NVIDIA Blackwell Ultra Fuels AI & HPC Innovation, Efficiency and Capability
March 21, 2025
I/O Fund
Team
NVIDIA’s groundbreaking hardware technologies and AI are unlocking unprecedented computational power. At the NVIDIA GTC 2025, NVIDIA unveiled its Blackwell Ultra GPU designed for the “Age of Reasoning” at its 2025 GPU Technology Conference (GTC). AI accelerators like GPUs are well suited for AI training and inference due to parallel processing, which allows for many calculations to be performed simultaneously. Only 30% of the top 500 supercomputers relied on accelerated computing; today, 80% do. The Green 500 ranking of supercomputers by energy efficiency shows an even more pronounced trend.
NVIDIA Blackwell Ultra GPU and GB300 NVL72 server key specifications included.
Source: NVIDIA
Blackwell Ultra GPUs for the Age of Reasoning
AI reasoning models emulate how the brain thinks to render a conclusion, popularized by OpenAI’s o1, Google’s Gemini 2.0 Flash Thinking and DeepSeek’s R1 A1 models. Reasoning models improve responses to queries and more powerful GPUs improve the performance of these models. Blackwell Ultra GPUs are the next generation of the evolution of the GB200 bolstered by more inference power horsepower, packing 50% FLOPS at 1.1 exaFLOPS of FP dense compute.
NVIDIA Blackwell Ultra AI Factory Output chart shows 50x performance increase.
Source: NVIDIA
At the NVIDIA GTC 2025, in his March 18 presentation titled, “The Next Frontier of AI Supercomputing: Efficiency With Unprecedented Capability”, NVIDIA’s Vice President of Hyperscale and HPC Computing, Ian Buck, stated, “Blackwell Ultra takes GB200’s 40x data center revenue opportunity to 50x”, citing faster token serving and higher throughput ideal for post-training for models like DeepSeek, which chomp through 100 trillion tokens.
NVIDIA GB300 NVL72 Unleashes Inference Horsepower
NVIDIA’s GB300 superchip combines two Blackwell Ultra GPUs with one Grace CPU. Blackwell Ultra GPUs can be used in the NVL72 rack server, which integrates 72 Blackwell Ultra GPUs and 36 Grace CPUs. The NVIDIA GB300 NVL72 has a fully liquid-cooled rack-scale design. AI factories achieve 50X higher output for reasoning model inference with the NVIDIA GB300 NVL72 compared to the NVIDIA Hopper platform when used with the NVIDIA Quantum-X800 InfiniBand or Spectrum-X Ethernet paired with ConnectX-8 SuperNICS.
Blackwell Ultra’s Silicon Photonics Slashes Power Consumption by Up to 77%
NVIDIA’s Blackwell Ultra GPUs use co-packaged optics with silicon photonics, which integrates optical and silicon components onto a single substrate. This reduces power consumption by eliminating the need for external lasers and pluggable transceivers to achieve a significant reduction in power from 39 watts to 9 watts. Buck said that silicon photonics "… gives you that benefit from going from 30 watts of power down to only 9 watts of power for the same number of ports, and that's huge. It doesn't sound like 39 sounds a lot. But if you get 400,000 GPUs in an AI supercomputer, there's like 24 megawatts of lasers like so that's a lot of laser light that could be optimized and made more efficient.”
Join thousands of investors who trust I/O Fund’s expert stock analysis on AI, semiconductors, cryptocurrency, and adtech — sign up for free! Click here!
Beth Kindig, Lead Analyst at the IO Fund, pointed out in her “AI Power Consumption: Rapidly Becoming Mission-Critical” blog article that, "In my analysis last month on the Blackwell architecture, I made the argument these estimates are too low and that my firm expects we will see a $200 billion data center segment by end of CY2025 propelled forward by the B100, B200 and GB200, including the following points: “Taiwan Semi’s CoWos capacity, which is essential for Blackwell’s architecture, is estimated to rise to 40,000 units/month by the end of 2024, which is more than a 150% YoY increase from ~15,000 units/month at the end of 2023. Applied Materials has boosted its forecast for HBM packaging revenue from a prior view for 4X growth to 6X growth this year.””
