Opinion
Losing TSMC Wouldn't Be Disastrous Because Intel Is Only 2x Behind
“And my contrarian take is that it wouldn't be that bad. Supply would take a shock for sure, but The best processes that we have in the West, like Intel, not that far behind, at worst, like maybe two X worst performance per watt. And the gap is just far smaller…”
Disclosure
Movva: Sail Research will buy data centers with only 95% uptime
“You'd have missing zero buyers for a data center that is 95% uptime. I'm that first buyer. I will buy 95% uptime.”
Assertion Not checkable as stated
Nvidia's BF16 Efficiency Barely Improves Across Hopper, Blackwell, and Rubin
“If you look at, you know, Hopper to Blackwell to Rubin, and you compare like for like, what is the performance per watt of a beefload-sixteen multiply? It hasn't improved all that much.”
Assertion Not checkable as stated
TSMC Performance Per Watt Barely Changes From 5nm to 2nm
“If you look at TSMC five nanometer versus four versus three versus two, the performance per watt on these chips doesn't change like a dramatic amount.”
Assertion Not checkable as stated
Cursor Forced Baseten, Fireworks, and Together to Prioritize Low Latency
“The challenge is, all those companies, you could take your pick, Base-Ten, Fireworks together, they all focus on low latency inference. And they were pulled in that direction by one very important customer Cursor.”
Prediction Not checkable as stated
Future AI Agents Will Prioritize Long-Horizon Task Completion Over Speed
“You wanted more persistence, more long horizon tasks, and now it's, to me, very obvious that the future of agentic inference is long horizon tasks. You're gonna run the machine for hours or days at a time. It doesn't matter if it spits out tokens at a hundred …”
Prediction Not checkable as stated
AI Workloads Will Eventually Shift to 90% Background and 10% Real-Time
“So long-term, I think, you know, we're going to end this year at maybe fifty-fifty background and real-time workloads, but I see this going to ninety-ten in favor of background.”
Opinion
Anthropic API Spend Is Now the Best Metric for Software Security
“When you want secure software, it's really a question of how many dollars did you spend on Anthropix APIs trying to break into your software. That is the best indication for how secure it is, because that's the best tool in the world.”
Prediction Not checkable as stated
AI Scientific Research Costs Could Soon Drop to Tens of Dollars
“And we have actually started to bring it within view, a dollar cost for these long horizon tasks that is reasonable. It's not millions, it's thousands, and maybe it could be hundreds or even tens of dollars in the near future, too. Have a definitive answer to …”
Opinion
Movva: NVLink is mandatory for low-latency AI inference
“NVLink is mandatory, I would say, for low latency inference.”
Prediction Not checkable as stated
Movva: Rival chipmakers will not replicate NVIDIA's NVLink quickly
“I think I'm not holding my breath for other companies broadly to figure out NVLink quickly. It's challenging technology to figure out. It's hard to scale. It's hard to productionize.”
Assertion Supported
Movva: Non-NVIDIA chips beat NVIDIA on FLOPS per dollar
“And so there are other chips that definitely rank higher than NVIDIA on flops per dollar. But they may not have as much interconnect”
Prediction Not checkable as stated
Movva: Cerebras and Groq will serve as accelerators alongside traditional GPUs
“Cerebris and Grok and maybe a couple others, you should think of them as accelerators. What they are really good at is being used in conjunction with a more traditional GPU-like device that critically has this off-chip memory built in.”
Assertion Not checkable as stated
Nvidia Strategically Rations Chips to Prevent Deep-Pocketed Buyers From Gaining Power
“NVIDIA sees the, if they just let the most deep pockets buy all the chips that Maybe hurts them in the long term if that customer ends up accruing a lot of more power. They understand that compute is power today, and so they're quite strategic about how they a…”
Insight
Movva: There are no bad AI chips, only bad pricing
“I like to say there's no bad chips, there's only bad pricing. And I will make any chip work at the right price.”
Assertion Not checkable as stated
Building 100-Megawatt Data Centers in the US Is Now Practically Impossible
“And basically there's no way to build a gigawatt data center in the United States easily anymore. Even a hundred megawatts is, is increasingly hard. It's basically impossible unless you're a very special set of customers. 10 megawatts is probably on the edge o…”
Insight
Movva: Uncontrolled P99 latency yields unbeatable economics for background agents
“We tell our customers, look, our average throughput is going to be very competitive, but our P 99, our 99 percentile latency is not going to be controlled. It cannot be. And in return, I'll give you unbeatable economics. And I think that's the right fit for ba…”
Insight
Tolerating Data Center Outages Unlocks Cheap Power for Inexpensive AI Chips
“The trick is that it's going to give me better access to power that no one else is going to touch, because it is so annoying to deal with that kind of outage. And if my chips are cheap enough, they're probably not going to be Nvidia racks. And if my chips are …”
Assertion Open · timeframe Dec 2026
Movva: NVIDIA is producing 5 million Blackwell chips this year
“NVIDIA's pumping out five million Blackwell chips this year.”
Opinion
Global GPU Utilization Is Actually Far Worse Than xAI's Clusters
“We all make fun of XAI for having, you know, some challenges with total flop utilization on its clusters, but the reality for the rest of the world is it's far worse. A ton of GPUs just sit in warehouses or sit In private pools allocated to a specific customer…”
Opinion
Movva: Frontier AI labs pay huge premium for 3-6 month lead
“In a line, I would say the labs pay an immense premium to be three to six months ahead of everything else.”
Prediction Not checkable as stated
Movva doubts commercial premium for short-term AI lead will last
“I don't know that the premium for being three to six months ahead is going to last that long. I mean, if you look at, like, enterprise deployments they don't move at three to six months speed.”
Disclosure
Movva: Sail's Custom Chip Strategy Sidesteps HBM by Offloading to Flash
“When I talk about building custom chips, and they ask me, oh, so what's different? Basically, it's about sidestepping the HBM shortage and focusing on More extreme offload to other forms of memory, such as flash.”
Insight
Movva: Compute startups should primarily attack Nvidia's reliance on HBM
“You want to pick something and say, I think they've underpriced the impact of how short we're going to be on HBM. We're going to push really hard in this other direction instead, which, you know, as an aside, I do think is probably the thing to attack most.”
