Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

A TWO FLOOR BITCOIN ASIC FACILITY WITH AN ON SITE TECH FIXING DEAD HASHBOARDS 24/7 STILL PRINTS LESS PER KILOWATT THAN AN 8 CARD RTX 4090 CLOSET ON VAST AI WITH ZERO REPAIR STAFF 01:48 the narrator stops between the racks, "this is the sound of money printing machine,... literally this machine they just print money", he is standing next to $400,000 of stranded SHA-256 ASICs a SHA-256 ASIC at 240 TH/s in 2026 prints $5 a day after electricity in a $0.04/kWh facility, the same operator pays $80,000 a year for the repair tech because the chips literally cook themselves under sustained load ASICs cannot pivot, they only do bitcoin hashing, the whole facility is locked into a single workload that pays 5x less per kilowatt than what a basement RTX 4090 rents for on vast ai right now 8 used RTX 3090s in a residential closet at $0.12/kWh clear $1,200 a month net on vast ai, zero repair staff, zero water cooling pipes, the operator restarts the docker container when something glitches the narrator says "this machine just prints money" while standing next to $400,000 of stranded SHA-256 hardware that cannot serve a single openai customer, the GPU farms three blocks away pivoted in 2024 and never came back bookmark this and read the article belowshow more

starmex

9,671 subscribers

18,281 views • 3 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

14 Comments

Gipp 🦅3 months ago

used video cards will win anyway

starmex3 months ago

fr the secondary market is where the real margins live

Philomena Harry3 months ago

i've never been described so accurately by a stranger on the internet...

Virexontic3 months ago

$400,000 facility. Full-time repair tech. Two floors of hardware. Gets outprinted by 8 GPUs in a residential closet with zero staff. 😅 ASICs can only do one thing. The moment that one thing stopped paying — $400,000 became a very expensive space heater. Flexibility is the moat now. Not hashrate.

wladwtf3 months ago

wonder what the power bill looks like for a setup that size ?

starmex3 months ago

bro the kw efficiency is the whole point, way lower than asics for the same output

アキラ3 months ago

Two floors of Bitcoin ASICs print less than RTX 4090s.

Primee323 months ago

$400k in asics that can't serve a single openai customer. the pivot happened and they missed it

アキラ3 months ago

An interesting look at the efficiency of different mining setups.

SirNastynate3 months ago

@FBIDirectorKash

Virexontic3 months ago

$400,000 facility with a full-time repair tech. Prints less per kilowatt than 8 used GPUs in a residential closet. 😅 "This machine just prints money" he says. Standing next to hardware that can't serve a single AI customer. The sound of money printing is now coming from the basement three blocks away. 👀

Ora3 months ago

the efficiency delta here is actually crazy to think about

SITA3 months ago

flexibility is becoming more valuable than raw hash rate

｡˚ ☁️ ˚｡⋆Bᚱand❃₦⋆｡˚ ☁️ ˚｡3 months ago

$BTC is so uneconomical, adopt $kas FTW

Related Videos

$0.05 PER KWH IN PARAGUAY VS $0.15 IN THE US, A 5,000 GPU FARM NEAR ASUNCION CLEARS $4.7M/MO ON VAST AI WHILE US OPERATORS NET $1.8M ON THE SAME RACK the warehouse in the clip is somewhere outside asuncion, multi story racks running floor to ceiling, the operator built it next to a hydroelectric dam where electricity costs 5 cents per kilowatt hour a US warehouse running the same 5,000 GPU rack burns through $1.6M a month in electricity at $0.15/kWh, the same rack in paraguay burns $530K, the gap is $1M of pure margin every 30 days the racks host RTX 4090s, A100s and H100s on vast ai and clore ai, indie AI startups and image generation services rent them by the hour because aws is rationing GPU access to fortune 500 a 4090 nets the US operator $500 to $1,000 a month, the paraguay operator clears $850 to $1,400 on the same card because power eats half what it does north of the equator bitfarms, riot and core scientific have all expanded into paraguay since 2024, the next 10,000 GPU AI farm coming online in 2026 is more likely to sit near a paraguayan dam than in texas bookmark this and read the article below

$0.05 PER KWH IN PARAGUAY VS $0.15 IN THE US, A 5,000 GPU FARM NEAR ASUNCION CLEARS $4.7M/MO ON VAST AI WHILE US OPERATORS NET $1.8M ON THE SAME RACK the warehouse in the clip is somewhere outside asuncion, multi story racks running floor to ceiling, the operator built it next to a hydroelectric dam where electricity costs 5 cents per kilowatt hour a US warehouse running the same 5,000 GPU rack burns through $1.6M a month in electricity at $0.15/kWh, the same rack in paraguay burns $530K, the gap is $1M of pure margin every 30 days the racks host RTX 4090s, A100s and H100s on vast ai and clore ai, indie AI startups and image generation services rent them by the hour because aws is rationing GPU access to fortune 500 a 4090 nets the US operator $500 to $1,000 a month, the paraguay operator clears $850 to $1,400 on the same card because power eats half what it does north of the equator bitfarms, riot and core scientific have all expanded into paraguay since 2024, the next 10,000 GPU AI farm coming online in 2026 is more likely to sit near a paraguayan dam than in texas bookmark this and read the article below

starmex

27,045 views • 3 months ago

ONE OPERATOR STACKED 300 GPUS ACROSS TWO APARTMENTS IN THE SAME BUILDING AND RUNS A $48K/MONTH AI INFERENCE FARM ON VAST AI FROM HIS LIVING ROOM 00:17 he walks past stacks of GPU boxes, "and probably another 100 GPU boxes in the second apartment, let me know in the comments if you want to see them" he rents 2 units in the same building, one as his living space with 200 GPUs in the bedroom and hallway, the second is dedicated and climate controlled just for the other 100 cards a 300 RTX 4090 setup pulls 135 kilowatts fully loaded, his power bill runs $9,800 a month at $0.10 per kwh, on vast ai the same fleet clears $48,000 in gross monthly rental income he never built this in a warehouse because residential electricity in his city is cheaper than commercial under 150 kw, the split apartment trick keeps him under that ceiling while doubling his rack space the same hardware would have cleared maybe $9,000 a month mining ethereum classic in 2022, vast ai pays 5 times that for AI inference because nobody can ship enough H100s to meet startup demand bookmark this and read the article below

