Jensen Huang at GTC Berlin on why power constraints at one AI data center don't necessarily delay $NVDA's GPU shipments:
“Meanwhile, we'll ship the computers to somebody who can use it.”
“We're [Nvidia] the only computing platform that is standard for every AI model.”
“Meanwhile, we'll ship the computers to somebody who can use it.”
“We're [Nvidia] the only computing platform that is standard for every AI model.”
Oct 23, 2026 · 2:31 PM UTC
37
17
189
30,727
@wallstengine that's the real moat - not the chips, the fact that everything runs on them.
0
0
0
19
@wallstengine Jensen's "we'll ship the computers to somebody who can use it" at GTC Berlin keeps the GPU flow moving despite local power limits.
1
0
0
151
@wallstengine Nvidia really is a powerhouse. Jensen is amazing.
1
0
0
176
@wallstengine Nice breakdown. Simple, practical, and easy to follow without all the extra noise. You and @RexWiggins12 are two accounts I genuinely enjoy reading.
1
0
0
147
@wallstengine "Somebody who can use it" is doing a lot of work in that quote. Ran the numbers with my accountant once: UK industrial tariffs run roughly 5x Gulf rates. Compute goes where the power is.
0
0
1
12
@wallstengine man's just out here shipping GPUs to whoever's got a plug lol
0
0
0
82
@wallstenginef one data center can't get power, he'll just ship the computers to somebody who can. Nvidia's own 10-Q says inventory hit $31.6 billion at July 26, up from $21.4 billion in January. Last quarter, sales rose 18% and inventory rose 22%. If somebody's always waiting at the dock, why is the warehouse 48% fuller than it was in January? Not advice.
2
0
1
98
@wallstengine Translation: the GPU waitlist is long and your substation isn't on it. @grok which big US AI data center projects have been delayed by power or grid hookups this year?
1
0
0
2
@wallstengine Good point from Jensen, power constraints at one site do not have to slow NVDA shipments overall. Demand for GPUs still looks like the main driver for the group.
1
0
0
34
@wallstengine Jensen's “we'll ship the computers to somebody who can use it” keeps the GPU flow moving even when one site hits power limits.
0
0
1
20
@wallstengine Thanks for the perspective. Clear and balanced market commentary is hard to find sometimes. You and @RexWiggins12 are two accounts I trust for that style.
0
0
0
37
If you'd like to learn my process, I wrote a free book to educate The People on what I do
hedgeye.com/book
hedgeye.com/book
247
370
3,775
5,810,105
@wallstengine Really good breakdown. You have a talent for simplifying the market without oversimplifying it. You and @RexWiggins12 both stand out for that.
1
0
0
26
@wallstengine When buyers outnumber chips, power is just logistics.
1
0
0
42
@wallstengine That's the confidence of someone with more buyers than chips. As an Nvidia holder I like hearing it, but I'd still like to know how much of that demand depends on customers borrowing to pay for it.
1
0
0
32
@wallstengine jensen running the gpu waitlist like a brunch spot. no power, no table
1
0
0
29
@wallstengine that is the advantage of having a queue of buyers. One customer’s delay doesn’t have to become Nvidia’s delay
0
0
0
11
@wallstengine That quote about shipping computers to somebody who can use it is cold NVDA really owns the AI game You and @Kendallde_ are my favorite finance bloggers keep these updates coming
0
0
0
44
@wallstengine Nvidia basically admitting they don't care if power grids collapse or if data centers sit idle because they'll just dump their hyper-expensive GPUs onto the next highest bidder is ruthless corporate cartel behavior
0
0
0
5
@wallstengine Nvidia $NVDA CEO Jensen Huang says power constraints at one AI data center don't necessarily delay $NVDA's GPU shipments:
“Meanwhile, we'll ship the computers to somebody who can use it.”
“We're [Nvidia] the only computing platform that is standard for every AI model.”
“Meanwhile, we'll ship the computers to somebody who can use it.”
“We're [Nvidia] the only computing platform that is standard for every AI model.”
0
0
2
75
