Forwardhour (@forwardhour) header
Forwardhour (@forwardhour) profile picture

the forward book for gpu owners

San Francisco
Joined November 2026
Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)Photo by Forwardhour (@forwardhour)
What 3 days of GPU news told us: A JPMorgan note on Nvidia's financing knocked neocloud stocks down 6 to 10%. Money for GPU owners just got more expensive. SemiAnalysis cut its 2030 AI accelerator capex forecast 40%, partly on lower chip pricing. The traded H100 hour bounced to $3.16 while posted on-demand prices kept falling. OpenAI signed a 6-year, 457 MW AMD capacity deal. No price disclosed. Debt costs more, forecasts get cut, and the biggest deals leave no price behind. GPU owners need a floor they can see That's what Forwardhour is building. Beta opens 22 Nov
2
87
Same GPU hour and one month apart on a public H100 rental index: $2.92 on 29 Oct. $3.53 on 3 Nov. $2.93 on 18 Nov. That is a $0.61 swing. On 1,000 GPUs it is $445,300 a month. The rent you get depends on which day your contract ends
1
1
83
Two GPU price indexes, one H100, last week: @OrnnExchange (traded): $3.06, down 9.2% @silicondata (posted): $3.08, up 2.3% Two cents apart. Opposite directions. H200: $6.28 vs $3.57, a 1.76x gap. Both are lined up to settle futures. Which should a GPU owner price against?
1
2
4
118
Sources: Ornn OCPI close 15 Nov, Silicon Data index 16 Nov, Compute Terminal board 16 Nov (CC BY 4.0). Week on week: Ornn 8 to 15 Nov, Silicon Data 9 to 16 Nov. Beta opens 22 Nov. No floor is sold today.
1
19
Nvidia reports tomorrow. Today, Lambda raised its on-demand price for an 8× H100 instance by 15%. If you rent or operate H100s, how did your last renewal compare with the previous contract? Up, flat, or down
1
1
88
Nvidia's guarantees helped GPU buyers get financed. Now investors are questioning them. For a GPU owner the risk is simpler: the day a rental contract ends and the GPUs go empty. In our example, 448 H100s come free within 90 days. That's $707,808 a month in rent to replace. When do yours?
🚨 Another major scandal is coming out of NVIDIA. And this one has nothing to do with Jensen Huang's affairs. According to a Wall Street Journal report, Nvidia was offering billions in financial guarantees to AI companies that could not afford its chips, so they could keep buying them. Some of these deals were being pushed by Jensen's own son-in-law, Nico Caprez, who became a vice president in January. His team was telling small cloud companies and data center builders that Nvidia could arrange credit support for them within a couple of weeks. The goal was simple: turn them into long term buyers of Nvidia GPUs. And the size of Nvidia's commitments has exploded. Its own filing shows $530 BILLION in commitments and guarantees, up from $184 billion just one quarter earlier. Inside the company, CFO Colette Kress was raising concerns. She kept asking how these guarantees were priced and what Nvidia was really getting back. In an August meeting, Jensen cut her off and called her "a deer in the headlights." She was also the one who cut Nvidia's guarantee on OpenAI's Ohio data center from $250 billion to $105 billion. Now Jensen is gone, and the CFO who questioned these deals is running Nvidia. To understand why this is a big deal, you need to understand the AI funding loop. Here is how it works: 1. Nvidia puts money into an AI company, or guarantees its loans and leases. 2. That company uses the money and credit to build a data center. 3. The data center is filled with Nvidia GPUs. 4. Nvidia reports those GPU sales as revenue. 5. If the company cannot rent out all its computing power, Nvidia agrees to buy the unused part. SO NVIDIA IS THE INVESTOR, THE LENDER, THE SUPPLIER AND THE BACKUP CUSTOMER, ALL IN THE SAME DEAL. The money goes out of Nvidia and comes back to Nvidia as sales. CoreWeave is the clearest example. Nvidia owns a stake in it, sells it chips, and agreed to buy up to $6.3 billion of its unused computing capacity through 2032. OpenAI is the biggest. Nvidia is guaranteeing up to $105 billion for its Ohio data center, supplying all the chips inside it, and invested $1.5 billion in the company building it. Nvidia's own filing for the quarter ending July 26 shows where the money is tied up: • Supply commitments for memory and chip wafers: $279 BILLION, up from $119 billion. • Guarantee on the OpenAI Ohio data center: up to $105 BILLION. • Deals to buy unused cloud capacity from neoclouds: $36 BILLION. Nvidia's actual liabilities on its balance sheet: $91 BILLION. That means the commitments outside the balance sheet are almost 6 times bigger than the debts on it. Nvidia does not lose this money unless something breaks. A guarantee only costs Nvidia if a customer cannot pay. But this is where it gets uncomfortable. The WSJ says Jensen himself had started to panic. He warned executives that the biggest cloud companies were moving to their own chips to depend less on Nvidia. At the same time, the wafers and memory Nvidia had committed to buy were coming at him like an "avalanche." So the biggest customers were trying to leave, Nvidia was locked into buying huge amounts of supply, and smaller customers were being kept alive with Nvidia's credit. And things were already starting to crack: • Nvidia paused some of its financing deals in August. • Nearly half of the $80 billion in data center bonds issued this year were rated junk. • Neocloud stocks crashed this week, with Nebius down over 20% and CoreWeave down 17% at the lows. • OpenAI raised less than it wanted, at a lower valuation. And now Kress is the one deciding whether this continues. If she pulls back on the financing, the companies that depended on Nvidia's money will have to borrow elsewhere, in a debt market that is already nervous. If they cannot, they stop buying chips. And if they stop buying chips, Nvidia's revenue drops. That is the risk of a loop. It works perfectly when money keeps flowing, and it hurts on the way down.