ONE OPERATOR STACKED 300 GPUS ACROSS TWO APARTMENTS IN THE SAME BUILDING AND RUNS A $48K/MONTH AI INFERENCE FARM ON VAST AI FROM HIS LIVING ROOM 00:17 he walks past stacks of GPU boxes, "and probably another 100 GPU boxes in the second apartment, let me know in the comments if you want to see them" he rents 2 units in the same building, one as his living space with 200 GPUs in the bedroom and hallway, the second is dedicated and climate controlled just for the other 100 cards a 300 RTX 4090 setup pulls 135 kilowatts fully loaded, his power bill runs $9,800 a month at $0.10 per kwh, on vast ai the same fleet clears $48,000 in gross monthly rental income he never built this in a warehouse because residential electricity in his city is cheaper than commercial under 150 kw, the split apartment trick keeps him under that ceiling while doubling his rack space the same hardware would have cleared maybe $9,000 a month mining ethereum classic in 2022, vast ai pays 5 times that for AI inference because nobody can ship enough H100s to meet startup demand bookmark this and read the article below

starmex

11,545 views • 3 months ago

WAREHOUSES THE SIZE OF A WALMART WITH 4,000 GPUS WERE EARNING $320K A MONTH MINING ETH, THE SAME RACKS NOW EARN $4M A MONTH RENTING TO AI STARTUPS ON VAST AI the warehouse in the clip is one of hundreds across the US, the operator at 6'6" tall for scale, electricity contracts and cooling already in place ethereum's merge in september 2022 killed gpu mining for everyone without a power plant, by 2025 alt coin mining was barely covering the electricity bill the same racks now host RTX 3090s and 4090s on vast ai, indie AI developers priced out of AWS rent them by the hour for inference and image generation a 4090 nets $500 to $1,000 a month renting to AI startups, the same card was clearing $80 to $150 mining, 4 to 7 times the revenue with zero new hardware core scientific, hut 8 and riot have all signed multi billion dollar AI hosting deals in 2025, the warehouses you scrolled past on tiktok are quietly becoming the inference backbone of every AI tool you use bookmark this and read the article below

WAREHOUSES THE SIZE OF A WALMART WITH 4,000 GPUS WERE EARNING $320K A MONTH MINING ETH, THE SAME RACKS NOW EARN $4M A MONTH RENTING TO AI STARTUPS ON VAST AI the warehouse in the clip is one of hundreds across the US, the operator at 6'6" tall for scale, electricity contracts and cooling already in place ethereum's merge in september 2022 killed gpu mining for everyone without a power plant, by 2025 alt coin mining was barely covering the electricity bill the same racks now host RTX 3090s and 4090s on vast ai, indie AI developers priced out of AWS rent them by the hour for inference and image generation a 4090 nets $500 to $1,000 a month renting to AI startups, the same card was clearing $80 to $150 mining, 4 to 7 times the revenue with zero new hardware core scientific, hut 8 and riot have all signed multi billion dollar AI hosting deals in 2025, the warehouses you scrolled past on tiktok are quietly becoming the inference backbone of every AI tool you use bookmark this and read the article below

starmex

428,022 views • 3 months ago

AMD CEO Lisa Su just killed Nvidia’s $4,000 AI box with a $1,499 lunchbox. She walked on stage, held it in one hand, and ran a 235 billion parameter model live. No data center. No cloud. No rented GPU. The chip inside is something nobody saw coming. AMD’s Ryzen AI Max+ 395 is the first x86 silicon where CPU and GPU share the same 128GB of memory. That single trick lets a desktop run models that used to need a server rack. Out of those 128GB, Linux hands the GPU 110GB to play with. For context, an RTX 5090 gives you 32GB. A 4090 gives you 24. This box gives you more than three times either of them, in a chassis the size of a thick paperback. The benchmark that broke the room: this chip beat an Nvidia RTX 5080 by more than 3x on DeepSeek R1 inference. A $1,499 lunchbox outrunning a $1,000 discrete graphics card on a real AI workload. Nvidia spent a decade convincing the world you needed their hardware for serious AI. AMD just put that on a desk for half the price. Here is what nobody is telling you. A heavy AI user right now pays $200 for Claude Code Max, $200 for ChatGPT Pro, $20 for Cursor, $20 for Gemini. That is $5,280 a year leaving your account. The box pays itself off in 9 months and then runs free for the rest of its life. Install Ollama. Pull Qwen3 235B. Point Claude Code at localhost. Same interface you already use, except now nothing leaves your machine, nothing costs per request, and no company throttles your usage at 3am when you finally have time to build. This is the moment every AI subscription becomes optional. Lawyers stop fearing OpenAI leaks. Developers stop watching the token meter. Founders stop renting H100s for prototypes that never ship because the bill scared them. The first thousand people to figure this out will own the next two years of private AI consulting. Save this, and read the full breakdown article below you are watching the next shift hit before everyone else does.

AMD CEO Lisa Su just killed Nvidia’s $4,000 AI box with a $1,499 lunchbox. She walked on stage, held it in one hand, and ran a 235 billion parameter model live. No data center. No cloud. No rented GPU. The chip inside is something nobody saw coming. AMD’s Ryzen AI Max+ 395 is the first x86 silicon where CPU and GPU share the same 128GB of memory. That single trick lets a desktop run models that used to need a server rack. Out of those 128GB, Linux hands the GPU 110GB to play with. For context, an RTX 5090 gives you 32GB. A 4090 gives you 24. This box gives you more than three times either of them, in a chassis the size of a thick paperback. The benchmark that broke the room: this chip beat an Nvidia RTX 5080 by more than 3x on DeepSeek R1 inference. A $1,499 lunchbox outrunning a $1,000 discrete graphics card on a real AI workload. Nvidia spent a decade convincing the world you needed their hardware for serious AI. AMD just put that on a desk for half the price. Here is what nobody is telling you. A heavy AI user right now pays $200 for Claude Code Max, $200 for ChatGPT Pro, $20 for Cursor, $20 for Gemini. That is $5,280 a year leaving your account. The box pays itself off in 9 months and then runs free for the rest of its life. Install Ollama. Pull Qwen3 235B. Point Claude Code at localhost. Same interface you already use, except now nothing leaves your machine, nothing costs per request, and no company throttles your usage at 3am when you finally have time to build. This is the moment every AI subscription becomes optional. Lawyers stop fearing OpenAI leaks. Developers stop watching the token meter. Founders stop renting H100s for prototypes that never ship because the bill scared them. The first thousand people to figure this out will own the next two years of private AI consulting. Save this, and read the full breakdown article below you are watching the next shift hit before everyone else does.

