AMD Acquired ZT Systems: A Challenge to NVDA's Dominance?
A major acquisition grabbed market attention yesterday: $Advanced Micro Devices(AMD)$ is set to pay $4.9 billion to acquire ZT Systems, a company specializing in cloud computing and AI data center equipment.
Media outlets are interpreting this move as AMD's challenge to $NVIDIA Corp(NVDA)$ 's dominance in AI computing.
Looking at stock performance, AMD rose 4.52% yesterday, while NVIDIA climbed 4.35%. Given the vast difference in their market caps, AMD's acquisition had little negative impact on NVIDIA.
AMD can't Surpass NVDA
Why was the market's reaction so subdued?
ZT Systems specializes in AI and general computing infrastructure, offering high-performance computing (HPC) servers, storage servers, and other custom equipment for major cloud companies and enterprises.
In the AI era, high-performance computing demands more sophisticated server integration. AMD aims to leverage ZT Systems to provide broader chip, software, and system design products for big data center clients like $Microsoft(MSFT)$ and $Meta Platforms, Inc.(META)$ , promising improved performance through tightly integrated products.
AMD CEO Lisa Su emphasized that ZT's value lies in providing direct support to clients building massive AI data center clusters. ZT's engineers will help AMD accelerate the testing and release of its latest AI GPUs to meet the needs of large cloud companies.
ZT Systems’ business resembles that of other major server firms like $SUPER MICRO COMPUTER INC(SMCI)$ and $Dell Technologies Inc.(DELL)$ . For instance, in June, ZT Systems launched the ACX200 platform with the GB200 Grace Blackwell superchip, featuring liquid-cooled rack-mounted servers and high-speed GPU interconnects via NVLink.
Lisa Su clarified that after the acquisition, AMD plans to spin off and sell ZT’s manufacturing assets, retaining only its system design capabilities, and it will not compete with companies like Supermicro.
Despite the hefty $4.9 billion price tag, ZT Systems' valuation isn't sky-high. With over $10 billion in annual sales, ZT's price-to-sales ratio is around 0.5, similar to the valuation of pre-AI boom Supermicro.
From a valuation perspective, ZT Systems doesn't add much value, and AMD can't surpass NVIDIA just by acquiring a server integrator!
NVIDIA’s Strongest Moat
Recently, former $Alphabet(GOOG)$ $Alphabet(GOOGL)$ CEO Eric Schmidt spoke at a Stanford University conference about NVIDIA. He admitted past mistakes, including underestimating CUDA, NVIDIA’s programming language, which he once thought was foolish.
Key Speakers At Aspen Ideas: Climate Conference
Now, he said CUDA is NVIDIA’s strongest moat. All major AI models run on CUDA, and only NVIDIA GPUs support CUDA—something no other chip can rival.
CUDA (Compute Unified Device Architecture) is NVIDIA’s integrated software and hardware technology, which allows GPUs to perform computations beyond graphics processing, marking the first time GPUs could be used as a development environment for C compilers.
In 2007, NVIDIA held its first CUDA technology conference and has invested $500 million annually in CUDA research and development since, representing about one-sixth of its revenue.
At that time, the market didn’t see CUDA’s value. Despite CEO Jensen Huang’s repeated explanations, Wall Street analysts weren’t convinced. NVIDIA’s market cap hovered around $10 billion, and its net and gross profits dropped significantly post-2008, even facing losses.
It wasn’t until the end of 2022, with the release of ChatGPT, that CUDA’s value became undeniable. NVIDIA’s long-term investment and the millions of developers it attracted form its biggest competitive edge, making it tough for AMD and Intel to catch up.
AI Revolution
In terms of AI GPU sales, AMD’s revenue this year is around $4.5 billion, while NVIDIA’s data center business is expected to surpass $100 billion!
While AMD’s $4.9 billion acquisition of ZT Systems will aid its AI GPU business, challenging NVIDIA’s dominance is unrealistic. The key in data centers is AI GPU performance, not server integration.
However, both NVIDIA and AMD still have room for growth in valuation. This AI revolution is set to exceed the dot-com bubble of the 2000s!
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
