I last looked at NVIDIA (NVDA) as the Bull of the Day after another stunning beat-and-raise quarter, reported on May 22, launched the stock from $95 ($950 pre-split) to my price target of $140 ($1,400 pre-split).
We knew the new Grace-Blackwell “super-GPU” was coming, but apparently investors and analysts needed to hear it from Jensen himself that the demand and deliveries were going to be off the charts by Q4.
Tectonic Shift to AI Economy
The other huge dynamic we learned on the conference call with Jensen on May 22 essentially put to rest the idea that “cannibalization” was about to occur as datacenter and LLM-builders would stop buying older systems like the GH200 because they wanted to wait for the GB200.
Here’s what I wrote to my investor group on May 23 as we sat back and enjoyed the rocket launch…
TAZR Traders
You know the good news on NVDA: Wall Street continues to capitulate to the “tectonic shift” that has occurred with AI/ML/DL as more price targets play catch-up to our $1200…
Even as I raise mine to $1400 for the next 18 months.
And even as everyone and their Grandpa wring their hands about the NVDA market cap being worth more than half the world.
If you follow me on Twitter @KevinBCook you got to see the stunning “chapter and verse” from Jensen on the conference call yesterday… with all the analysts trying to find a fly in the ointment.
I used Otter.AI to transcribe the whole thing and will share more over the coming days… what a sermon Master Huang delivered!
It made me coin a new phrase about how he abolished the fear of “cannibalization” during this Global 2,000 — plus nation-states — deployment transition from GH200 to GB200 systems…
“Multiplicative Integration.”
Jensen didn’t say this, but this is what I felt he was talking about. I actually posted his final 5-paragraph “mic drop” because it captured the essence of the technological transformation we are in for science, industry, medicine, education, and society.
Jensen & Co. systems have “the transition” covered with TCO (total cost of operations) fully seamless — and bootstrapped — on the upgrades to the Blackwell platform.
Corporations won’t have to worry about how the current GH200 systems integrate with the new Blackwell systems because Jensen and his top-tier CUDA engineers have already built that seamless transition into everything.
(end of TAZR commentary excerpt from May 23)
Huang’s Law > Moore’s Law
Then in the first week of June came the annual Computex conference in Taiwan. And Jensen made it clear that the Acceleration of GPU computing was driving down nearly all Costs, including Energy usage, and providing powerful platforms for Science, Smart cities, and Simulation.
The oddly capitalized words in my previous sentence make up a new handy acronym I devised for technology investors to memorize when trying to understand the revolution that NVIDIA has birthed: ACESx3 is for Acceleration-Cost-Energy-Science-Smart-Simulation.
This is how I remember the power of NVIDIA, and the exponential reasons for staying invested in NVDA shares. And all of these letters stand for one of the most powerful ACESx3 innovations for human kind: advances in medicine and cures for our worst diseases.
After Computex, and the NVDA 10-for-1 stock split which took effect on June 10, I raised my price target to $200. I had already told investors (and savvy traders) in May that the gap up from $95 ($950 pre-split) was likely never going to be filled (barring WWIII).
Here was my commentary for Zacks TAZR members…
NVDA to $200
Posted on 6/10/24
TAZR Traders
This morning we taped the Zacks Ultimate Strategy Session (ZUSS).
My Top Stock Pick was, of course, NVIDIA.
And I tried to cram a lot (too much) into my 10-minute segment, but of course got lost along the way.
I neglected to reiterate the emerging potential of Jensen’s roadmap to $1 trillion in annual revenues by 2030 (a slide I showed you from last November).
And then I forgot to speak this key statement from my weekend notes that sums up my “No Cannibalization, Just Multiplicative Bootstrapping Integration” (that I introduced last week during Computex)…
“ALL OLD and NEW DGX-GPU systems BENEFIT FROM CUDA SOFTWARE UPGRADES, SUCH THAT ALL PREVIOUS GENERATION CHIP-SETS ACTUALLY GET FASTER!”
This is way beyond the “backward compatibility” thesis, so it gets lost on most analysts.
This is built-in “acceleration multiplicity.”
To say that the NVIDIA stack of enlightened software that extracts the most from the accelerated hardware is “an exponential event-horizon for enterprise and science” is the understatement of the decade.
We may not like what Meta, Google, and Amazon do with this massive data-power over society, but the genie is long out of the bottle.
Now, we’re aboard the ETI (Exponential Technology Innovation) Rocket that I previewed in 2017 in my Technology Super Cycle thesis.
And as NVIDIA CUDA-DGX-GPU acceleration has surprised almost everyone to rival Apple and Microsoft valuations, it’s also deliciously ironic that the “Agree to Disagree” topic for this month’s ZUSS was…
“Ben Rains and Kevin Cook will Agree To Disagree on whether the U.S. National Debt matters to the U.S. economy and the market.”
When I last addressed this debate with some degree of interest, it was 3 years ago here…
In any case, let’s buy all the NVDA shares we can below $120 because my new price target for 2026 is…
$200!
No one had a $2,000 price target pre-split so just ignore the Wall Street noise and stick with Cooker on the ETI that has yet to be seen!
Cooker
What About Gloomy Goldman on the AI Bubble?
You’ve probably heard the hub-bub by now that Goldman research says the “ROI on AI” is non-existent. And that means of course that AI is a massive bubble ready to pop.
That’s some cute “could happen” analysis. But it ignores what the Fortune 1000, the Global 2000, and the 20 largest nation-states are really up to: investing deeply for the future they see 5 years from now where if you don’t have your own deep learning, automation and LLM expertise, you will be mere gold dust against your competitors.
And as I’ve been predicting since February, the sales estimates for NVIDIA were going to cross $150 billion and head to $200 billion very soon.
Well it finally happened this month with an unnamed analyst posting a high-Street peg of $180 billion for next year.
I think it might be Tim Arcuri from UBS because on July 8 he talked about his bull case seeing $204 billion for next year. Here’s what I posted on Twitter/X that day…
UBS analyst Tim Arcuri leads the bulls with the call today for $204 billion next year, despite Goldman handwringing about “no ROI from all those GPUs yet.” Still building the field of dreams AI factories for science and innovation!
Bottom line: As I’ve been saying for 2 years, the analysts are still way behind the curve of what is emerging for NVDA sales over the next 3-5 years. Join me in buying every dip below $120 and if $100 scares you, buy more.
— Kevin Cook
Want the latest recommendations from Zacks Investment Research? [sponsor]Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report.
Source: Zacks