AdiiX

3,397,308 views • 3 months ago

A FARM OF 90 RTX 3090s BUILT FOR ETHEREUM MINING IN 2021 IS NOW SERVING LOCAL LLM INFERENCE FROM THE SAME GARAGE. THE PART THAT MATTERS ISN'T THE SIZE, IT'S THAT THE OWNER NEVER UNPLUGGED THE RIG, HE JUST CHANGED THE WORKLOAD the camera walks through the room. three custom wood frames stacked floor to ceiling. ninety 3090s breathing rainbow LEDs in unison. a yellow industrial duct ripped through the wall to pull the heat out. each card pulling 250W under load, the whole room drawing 22kW continuous. this is not a hobby setup. it is a small datacenter dressed up as a bedroom do the math. 90 cards times 24GB VRAM is 2,160GB of pooled memory, enough to load every frontier open weight model at once. at $700 used per card, each 3090 pays for itself against a Claude Code Max subscription in about three months. the same RTX 3090 i wrote about last week as the best value per dollar for a local AI rig, stacked 90 times in a garage this is not a crypto relic. it is the new datacenter, built from cards that gamers retired in 2022 and crypto miners gave up on in 2023, now quietly answering prompts that OpenAI charges you $200 a month for

A FARM OF 90 RTX 3090s BUILT FOR ETHEREUM MINING IN 2021 IS NOW SERVING LOCAL LLM INFERENCE FROM THE SAME GARAGE. THE PART THAT MATTERS ISN'T THE SIZE, IT'S THAT THE OWNER NEVER UNPLUGGED THE RIG, HE JUST CHANGED THE WORKLOAD the camera walks through the room. three custom wood frames stacked floor to ceiling. ninety 3090s breathing rainbow LEDs in unison. a yellow industrial duct ripped through the wall to pull the heat out. each card pulling 250W under load, the whole room drawing 22kW continuous. this is not a hobby setup. it is a small datacenter dressed up as a bedroom do the math. 90 cards times 24GB VRAM is 2,160GB of pooled memory, enough to load every frontier open weight model at once. at $700 used per card, each 3090 pays for itself against a Claude Code Max subscription in about three months. the same RTX 3090 i wrote about last week as the best value per dollar for a local AI rig, stacked 90 times in a garage this is not a crypto relic. it is the new datacenter, built from cards that gamers retired in 2022 and crypto miners gave up on in 2023, now quietly answering prompts that OpenAI charges you $200 a month for

Ridark

64,499 views • 3 months ago

NVIDIA'S BIGGEST COMPETITOR AMD BUILT A BOX THAT RUNS A 235B MODEL AND KILLS $200 OPENAI AND $200 CLAUDE CODE FOR $9 A MONTH amd's ryzen ai max+ 395 is the first x86 chip that runs a 200 billion parameter model on a single piece of silicon with 128gb of unified memory the gmktec evo-x2 runs qwen3 235b fully, deepseek v3 comfortably and llama 3.3 70b with headroom. on linux you get 110gb of usable vram out of 128gb amd claimed the chip beat an nvidia rtx 5080 by more than 3x on deepseek r1 inference. a lunchbox sized pc outrunning a $1,000 discrete gpu on a real ai workload a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and the box pays itself off in 9 to 10 months ollama installs in one command and claude code points at localhost so nothing leaves the machine and nothing costs per request bookmark this and read the article below

NVIDIA'S BIGGEST COMPETITOR AMD BUILT A BOX THAT RUNS A 235B MODEL AND KILLS $200 OPENAI AND $200 CLAUDE CODE FOR $9 A MONTH amd's ryzen ai max+ 395 is the first x86 chip that runs a 200 billion parameter model on a single piece of silicon with 128gb of unified memory the gmktec evo-x2 runs qwen3 235b fully, deepseek v3 comfortably and llama 3.3 70b with headroom. on linux you get 110gb of usable vram out of 128gb amd claimed the chip beat an nvidia rtx 5080 by more than 3x on deepseek r1 inference. a lunchbox sized pc outrunning a $1,000 discrete gpu on a real ai workload a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and the box pays itself off in 9 to 10 months ollama installs in one command and claude code points at localhost so nothing leaves the machine and nothing costs per request bookmark this and read the article below

starmex

93,580 views • 3 months ago

$🚨 I DON'T KNOW WHAT PEOPLE ARE WAITING FOR! This guy just deployed a full GPU compute farm in a data center Not renting cloud instances, not paying per-token, just racks of consumer GPUs on warehouse shelving doing AI inference 24/7 The setup is wild: open-frame rigs with rows of white fans, PCIe riser cables, server-grade PSUs, and a PDU for power distribution. All mounted on red industrial shelving with raised-floor cooling underneath. It looks like a crypto mining op, because it probably was one Here's what's actually happening: > Ex-mining rigs are being redeployed for AI inference at scale > Each rig runs 6-8 GPUs on PCIe risers connected to a single motherboard > Fans pull cold air from the raised floor through the open frames > A PDU on each rack handles power distribution across all rigs > The whole thing runs in a colocation facility with cheap power rates The economics make sense: cloud GPU pricing runs $2-4/hr per card. Own the hardware, colocate it for $0.08-0.12/kWh, and you're printing compute at a fraction of the cost The crypto-to-AI pipeline is real. Same hardware, same facilities, same power contracts, completely different revenue model Bookmark this!$

