The majority of AI applications today run on Nvidia's chips. It's why Nasdaq calls the company the “King of AI.”
One software company Keith Kohl calls the “Nvidia Killer” is coming for the crown and it's on the brink of making a major announcement that could hand investors quadruple-digit gains.
Keith says engineers working in this nondescript building in Sunnyvale, California, are on the brink of unleashing the most powerful technology humanity has ever seen.
Keith Kohl is a 20-year market veteran who mainly covers energy and technology investments as the editor of Angel Publishing’s Technology and Opportunity, Energy Investor, and Topline Trader advisory services. I am intimately familiar with Keith's work, having previously reviewed his NH3 Company and “Algo Meds” Company teasers, among others.
Forbes calls this technology “the greatest profit engine in history.”
As you may have already guessed, Keith is talking about artificial intelligence (AI). However, this is not about an upgrade to ChatGPT, robots coming for your job, or self-driving cars finally going mainstream. This is about an epic problem standing in the way of AI's mass adoption, that one company is solving.
The Fatal Flaw Tech Giants Don't Want You to Know About
The truth is, AI is NOT ready for prime time because it has one fatal flaw.
As dedicated readers of Greenbull already know, we have been saying that AI is still very, very early stage.
Tools like ChatGPT are fun to use, but they can't be trusted to produce factual work any more than a guy with an eye patch and a habit that you met in a dark alleyway can.
Tech giants like Microsoft, Nvidia, and Google don't want us to know about this major flaw, because their sky-high share prices depend on everyone buying into the AI promise.
The major flaw Keith promises to reveal today lies at the heart of it all and if this problem isn't solved, AI will remain half-baked, at best, for years to come.
Keith says the little-known company from Sunnyvale he discovered owns the patent-protected key to unlock AI's real potential and he predicts it's going to be flooded with hundreds of millions and even billions of dollars.
All the information we need to move on this company can be found in a recently published investment report called “The Nvidia Killer: Unlocking the $100 Trillion AI Boom.”
The report is only available to subscribers of Keith's research service, Technology and Opportunity. The regular price for a year's subscription is $249 but for a limited time we can we can get one for just $99, a 60% discount.
A subscription comes with a six-month money-back guarantee, three research reports, including the one being promoted here, 12 monthly briefings, flash market alerts, and more.
A.I.'s Major Flaw
AI runs on high-performance chips, and almost all AI apps today use this H100 chip by Nvidia.
Demand for it is so high, that the CEO of one startup says, “It's like toilet paper during the pandemic.”
The regular list price for one is $10,000, but the current shortage made some people pay more than $40,000 on eBay earlier this year.
Needless to say, it takes a lot more than a single chip to build and run an AI model. Analysts estimate the training of OpenAI's ChatGPT requires 10,000 chips.
For $40,000 per chip, you'd have to spend $400 MILLION just to build a single AI model.
Besides the cost, AI's major flaw is the “bottleneck problem.”
We learn that this is just a nice way of saying AI melts Nvidia's chips.
Because AI applications constantly move large datasets between different parts of the chip, from memory to computing and back again, it uses a tremendous amount of time and power. Or to put it another way, running AI models on these chips originally made for gaming is like trying to loosen a screw with a knife.
AI also has another major flaw: It uses more energy than other forms of computing.
It's not just electricity consumption that drives up prices, these chips require a stunning amount of water for cooling. One study estimates you could fill a nuclear reactor's cooling tower with the amount of freshwater required to cool ChatGPT servers.
This is why an exchange with ChatGPT is estimated to be 10 times more expensive than a Google query.
This is hardly sustainable.
AI Is a Pipe Dream Without This Solution
Keith's Sunnyvale company has developed a solution that addresses these major flaws.
This firm's patent-protected flagship chip outperforms other chips by a factor of 100 on big-data workloads, but that's not all.
Its chip runs on only 60 watts of power. When we compare this to Nvidia's H100, which devours 400 watts of electricity, that's an impressive 85% less power consumption.
This kind of performance drastically slashes the cost of running AI systems and unleashes AI's true potential.
Yet hardly any retail investors know this company's name. Let's find out what it is together.
Revealing Keith Kohl’s “Nvidia Killer” Company
We get clues sprinkled throughout the presentation about the company:
- This company from Sunnyvale has specifically designed a chip to run AI applications on
- Its technology is protected by 125 patents, building an impenetrable moat around its vital technology.
- The U.S. Air Force and Space Force secured this high-performance chip for a top-secret project.
- This firm is still about 6,500x smaller than Nvidia, but its shares did jump 512% earlier this year.
This came down to two companies, but the “Nvidia Killer” is GSI Technology Inc. (Nasdaq: GSIT). How can we be so sure?
- GSI is headquartered in Sunnyvale and the company designs and makes semiconductor memory solutions for a variety of business and government customers.
- The Crunchbase company database assigns 142 registered patents to GSI, making it maybe the only time a newsletter writer has ever understated his stock pick.
- GSI announced a partnership with the U.S. Air Force and Space Force back in July.
- The company's market cap is only $55 million, or 20,000x smaller than Nvidia after its big summer run-up came crashing back down.
Real Chance to Make 120x Your Investment?
Keith's growth projections are based on the company announcing the release of a second-generation AI chip, which will “swamp it with orders.”
Something along the lines of this happened on October 18th, when GSI announced the “tape-out” of its Gemini-II Associative Processing Unit (APU) for AI.
The Gemini-II claims to bring significant performance enhancements over Gemini-I, with more than ten times the processing performance and eight times the memory density compared to Gemini-I. However, it won't be ready to roll out anytime soon. GSI hopes to have the chip ready early next year, with sampling scheduled for the second half of 2024.
Is it likely to displace Nvidia and other big boys, such as IBM, who are rolling out their own next-gen AI chips?
While almost nothing is impossible, this would be like all of the Avengers getting killed at the beginning of a movie.
Even if Gemini-II removes the bottleneck described earlier between the processors and memory arrays of a chip by performing massive parallel searches directly in the memory array where data is stored. It is still up against existing supply contracts the big chipmakers have in place with their customers, not to mention the capability to ship customer orders, even if they were to hypothetically receive them.
GSI currently trades for little more than double the cash on its balance sheet, with no debt, so it's a nice little speculation at this price. Just don't expect any major moves, anytime soon.
Quick Recap & Conclusion
- The majority of AI applications today run on Nvidia's chips. It's why Nasdaq calls the company the “King of AI”, so Keith Kohl thinks he's found a“Nvidia Killer” company that could hand investors quadruple-digit gains.
- This company has built a next-gen AI chip that solves the major flaws of today's chips, namely the bottleneck of processing power and high energy use.
- All the information we need on this company can be found in a recently published investment report called “The Nvidia Killer: Unlocking the $100 Trillion AI Boom.” The report is only available to subscribers of Keith's research service, Technology and Opportunity, which we can get for $99 for the first year.
- Keith committed the cardinal sin of dropping one too many clues and we were able to reveal his “Nvidia Killer” as GSI Technology Inc. (Nasdaq: GSIT).
- GSI's next-gen chip won't likely be ready to ship commercially until 2025. The stock is cheap on an absolute basis, but you may be waiting a long time for the stock to make its move if it ever does.
Is any company likely to beat Nvidia in the AI chip wars? Tell us what you think in the comments.