The Next Generation CPU: Vera CPU: Grace’s Successor
NVIDIA’s next-generation CPU is Vera, a follow-on to Grace. With 88 cores (176 threads via spatial multithreading), Vera doubles Grace’s performance 2X, memory bandwidth by 5X per watt, and has a beefier chip-to-chip link for the upcoming Rubin GPU. “Every core talks to every other core,” Buck stressed, contrasting x86’s front-end focus. Vera’s 12-thread memory saturation trounces traditional CPUs, feeding GPUs for AI and HPC back-end tasks. Vera Rubin will launch in 2026. Vera Rubin NVL 144 will launch in the second half of 2026. FYI, Vera Rubin was an American astronomer who discovered dark matter. Rubin will mark the shift from HBM3/HBM3e to HBM4 and HBM4e for Rubin Ultra.
The Next Generation GPU Architecture: Rubin Ultra
NVIDIA will be launching Vera Rubin NVL 576 in the second half of 2027, which will have 14X the performance of GB300 NVL72. Rubin will have 1.2 ExaFLOPS of FP8 training compared to just 0.36 ExaFLOPS for B300, resulting in 3.3X compute performance. Bandwidth will improve from 8 TB/s to 13 TB/s. It will have 576 Rubin GPUs in a rack. Compute density is boosted by featuring four dies per package. Rubin Ultra NVL576 will have 365 TB of memory. The inference compute with FP4 rises to 15 ExaFLOPS with 5 ExaFLOPS of FP8 training compute. NVIDIA hinted the next-generation architecture after astronomer Vera Rubin will be named after theoretical physicist Richard Feynman.
The I/O Fund recently entered five new small and mid-cap positions that we believe will be beneficiaries of this AI spending war. We discuss entries, exits, and what to expect from the broad market every Thursday at 4:30 p.m. in our 1-hour webinar. For a limited time, get $110 off an Annual Pro plan with code PRO110OFF [Learn more here.]
Disclaimer: This is not financial advice. Please consult with your financial advisor in regards to any stocks you buy.
Recommended Reading:
Get a bonus for subscription!
Subscribe to our free weekly stock
analysis and receive the "AI Stock: 5
Things Nobody is Telling you" brochure
for free.
More To Explore
Newsletter
2025 Market Outlook: Why Stocks and Bonds Are Signaling More Volatility
As the S&P 500 reaches a key bounce target, troubling signs in bonds and consumer behavior suggest this market rally may be on thin ice. I/O Fund’s Knox Ridley explains why volatility may intensify an
The Impact of Tariffs on the Stock Market: Q1 Preview
Rising tariffs are injecting significant uncertainty into the stock market, triggering daily volatility and forcing analysts to revise earnings estimates. Our Q1 preview dives into the potential impac
Tesla Stock Faces Recalibration of Growth Expectations
Tesla’s stock is now facing a recalibration of expectations after Q1’s delivery report missed by a wide margin. Q1’s analyst consensus has gone from $25.98B at the start of the year to $23.97B in earl
The Fed Can’t Save This One: Why Bonds May Break the Stock Market in 2025
In early 2025, as markets rallied to new highs, we warned that divergence across key sectors signaled a looming correction. Now, with all major indexes in a technical bear market and bond market dysfu
Oracle Stock Outlook: Revenue Could Double by FY2029, yet Targets Seem Lofty
Late in 2024, Oracle outlined an ambitious plan to nearly double its revenue by fiscal 2029, hinging on long-term growth in enterprise AI and cloud spending. Oracle sets itself apart from its hypersca
I/O Fund Reports 210% Cumulative Return -- Ranking Above Wall Street's Best
In 2024, I/O Fund posted a 35% return, significantly outperforming popular tech ETFs, which recorded an 8% return over the same period. On a cumulative basis, the results translate to a remarkable 219
The Harsh Truth: Retail Investors Take the Brunt of Market Losses
Retail investors face significant disadvantages in the stock market, often underperforming institutional investors by a wide margin. Studies show that high-frequency trading firms dominate market acti
NVIDIA’s GB200s for up to 27 Trillion Parameter Models: Scaling Next-Gen AI Superclusters
Supercomputers and advanced AI data centers are driving the AI revolution, enabling breakthroughs in deep learning and large-scale model training. As AI workloads become increasingly complex, next-gen
NVIDIA Blackwell Ultra Fuels AI & HPC Innovation, Efficiency and Capability
NVIDIA’s latest Blackwell Ultra GPU, unveiled at NVIDIA GTC 2025, is transforming AI acceleration and high-performance computing (HPC). Designed for the “Age of Reasoning,” these cutting-edge GPUs del
Nvidia CEO Predicts AI Spending Will Increase 300%+ in 3 Years
Nvidia has traversed choppy waters so far in 2025 as concerns have mounted about how the company plans to sustain its historic levels of demand. At GTC, Huang threw cold water on many of the Street’s