🚨 I DON'T KNOW WHAT PEOPLE ARE WAITING FOR! This guy just deployed a full GPU compute farm in a data center Not renting cloud instances, not paying per-token, just racks of consumer GPUs on warehouse shelving doing AI inference 24/7 The setup is wild: open-frame rigs with rows of white fans, PCIe riser cables, server-grade PSUs, and a PDU for power distribution. All mounted on red industrial shelving with raised-floor cooling underneath. It looks like a crypto mining op, because it probably was one Here's what's actually happening: > Ex-mining rigs are being redeployed for AI inference at scale > Each rig runs 6-8 GPUs on PCIe risers connected to a single motherboard > Fans pull cold air from the raised floor through the open frames > A PDU on each rack handles power distribution across all rigs > The whole thing runs in a colocation facility with cheap power rates The economics make sense: cloud GPU pricing runs $2-4/hr per card. Own the hardware, colocate it for $0.08-0.12/kWh, and you're printing compute at a fraction of the cost The crypto-to-AI pipeline is real. Same hardware, same facilities, same power contracts, completely different revenue model Bookmark this!

cristal

16,305 views • 3 months ago

AMD CEO LISA SU HELD A MINI PC ON STAGE THAT RUNS A 235B MODEL AND REPLACES YOUR $440/MONTH AI STACK amd's ryzen ai max+ 395 is the first x86 chip that runs a 200 billion parameter model on one piece of silicon. cpu and gpu share 128gb of unified memory, no separate graphics card needed the gmktec evo-x2 runs qwen3 235b fully, deepseek v3 comfortably and llama 3.3 70b with headroom. on linux you get 110gb of usable vram out of 128gb amd claimed the chip beat an nvidia rtx 5080 by more than 3x on deepseek r1 inference. a lunchbox sized pc outrunning a $1,000 discrete gpu on a real ai workload a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and the box pays itself off in 9 to 10 months install ollama, pull the model, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

AMD CEO LISA SU HELD A MINI PC ON STAGE THAT RUNS A 235B MODEL AND REPLACES YOUR $440/MONTH AI STACK amd's ryzen ai max+ 395 is the first x86 chip that runs a 200 billion parameter model on one piece of silicon. cpu and gpu share 128gb of unified memory, no separate graphics card needed the gmktec evo-x2 runs qwen3 235b fully, deepseek v3 comfortably and llama 3.3 70b with headroom. on linux you get 110gb of usable vram out of 128gb amd claimed the chip beat an nvidia rtx 5080 by more than 3x on deepseek r1 inference. a lunchbox sized pc outrunning a $1,000 discrete gpu on a real ai workload a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and the box pays itself off in 9 to 10 months install ollama, pull the model, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

starmex

981,550 views • 3 months ago

Good Alexander explains why he thinks OpenAI is today's Amazon “Everyone was in Amazon because of AWS and they all marked the retail business at zero” “That business they thought was worth zero a decade ago is actually this money printing machine because it has an advertising engine. As of Q1 2026 it's a $69B a year run rate advertising business” “OpenAI has three times the number of users as Amazon” “Currently the market narrative and the market consensus is that Sam Altman's an idiot because he has 900 million DAUs and he's losing money on the DAUs” “It's literally the exact same thing that caused everyone to be sidelined in Amazon for like a solid 10 year run. They didn't understand that it's really hard to get 900M users and appealing to normies is not easy”

Good Alexander explains why he thinks OpenAI is today's Amazon “Everyone was in Amazon because of AWS and they all marked the retail business at zero” “That business they thought was worth zero a decade ago is actually this money printing machine because it has an advertising engine. As of Q1 2026 it's a $69B a year run rate advertising business” “OpenAI has three times the number of users as Amazon” “Currently the market narrative and the market consensus is that Sam Altman's an idiot because he has 900 million DAUs and he's losing money on the DAUs” “It's literally the exact same thing that caused everyone to be sidelined in Amazon for like a solid 10 year run. They didn't understand that it's really hard to get 900M users and appealing to normies is not easy”

Jack

24,053 views • 3 months ago

NVIDIA CEO JENSEN HUANG UNVEILED THE RTX 3090 AND THE USED VERSION ON EBAY NOW KILLS A $400/MONTH CLAUDE CODE AND OPENAI BILL the rtx 3090 has 24gb of vram, the same as a new rtx 4090 that costs $2,000. for local ai inference vram matters more than anything else and the 3090 sells for $650-800 used it runs qwen 3.6 27b comfortably, deepseek r1 14b at 60+ tokens per second and mistral 7b for fast automation. 16gb of qwen 3.6 fits with 8gb of headroom for long context alibaba's qwen 3.6 27b beats claude 4.5 opus on realworldqa vision 84.1 vs 77.0 and on ifbench instructions 76.5 vs 58.0. a free open weight model out-benchmarks anthropic's flagship a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and a used 3090 pays itself off in 3.5 months install ollama, pull qwen 3.6 27b, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

NVIDIA CEO JENSEN HUANG UNVEILED THE RTX 3090 AND THE USED VERSION ON EBAY NOW KILLS A $400/MONTH CLAUDE CODE AND OPENAI BILL the rtx 3090 has 24gb of vram, the same as a new rtx 4090 that costs $2,000. for local ai inference vram matters more than anything else and the 3090 sells for $650-800 used it runs qwen 3.6 27b comfortably, deepseek r1 14b at 60+ tokens per second and mistral 7b for fast automation. 16gb of qwen 3.6 fits with 8gb of headroom for long context alibaba's qwen 3.6 27b beats claude 4.5 opus on realworldqa vision 84.1 vs 77.0 and on ifbench instructions 76.5 vs 58.0. a free open weight model out-benchmarks anthropic's flagship a heavy ai user pays $200 for claude code max, $200 for chatgpt pro, $20 for cursor and $20 for gemini. that's $5,280 a year and a used 3090 pays itself off in 3.5 months install ollama, pull qwen 3.6 27b, point claude code at localhost. same interface, nothing leaves the machine, nothing costs per request bookmark this and read the article below

starmex

86,831 views • 3 months ago

A finance professor manages $200M with AI agents, and he told everyone why: "Large language models are at the level of a fourth-year PhD student in every field" Alejandro Lopez-Lira's AI fund, Autopilot, returned 56% last year. The S&P did 16%. There are 52,000 people with money in it, and most of them just watch the machine work. What he automated is the same six-step loop every fund on earth runs: find an idea, code it, backtest it, deploy it, read the autopsy, learn from it. A quant at Two Sigma runs that loop once a month, and the salary time alone costs around $50,000 per hypothesis. All steps from this loop now fit in AI trading text box. Plain English in, executable strategy out, five-year backtest in 12 seconds, live on a broker 90 seconds after you typed the sentence. He runs $200M with AI. You can run same AI fund in two clicks, free to try: Step 6 on this loop is where everyone is stuck. Your agent has no memory. Every strategy it kills goes into a log nobody reads, and the next one starts from zero. Nobody keeps negative results. Not Citadel, not Man Group, not a single repo on GitHub. Fix that and the agent remembers every hypothesis it killed and the regime it died in. It stops burning cycles on your old mistakes. Jane Street pays 3,500 people to run this cycle and made $39.6 billion doing it. Five sixths of it is now free. Bookmark & read full map of this loop in the article below. Most people still think AI trading is out of reach for them - it isn't. Don't want to spend a dollar for testing this? Kalshi just opened a perps exchange and gives US users $25 free to start ->

