator as evidenced by customers’ willingness to pay a significant price premium and to wait for over a year to obtain its chips. According to the New York Times
Nvidia chips are expensive; however, the company argues they enable companies to save time training their Large Language Models — which more than offsets the price premium.
The price for each Nvidia H100 ranges from $15,000 to more than $40,000. AWS Vice President David Brown said Amazon’s AI chips are a bargain compared to Nvidia’s.
Nvidia CEO Jensen Huang disagrees. “If you can reduce the time of training to half on a $5 billion data center, the savings is more than the cost of all of the chips.” he said in the Times report. “We are the lowest-cost solution in the world.”
Nvidia’s Competitive Advantages
Nvidia’s competitive advantages flow from its ability to adapt to new opportunities faster than rivals and a page it took from Apple
Naveen Rao is an entrepreneur who sold his startup that built GPU chips for gaming and AI to Intel. Rao saw Nvidia adapting GPUs for AI far more rapidly than did Intel. He left Intel and started a software company, MosaicML where he compared Nvidia chips to its rivals. Rao told the Times, “Everybody builds on Nvidia first. If you come out with a new piece of hardware, you’re racing to catch up.”
Nvidia’s competitive advantages make it difficult for rivals to gain market share. The Times reported these competitive advantages include:
- Launching CUDA. In 2006, Nvidia announced CUDA — software for programming GPUs for other fields such as physics or chemical simulation. In 2012, researchers used the chips to identify a cat with precision.
- Hiring an AI team. Nvidia hired a team to train LLMs — gaining early insights into what AI practitioners wanted. Using that market intelligence, Nvidia built libraries — to perform tasks common to AI development — thus saving the developers’ time.
- Building faster chips every few years and expanding to complete computers. Nvidia consistently delivered faster chips every few years. In 2017, it began tailoring GPUs to handle specific AI calculations, sold chips or circuit boards for other companies, and provided complete computers aimed at faster AI processing.
- Introducing RTX 4800 and B100 chips. In January 2024, Nvidia launched RTX 4800 chips for consumer devices — primarily aimed at gamers but also useful for AI — with the ability to “generate AI video 150% faster than the last-generation model,” CNBC wrote. In late 2024, Nvidia is expected to launch a new AI system called B100. Morgan Stanley analyst Joseph Moore called the B100 “a strong advancement of state of the art,” the Journal noted. Bank of America
BAC analyst Vivek Arya expects the price of B100 products to be 10% to 30% higher than the popular H100 system, reported the Journal. - Forming partnerships with Big Tech companies and investing in startups. Nvidia collaborates with large technology companies such as ServiceNow
NOW NOW and Snowflake and funds startups. Nvidia invested in a $1.3 billion June 2023 financing for Inflection AI, which used to money to help finance 22,000 H100 chips.
These competitive advantages are difficult for rivals to copy. Moor Insights semiconductor analyst Patrick Moorhead told CNBC, “Nvidia has a double moat right now in that they have the highest performance training hardware. Then on the input side of the software, in AI, there are libraries and CUDA.”
Where Nvidia Stock Will Go Next
If Nvidia can continue to exceed high growth expectations, its shares will keep rising. However, the case for an Nvidia growth slowdown includes the following headwinds:
- Demand from big tech buyers could slow. D.A. Davidson analyst Gil Luria wrote Nvidia sells large numbers of GPUs to Microsoft, Amazon, Meta and Google — any of which could scale down their AI investment plans if demand slackens, noted CNBC.
- GPU lead times are shortening. Analyst Paulo Santos noted lead times for Nvidia's GPU orders have decreased from between 8 and 11 months down to “just 3-4 months, indicating a potential future sequential growth peak,” noted Seeking Alpha.
Nvidia bulls include Melius Research which recommended buying the stock. “If the upgrade from the A100 to the H100 is any indication, the Total Cost of Ownership benefit for data center operators will be enticing enough to fuel the upgrade and make 2025 a growth year,” noted analyst Ben Reitzes according to CNBC.
Analysts do not see much upside. Fifty three of them set an average 12-month price target for Nvidia of $700 a share — representing 3% upside according to CNN Business.
In after hours trading, the chip designer’s stock blew threw that target.
Goldman Sachs’ trading desk called Nvidia — which tends to fall after earnings reports before resuming an upward climb — “the most important stock on planet earth,” noted Bloomberg.
After the boffo earnings report, Nvidia’s stock defied the sinking and recovery path after its earnings report and became even more important.
https://news.google.com/rss/articles/CBMiaGh0dHBzOi8vd3d3LmZvcmJlcy5jb20vc2l0ZXMvcGV0ZXJjb2hhbi8yMDI0LzAyLzIxL252aWRpYS1zdG9jay1zb2Fycy1hZnRlci1ob3Vycy1vbi0yNjUtcmV2ZW51ZS1ncm93dGgv0gFsaHR0cHM6Ly93d3cuZm9yYmVzLmNvbS9zaXRlcy9wZXRlcmNvaGFuLzIwMjQvMDIvMjEvbnZpZGlhLXN0b2NrLXNvYXJzLWFmdGVyLWhvdXJzLW9uLTI2NS1yZXZlbnVlLWdyb3d0aC9hbXAv?oc=5
2024-02-21 21:43:36Z
CBMiaGh0dHBzOi8vd3d3LmZvcmJlcy5jb20vc2l0ZXMvcGV0ZXJjb2hhbi8yMDI0LzAyLzIxL252aWRpYS1zdG9jay1zb2Fycy1hZnRlci1ob3Vycy1vbi0yNjUtcmV2ZW51ZS1ncm93dGgv0gFsaHR0cHM6Ly93d3cuZm9yYmVzLmNvbS9zaXRlcy9wZXRlcmNvaGFuLzIwMjQvMDIvMjEvbnZpZGlhLXN0b2NrLXNvYXJzLWFmdGVyLWhvdXJzLW9uLTI2NS1yZXZlbnVlLWdyb3d0aC9hbXAv
Tidak ada komentar:
Posting Komentar