1
1
1
42
The video uses example data, not a real fleet. forwardhour.xyz Beta opens 22 Nov. No floor is sold today.
6
Crusoe's H100 hour goes from $3.90 to $4.50 on 1 Dec. Lambda's new list starts 16 Nov (1x H100: $4.89). Next month's price is public. Next year's hour has no public price we could find. If you own GPUs: where is your floor?
1
2
46
Sources: Crusoe and Lambda price pages via Sub Pulse: sub-pulse.com/blog Silicon Data index, 10 Nov: H100 $3.06, B200 $6.97 forwardhour.xyz Beta opens 22 Nov. No floor is sold today.
13
Forwardhour retweeted
this is one of the clearest takes on GPU pricing i see talks about how many GPUs are being built but very few talk about what an hour on them will be worth in two or three years if you own GPUs, finance them, or rent them, it's worth the 8 minutes
We made an 8 minute investigation into GPU prices. HIVE's annual report lists H200s at $3.99 an hour on demand and $1.45 to $1.85 on six to twelve month contracts. Those are different commitments. The harder part is finding out what other contracts actually pay. A Jump Trading speaker described the term market as an "unlit market", with deals done over @SlackHQ , @X DMs and @telegram . In one contract filed with the SEC, the price is blacked out. We followed that into the cost of owning the chips. In the film's model, an H100 hour costs $1.48 assuming a six-year life, or $2.26 assuming three years. We also looked at a five-year loan backed by contracts averaging about three years. What happens after those contracts end decides whether the numbers work. This is the research Forwardhour is built on. We made it with four Grok @bot agents.
1
3
108
An H200 is priced like a hotel room. $3.99 an hour if you walk in, $1.45 to $1.85 if you book 6 to 12 months ($HIVE annual report). You can book a hotel room for next year's dates. Try finding a public price for an H200 hour in 2027. We spent 8 minutes on why. Full investigation below.
We made an 8 minute investigation into GPU prices. HIVE's annual report lists H200s at $3.99 an hour on demand and $1.45 to $1.85 on six to twelve month contracts. Those are different commitments. The harder part is finding out what other contracts actually pay. A Jump Trading speaker described the term market as an "unlit market", with deals done over @SlackHQ , @X DMs and @telegram . In one contract filed with the SEC, the price is blacked out. We followed that into the cost of owning the chips. In the film's model, an H100 hour costs $1.48 assuming a six-year life, or $2.26 assuming three years. We also looked at a five-year loan backed by contracts averaging about three years. What happens after those contracts end decides whether the numbers work. This is the research Forwardhour is built on. We made it with four Grok @bot agents.
41
Forwardhour retweeted
We made an 8 minute investigation into GPU prices. HIVE's annual report lists H200s at $3.99 an hour on demand and $1.45 to $1.85 on six to twelve month contracts. Those are different commitments. The harder part is finding out what other contracts actually pay. A Jump Trading speaker described the term market as an "unlit market", with deals done over @SlackHQ , @X DMs and @telegram . In one contract filed with the SEC, the price is blacked out. We followed that into the cost of owning the chips. In the film's model, an H100 hour costs $1.48 assuming a six-year life, or $2.26 assuming three years. We also looked at a five-year loan backed by contracts averaging about three years. What happens after those contracts end decides whether the numbers work. This is the research Forwardhour is built on. We made it with four Grok @bot agents.
1
2
2
2
225
We made an 8 minute investigation into GPU prices. HIVE's annual report lists H200s at $3.99 an hour on demand and $1.45 to $1.85 on six to twelve month contracts. Those are different commitments. The harder part is finding out what other contracts actually pay. A Jump Trading speaker described the term market as an "unlit market", with deals done over @SlackHQ , @X DMs and @telegram . In one contract filed with the SEC, the price is blacked out. We followed that into the cost of owning the chips. In the film's model, an H100 hour costs $1.48 assuming a six-year life, or $2.26 assuming three years. We also looked at a five-year loan backed by contracts averaging about three years. What happens after those contracts end decides whether the numbers work. This is the research Forwardhour is built on. We made it with four Grok @bot agents.
1
2
2
2
225
Forwardhour is about GPU pricing and what owners earn when contracts end. The film's sources include company filings, SemiAnalysis and public rental prices. forwardhour.xyz
2
48