A finance professor manages $200M with AI agents, and he told everyone why: "Large language models are at the level of a fourth-year PhD student in every field" Alejandro Lopez-Lira's AI fund, Autopilot, returned 56% last year. The S&P did 16%. There are 52,000 people with money in it, and most of them just watch the machine work. What he automated is the same six-step loop every fund on earth runs: find an idea, code it, backtest it, deploy it, read the autopsy, learn from it. A quant at Two Sigma runs that loop once a month, and the salary time alone costs around $50,000 per hypothesis. All steps from this loop now fit in AI trading text box. Plain English in, executable strategy out, five-year backtest in 12 seconds, live on a broker 90 seconds after you typed the sentence. He runs $200M with AI. You can run same AI fund in two clicks, free to try: Step 6 on this loop is where everyone is stuck. Your agent has no memory. Every strategy it kills goes into a log nobody reads, and the next one starts from zero. Nobody keeps negative results. Not Citadel, not Man Group, not a single repo on GitHub. Fix that and the agent remembers every hypothesis it killed and the regime it died in. It stops burning cycles on your old mistakes. Jane Street pays 3,500 people to run this cycle and made $39.6 billion doing it. Five sixths of it is now free. Bookmark & read full map of this loop in the article below. Most people still think AI trading is out of reach for them - it isn't. Don't want to spend a dollar for testing this? Kalshi just opened a perps exchange and gives US users $25 free to start ->

cvxv666

83,045 views • 1 month ago

JENSEN HUANG PUT 1 PETAFLOP AND 128GB OF GPU MEMORY IN A $150,000 BOX. THE $249 JETSON RUNS THE SAME MODELS AND CUTS A $200/MONTH AI BILL TO $2 IN ELECTRICITY the $249 jetson orin nano super runs llama 3, mistral and deepseek locally at 67 trillion operations per second. zero api fees, zero data leaving your machine, $2 a month in electricity andrej karpathy says 99% of ai users are missing 7 basics that have nothing to do with the hardware. a proper config file, a /raw and /wiki folder, saved reference pages. the model stops guessing and starts helping most people pay $200 a month and still get bad output because the problem was never the model. it was the workflow around it. fix the system and a $249 box handles 80% of what you open chatgpt for the jetson breaks even in 6 weeks at $100/month api spend. after that every month is $98 back in your pocket. first year that's over $1,000 saved from a one-time $249 purchase the people building local ai infrastructure in 2026 with a proper workflow are going to look very far ahead in 2028 bookmark this and read the article below

JENSEN HUANG PUT 1 PETAFLOP AND 128GB OF GPU MEMORY IN A $150,000 BOX. THE $249 JETSON RUNS THE SAME MODELS AND CUTS A $200/MONTH AI BILL TO $2 IN ELECTRICITY the $249 jetson orin nano super runs llama 3, mistral and deepseek locally at 67 trillion operations per second. zero api fees, zero data leaving your machine, $2 a month in electricity andrej karpathy says 99% of ai users are missing 7 basics that have nothing to do with the hardware. a proper config file, a /raw and /wiki folder, saved reference pages. the model stops guessing and starts helping most people pay $200 a month and still get bad output because the problem was never the model. it was the workflow around it. fix the system and a $249 box handles 80% of what you open chatgpt for the jetson breaks even in 6 weeks at $100/month api spend. after that every month is $98 back in your pocket. first year that's over $1,000 saved from a one-time $249 purchase the people building local ai infrastructure in 2026 with a proper workflow are going to look very far ahead in 2028 bookmark this and read the article below

starmex

87,065 views • 4 months ago

THIS $1,300 CARD HAS THE SAME 32GB OF VRAM AS A $2,000+ RTX 5090 - FOR LOCAL AI, THAT'S THE SPEC THAT DECIDES EVERYTHING. that clip is the gigabyte radeon ai pro r9700 ai top, 32gb of vram, built for local ai. the pitch is simple: the same memory capacity as an rtx 5090, for roughly a third of what a 5090 actually sells for. here's why that matters, and where it doesn't. for ai, vram is the gate. a model either fits in memory or it doesn't. 32gb loads models a 16gb card can't touch, full stop. so for running local llms and image models, this card does the one thing that counts, cheap. honest line: this is not a 5090 for gaming. the 5090 is far faster at raster and ray tracing. they only match on how much vram they carry. if you want frames, buy the 5090. if you want to fit big models locally, you just found the same 32gb for a third of the money. no $2,000+ price tag, no cloud gpu rental, no model too big for 16gb. the uncomfortable part isn't that amd is cheaper. it's that for local ai you were about to overpay for gaming power you'd never use. follow RetroChainer for more on paying for the spec that matters, and bookmark this before your next gpu.

THIS $1,300 CARD HAS THE SAME 32GB OF VRAM AS A $2,000+ RTX 5090 - FOR LOCAL AI, THAT'S THE SPEC THAT DECIDES EVERYTHING. that clip is the gigabyte radeon ai pro r9700 ai top, 32gb of vram, built for local ai. the pitch is simple: the same memory capacity as an rtx 5090, for roughly a third of what a 5090 actually sells for. here's why that matters, and where it doesn't. for ai, vram is the gate. a model either fits in memory or it doesn't. 32gb loads models a 16gb card can't touch, full stop. so for running local llms and image models, this card does the one thing that counts, cheap. honest line: this is not a 5090 for gaming. the 5090 is far faster at raster and ray tracing. they only match on how much vram they carry. if you want frames, buy the 5090. if you want to fit big models locally, you just found the same 32gb for a third of the money. no $2,000+ price tag, no cloud gpu rental, no model too big for 16gb. the uncomfortable part isn't that amd is cheaper. it's that for local ai you were about to overpay for gaming power you'd never use. follow RetroChainer for more on paying for the spec that matters, and bookmark this before your next gpu.

RetroChainer

38,844 views • 1 month ago

THIS GUY IS WALKING INTO LOCAL SHOPS WITH A MAC MINI IN A HARD CASE AND WALKING OUT WITH $1,200 CHECKS FOR INSTALLING AI THAT RUNS WITHOUT THE INTERNET I had to rewatch this twice because the business model is so stupidly simple my brain kept looking for the catch. He shows up. Plugs in a Mac Mini. Installs a private AI network running entirely on Ollama. The whole thing is done in under an hour. Zero cloud dependency. No API bills. No monthly charges bleeding the client dry. The shop owner gets a local AI system that works even if the WiFi dies, and this guy walks out with $1,200 for an installation that takes less time than a long lunch. The hardware costs $1,300. One deployment and the machine is paid for. Then he locks in a $149 monthly retainer for support, which means every client after the first is pure margin stacking on top of recurring revenue that compounds every single month. Most people hear "AI business" and think they need to build a SaaS platform or learn to code or raise funding from someone. This guy skipped all of that and went straight to walking into physical stores with a box under his arm. This is not going to scale to a billion dollar company. Obviously. But a solo operator clearing $5-10K a month from local installs while everyone else is still arguing about which LLM is best on Twitter is the kind of quiet hustle that does not make headlines until someone does a breakdown of their year-end numbers. The surface area of what AI can do right now is expanding faster than anyone can map it. One guy is doing local hardware installs. Another is generating AI personas that pull five figures a month. Someone else is running AI video pipelines off a laptop. Same wave of tools. None of them competing with each other. Hit play. Watch him unbox the setup and walk through the numbers. Then tell me you could not do this in your city by next Friday.

THIS GUY IS WALKING INTO LOCAL SHOPS WITH A MAC MINI IN A HARD CASE AND WALKING OUT WITH $1,200 CHECKS FOR INSTALLING AI THAT RUNS WITHOUT THE INTERNET I had to rewatch this twice because the business model is so stupidly simple my brain kept looking for the catch. He shows up. Plugs in a Mac Mini. Installs a private AI network running entirely on Ollama. The whole thing is done in under an hour. Zero cloud dependency. No API bills. No monthly charges bleeding the client dry. The shop owner gets a local AI system that works even if the WiFi dies, and this guy walks out with $1,200 for an installation that takes less time than a long lunch. The hardware costs $1,300. One deployment and the machine is paid for. Then he locks in a $149 monthly retainer for support, which means every client after the first is pure margin stacking on top of recurring revenue that compounds every single month. Most people hear "AI business" and think they need to build a SaaS platform or learn to code or raise funding from someone. This guy skipped all of that and went straight to walking into physical stores with a box under his arm. This is not going to scale to a billion dollar company. Obviously. But a solo operator clearing $5-10K a month from local installs while everyone else is still arguing about which LLM is best on Twitter is the kind of quiet hustle that does not make headlines until someone does a breakdown of their year-end numbers. The surface area of what AI can do right now is expanding faster than anyone can map it. One guy is doing local hardware installs. Another is generating AI personas that pull five figures a month. Someone else is running AI video pipelines off a laptop. Same wave of tools. None of them competing with each other. Hit play. Watch him unbox the setup and walk through the numbers. Then tell me you could not do this in your city by next Friday.

Framez

932,332 views • 2 months ago

INTEL JUST SHIPPED A WORKSTATION CARD WITH TWO GPUs ON ONE PCB AND 48GB OF VRAM. FOUR CARDS GIVE A SINGLE MOTHERBOARD 192GB OF POOLED INFERENCE MEMORY FOR THE PRICE OF TWO RTX 5090s 00:14 he holds one card up to the camera, two GPU dies side by side under the cooler, each one running its own x8 PCIe lane back to the chipset the cluster sees eight discrete accelerators in software. intel's Battle Matrix stack shards a model across all eight, so a 235B parameter network loads in slices and answers requests in parallel what 192GB of VRAM unlocks: an entire 200B class model in memory without quantization. a vision agent reading 100 invoices at once. a research box that holds three frontier models loaded simultaneously, switching between them in under a second intel is the slow side of inference. nvidia is faster per token, that is the honest tradeoff. but the only other path to this much VRAM is a $40,000 nvidia rack or three networked Mac Studios four B60 cards plus the chassis lands under $5,000. power draw averages 800 watts, $55 a month in electricity. one engineer paying $400 a month for combined ChatGPT Pro and Claude Code Max pays the hardware off in less than a year

INTEL JUST SHIPPED A WORKSTATION CARD WITH TWO GPUs ON ONE PCB AND 48GB OF VRAM. FOUR CARDS GIVE A SINGLE MOTHERBOARD 192GB OF POOLED INFERENCE MEMORY FOR THE PRICE OF TWO RTX 5090s 00:14 he holds one card up to the camera, two GPU dies side by side under the cooler, each one running its own x8 PCIe lane back to the chipset the cluster sees eight discrete accelerators in software. intel's Battle Matrix stack shards a model across all eight, so a 235B parameter network loads in slices and answers requests in parallel what 192GB of VRAM unlocks: an entire 200B class model in memory without quantization. a vision agent reading 100 invoices at once. a research box that holds three frontier models loaded simultaneously, switching between them in under a second intel is the slow side of inference. nvidia is faster per token, that is the honest tradeoff. but the only other path to this much VRAM is a $40,000 nvidia rack or three networked Mac Studios four B60 cards plus the chassis lands under $5,000. power draw averages 800 watts, $55 a month in electricity. one engineer paying $400 a month for combined ChatGPT Pro and Claude Code Max pays the hardware off in less than a year

NO1ennn

48,247 views • 3 months ago

Gilbert Strang, MIT professor: "Anthropic pays $500,000 for one skill: linear algebra. a professor who gave 66 years to MIT taught all of it in his final lecture, for free." the article above hands you an AI that prints trading models on demand. strip the wrapper off any of them and one object is left. a matrix. Strang spent his whole career on that one thing, and this was the last time he ever wrote it on a chalkboard. look at what he puts up. A x = 0. the null space. it is every mix of your signals that cancels to nothing. an AI will hand you a thousand factors and half of them live right there, pure risk, zero return. a model is a covariance matrix, a set of eigenvectors, one projection. that is the entire machine. no magic, no black box. he was 88, and this was his last class ever. the recording sits free on YouTube while people pay for a worse version of what he gave away. the AI builds the matrix. Strang teaches you to read it. only one of those two skills holds its price.

Gilbert Strang, MIT professor: "Anthropic pays $500,000 for one skill: linear algebra. a professor who gave 66 years to MIT taught all of it in his final lecture, for free." the article above hands you an AI that prints trading models on demand. strip the wrapper off any of them and one object is left. a matrix. Strang spent his whole career on that one thing, and this was the last time he ever wrote it on a chalkboard. look at what he puts up. A x = 0. the null space. it is every mix of your signals that cancels to nothing. an AI will hand you a thousand factors and half of them live right there, pure risk, zero return. a model is a covariance matrix, a set of eigenvectors, one projection. that is the entire machine. no magic, no black box. he was 88, and this was his last class ever. the recording sits free on YouTube while people pay for a worse version of what he gave away. the AI builds the matrix. Strang teaches you to read it. only one of those two skills holds its price.

Rossst.03

69,889 views • 14 days ago

For years, local AI had one answer: buy an NVIDIA GPU. RTX 3090. 24GB VRAM. RTX 4090. 24GB VRAM. RTX 5090. 32GB VRAM. And if you wanted to run something bigger, the price climbed fast. Then this video shows a different reality. An old M1 MacBook Air. 16GB unified memory. A DeepSeek R1 0528 Qwen3 8B model. Just 4.62GB on disk. Then the context gets pushed to almost 50,000 tokens. And the 16GB machine stays in the green. Meanwhile, the 8GB Mac is already sitting at 7.51GB used out of 8GB. That's the part I think people still underestimate. Local AI isn't one giant hardware market anymore. A $4,000 RTX 5090 setup, a 96GB Strix Halo box and an old M1 MacBook aren't competing for the same job. They're different rungs. The mistake is buying hardware for the model you wish you needed instead of the work you actually do. If an 8B model can handle your coding, documents, automation and daily tasks, why are you pricing a 32GB GPU? And that's where the subscription math gets interesting. A lot of people are still paying every month for AI without ever checking what the laptop already sitting on their desk can do. The best local AI machine might not be the fastest one. It might be the one you already own.

For years, local AI had one answer: buy an NVIDIA GPU. RTX 3090. 24GB VRAM. RTX 4090. 24GB VRAM. RTX 5090. 32GB VRAM. And if you wanted to run something bigger, the price climbed fast. Then this video shows a different reality. An old M1 MacBook Air. 16GB unified memory. A DeepSeek R1 0528 Qwen3 8B model. Just 4.62GB on disk. Then the context gets pushed to almost 50,000 tokens. And the 16GB machine stays in the green. Meanwhile, the 8GB Mac is already sitting at 7.51GB used out of 8GB. That's the part I think people still underestimate. Local AI isn't one giant hardware market anymore. A $4,000 RTX 5090 setup, a 96GB Strix Halo box and an old M1 MacBook aren't competing for the same job. They're different rungs. The mistake is buying hardware for the model you wish you needed instead of the work you actually do. If an 8B model can handle your coding, documents, automation and daily tasks, why are you pricing a 32GB GPU? And that's where the subscription math gets interesting. A lot of people are still paying every month for AI without ever checking what the laptop already sitting on their desk can do. The best local AI machine might not be the fastest one. It might be the one you already own.

KyzoroX

67,357 views • 1 month ago

ANTHROPIC LEAKED A FILE THAT TURNS AN EMPTY FOLDER INTO A PRODUCT WITH A PRICE IN 60 MINUTES AND AIMS AT $100,000 A MONTH in one hour you get a product with a price on it - a startup burns a quarter and $80,000 for that same hour problem → shape → surface → price → machine → back to the problem you write only the first 12 minutes: who hurts, and what it costs them right now Opus 5 gives it a shape and then a price, three Sonnet 5 build three versions of the page Haiku 4.5 stands up the machine: delivery, the queue and the weekly numbers $100,000 a month is 1,000 customers at $100, or 10 at $10,000 same number, four completely different businesses which of those four rows you're standing on gets decided in minutes 36-48, not a year from now at minute sixty you have a product, a price and a machine - and 0 customers the hour compresses the first clock to nothing, it never touches the second one the hour makes the product, it does not make the demand save this and paste it into Claude Code - let it build you these 5 blocks ↓

ANTHROPIC LEAKED A FILE THAT TURNS AN EMPTY FOLDER INTO A PRODUCT WITH A PRICE IN 60 MINUTES AND AIMS AT $100,000 A MONTH in one hour you get a product with a price on it - a startup burns a quarter and $80,000 for that same hour problem → shape → surface → price → machine → back to the problem you write only the first 12 minutes: who hurts, and what it costs them right now Opus 5 gives it a shape and then a price, three Sonnet 5 build three versions of the page Haiku 4.5 stands up the machine: delivery, the queue and the weekly numbers $100,000 a month is 1,000 customers at $100, or 10 at $10,000 same number, four completely different businesses which of those four rows you're standing on gets decided in minutes 36-48, not a year from now at minute sixty you have a product, a price and a machine - and 0 customers the hour compresses the first clock to nothing, it never touches the second one the hour makes the product, it does not make the demand save this and paste it into Claude Code - let it build you these 5 blocks ↓

Noisy

14,595 views • 21 days ago

dude runs a private AI host out of his basement on 8 stacked 3090s an accounting firm signed the moment they realized their client financials would physically never touch a third party the rig pays his mortgage now the setup is eight RTX 3090s lined up on an open frame down in his basement - fans roaring, yellow zip ties holding cables, a little monitor blinking stats beside a cheap keyboard. it was built as a mining rig and the dashboard still looks the part the accounting firm didn’t care that it looked like a crypto leftover they cared about one promise no cloud provider could make them - that their clients’ financial records would never leave a machine the firm could physically point to that promise closed the deal here’s the corner they were backed into: accounting firms hold the most sensitive data imaginable - tax filings, payroll, full financial histories. they wanted AI to speed up the tedious work, but routing that data through a cloud model was a liability their partners would never sign. so they sat frozen while the tech moved on without them he handed them the one version that cleared legal: a private model running on hardware in a basement, not a server farm they’d never see the build behind the rig: eight 3090s stack up to 192GB of combined vram - enough to run a 70B model with real context through vLLM. the firm’s documents get indexed into an isolated vector database, so every answer pulls only from their own files. queries hit a local endpoint, replies come back, nothing ever leaves the basement. used 3090s were the quiet genius move - around $700 a card instead of triple that for new silicon the rig that once mined coins for pennies now serves inference that bills like a service the contract that covers his house: → managing partner came down for an in-person demo → asked the only question that mattered: where does the data live → he pointed at the rig and said “right here, nowhere else” → they signed a monthly deal before leaving the basement the economics underneath: → 8 used 3090s: ~$5,600, paid off in the first two months → electricity: a few hundred a month under load → what the firm pays: enough to clear his mortgage every month → margin: almost all of it, because the data center is under his own house real firms burn a fortune on compliant cloud infrastructure he does the same job in a basement, and the data is safer for the dumbest reason possible - it has nowhere else it can go the rig that looks one loose cable from death is quietly the most trusted machine his client has ever touched

dude runs a private AI host out of his basement on 8 stacked 3090s an accounting firm signed the moment they realized their client financials would physically never touch a third party the rig pays his mortgage now the setup is eight RTX 3090s lined up on an open frame down in his basement - fans roaring, yellow zip ties holding cables, a little monitor blinking stats beside a cheap keyboard. it was built as a mining rig and the dashboard still looks the part the accounting firm didn’t care that it looked like a crypto leftover they cared about one promise no cloud provider could make them - that their clients’ financial records would never leave a machine the firm could physically point to that promise closed the deal here’s the corner they were backed into: accounting firms hold the most sensitive data imaginable - tax filings, payroll, full financial histories. they wanted AI to speed up the tedious work, but routing that data through a cloud model was a liability their partners would never sign. so they sat frozen while the tech moved on without them he handed them the one version that cleared legal: a private model running on hardware in a basement, not a server farm they’d never see the build behind the rig: eight 3090s stack up to 192GB of combined vram - enough to run a 70B model with real context through vLLM. the firm’s documents get indexed into an isolated vector database, so every answer pulls only from their own files. queries hit a local endpoint, replies come back, nothing ever leaves the basement. used 3090s were the quiet genius move - around $700 a card instead of triple that for new silicon the rig that once mined coins for pennies now serves inference that bills like a service the contract that covers his house: → managing partner came down for an in-person demo → asked the only question that mattered: where does the data live → he pointed at the rig and said “right here, nowhere else” → they signed a monthly deal before leaving the basement the economics underneath: → 8 used 3090s: ~$5,600, paid off in the first two months → electricity: a few hundred a month under load → what the firm pays: enough to clear his mortgage every month → margin: almost all of it, because the data center is under his own house real firms burn a fortune on compliant cloud infrastructure he does the same job in a basement, and the data is safer for the dumbest reason possible - it has nowhere else it can go the rig that looks one loose cable from death is quietly the most trusted machine his client has ever touched

regent0x

13,334 views • 3 months ago

A finance professor manages $200M with AI agents, and he told everyone why: "Large language models are at the level of a fourth-year PhD student in every field" Alejandro Lopez-Lira's AI fund, Autopilot, returned 56% last year. The S&P did 16%. There are 52,000 people with money in it, and most of them just watch the machine work. What he automated is the same six-step loop every fund on earth runs: find an idea, code it, backtest it, deploy it, read the autopsy, learn from it. A quant at Two Sigma runs that loop once a month, and the salary time alone costs around $50,000 per hypothesis. All steps from this loop now fit in AI trading text box. Plain English in, executable strategy out, five-year backtest in 12 seconds, live on a broker 90 seconds after you typed the sentence. He runs $200M with AI. You can run same AI fund in two clicks, free to try: Step 6 on this loop is where everyone is stuck. Your agent has no memory. Every strategy it kills goes into a log nobody reads, and the next one starts from zero. Nobody keeps negative results. Not Citadel, not Man Group, not a single repo on GitHub. Fix that and the agent remembers every hypothesis it killed and the regime it died in. It stops burning cycles on your old mistakes. Jane Street pays 3,500 people to run this cycle and made $39.6 billion doing it. Five sixths of it is now free. Bookmark & read full map of this loop in the article below. Most people still think AI trading is out of reach for them - it isn't.

A finance professor manages $200M with AI agents, and he told everyone why: "Large language models are at the level of a fourth-year PhD student in every field" Alejandro Lopez-Lira's AI fund, Autopilot, returned 56% last year. The S&P did 16%. There are 52,000 people with money in it, and most of them just watch the machine work. What he automated is the same six-step loop every fund on earth runs: find an idea, code it, backtest it, deploy it, read the autopsy, learn from it. A quant at Two Sigma runs that loop once a month, and the salary time alone costs around $50,000 per hypothesis. All steps from this loop now fit in AI trading text box. Plain English in, executable strategy out, five-year backtest in 12 seconds, live on a broker 90 seconds after you typed the sentence. He runs $200M with AI. You can run same AI fund in two clicks, free to try: Step 6 on this loop is where everyone is stuck. Your agent has no memory. Every strategy it kills goes into a log nobody reads, and the next one starts from zero. Nobody keeps negative results. Not Citadel, not Man Group, not a single repo on GitHub. Fix that and the agent remembers every hypothesis it killed and the regime it died in. It stops burning cycles on your old mistakes. Jane Street pays 3,500 people to run this cycle and made $39.6 billion doing it. Five sixths of it is now free. Bookmark & read full map of this loop in the article below. Most people still think AI trading is out of reach for them - it isn't.

sopersone

48,482 views • 21 days ago