Loading video...

Video Failed to Load

Go Home

Someone built a workstation that needs 4.5 kilowatts. That's more than any standard wall circuit can deliver. Renderboxes Molecule Air: eight RTX 6000 Ada, a Threadripper PRO 7975WX, 768GB of DDR5 ECC 5800, and three Corsair HX1500i supplies stacked together. 🔴 US 15A circuit → 1,800W 🔴 US 20A...

kyzoroXX's profile picture

KyzoroX

1,276 subscribers

21,491 views • 9 days ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

RTX 3090 was never designed to be a server GPU. Yet someone just built a 4-GPU AI workstation around four of them. 4× RTX 3090 96GB VRAM 256GB ECC RAM AMD EPYC 7K62 2000W PSU 11 fans The interesting part isn't the performance. It's the memory. Four 24GB cards can be found for a fraction of what you'd pay for modern datacenter GPUs with similar total VRAM. The catch is obvious: 96GB doesn't behave like one 96GB GPU. You still have to deal with GPU-to-GPU communication, power and heat. But for local AI, this is exactly why the 3090 refuses to disappear. Would you build this today?
0:51

RTX 3090 was never designed to be a server GPU. Yet someone just built a 4-GPU AI workstation around four of them. 4× RTX 3090 96GB VRAM 256GB ECC RAM AMD EPYC 7K62 2000W PSU 11 fans The interesting part isn't the performance. It's the memory. Four 24GB cards can be found for a fraction of what you'd pay for modern datacenter GPUs with similar total VRAM. The catch is obvious: 96GB doesn't behave like one 96GB GPU. You still have to deal with GPU-to-GPU communication, power and heat. But for local AI, this is exactly why the 3090 refuses to disappear. Would you build this today?

KyzoroX

36,030 views • 4 days 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
1:16

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 • 2 months ago

48GB of VRAM rents for $500 a month, and this rig has room to double it. A vertical frame. A bare motherboard. One graphics card, exposed and humming. No rack. No server room. Just a desk. Ryzen 9 9950x. AsRock Taichi X870E. 96GB of DDR5 at 6000MT/s. A Thermalright Royal Pretor 130 Ultra on top. A 1300-watt Super Flower supply feeding all of it. One card. 48GB of VRAM. On the open market, that card alone rents for $0.58 to $0.82 an hour. Run it around the clock, and it clears $500 a month, crunching someone else's AI workloads. The motherboard and the power supply were not picked at random. Both, for one reason: a second GPU. Slide it in, and VRAM doubles to 96GB. So does the money: $1,000 a month, from a machine that fits under a desk. Most people rent that power by the hour. This desk prints it.
0:54

48GB of VRAM rents for $500 a month, and this rig has room to double it. A vertical frame. A bare motherboard. One graphics card, exposed and humming. No rack. No server room. Just a desk. Ryzen 9 9950x. AsRock Taichi X870E. 96GB of DDR5 at 6000MT/s. A Thermalright Royal Pretor 130 Ultra on top. A 1300-watt Super Flower supply feeding all of it. One card. 48GB of VRAM. On the open market, that card alone rents for $0.58 to $0.82 an hour. Run it around the clock, and it clears $500 a month, crunching someone else's AI workloads. The motherboard and the power supply were not picked at random. Both, for one reason: a second GPU. Slide it in, and VRAM doubles to 96GB. So does the money: $1,000 a month, from a machine that fits under a desk. Most people rent that power by the hour. This desk prints it.

AiMind

30,720 views • 19 days ago

Elon Musk: Europe actually has more bureaucracy than the US does because you don't just have the provincial and national level, you also have the EU on top of that And to be totally frank, EU headquarters in Brussels is essentially a giant cathedral to bureaucracy. So if one wishes to visit the Grand Cathedral of Bureaucracy, visit EU headquarters in Brussels I do think that Europe probably needs government efficiency more than the US does. And this needs to happen at the country level and at the EU level. And actually, I think the job is more difficult in Europe than it is in the US I think over-regulation has inhibited a tremendous amount of progress and innovation in Europe. Slow strangulation by over-regulation It's a slow strangulation because it's not like at any one moment you feel like this is it. But the noose around the neck just gets a little tighter every year, and eventually it's going to kill Europe So there must be, in my opinion, immediate action to reduce regulation at the EU level, at the country level, and at the local level
1:10

Elon Musk: Europe actually has more bureaucracy than the US does because you don't just have the provincial and national level, you also have the EU on top of that And to be totally frank, EU headquarters in Brussels is essentially a giant cathedral to bureaucracy. So if one wishes to visit the Grand Cathedral of Bureaucracy, visit EU headquarters in Brussels I do think that Europe probably needs government efficiency more than the US does. And this needs to happen at the country level and at the EU level. And actually, I think the job is more difficult in Europe than it is in the US I think over-regulation has inhibited a tremendous amount of progress and innovation in Europe. Slow strangulation by over-regulation It's a slow strangulation because it's not like at any one moment you feel like this is it. But the noose around the neck just gets a little tighter every year, and eventually it's going to kill Europe So there must be, in my opinion, immediate action to reduce regulation at the EU level, at the country level, and at the local level

X Freeze

891,792 views • 9 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.
1:52

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

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
0:33

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 • 2 months ago

THIS WHOLE 3-NODE CLUSTER RUNS PROXMOX, CEPH AND KUBERNETES - ON 10 WATTS TOTAL. that clip is three zima boards wired to one little switch, and the power meter reads 10 watts for the entire thing. on those three tiny fanless boards: proxmox running a real cluster, three nodes as one datacenter ceph spreading storage across all three, so a node can die and your data doesn't kubernetes on top, kube1/2/3, plus docker and high-availability failover this is the exact stack that runs in datacenters, shrunk onto a shelf. kill one node and the vms migrate to the other two. it heals itself. and it pulls less power than a light bulb. no server rack, no cooling, no power bill you'd ever notice. the uncomfortable part isn't that it's small. it's that "enterprise infrastructure" was mostly the rack and the noise, and none of that was the actual capability. follow RetroChainer for more ways enterprise gear shrinks to a shelf, and bookmark this before you rent another vps.
0:28

THIS WHOLE 3-NODE CLUSTER RUNS PROXMOX, CEPH AND KUBERNETES - ON 10 WATTS TOTAL. that clip is three zima boards wired to one little switch, and the power meter reads 10 watts for the entire thing. on those three tiny fanless boards: proxmox running a real cluster, three nodes as one datacenter ceph spreading storage across all three, so a node can die and your data doesn't kubernetes on top, kube1/2/3, plus docker and high-availability failover this is the exact stack that runs in datacenters, shrunk onto a shelf. kill one node and the vms migrate to the other two. it heals itself. and it pulls less power than a light bulb. no server rack, no cooling, no power bill you'd ever notice. the uncomfortable part isn't that it's small. it's that "enterprise infrastructure" was mostly the rack and the noise, and none of that was the actual capability. follow RetroChainer for more ways enterprise gear shrinks to a shelf, and bookmark this before you rent another vps.

RetroChainer

41,828 views • 1 month ago

⚠️NOBODY AT THIS PARTY IS REACTING TO THE EXPOSED CIRCUIT BOARD IN HER CHEST, AND THAT'S THE PART THAT SHOULD UNSETTLE YOU. She's crouched in the middle of a formal event, dressed and posed like anyone else there, except half her body is peeled back to show mechanical joints, wiring, and a green circuit board sitting exactly where her ribcage should be. The people around her are holding wine glasses and talking, like there's nothing unusual standing a few feet away. That's the actual point of the image, not the machinery itself. It's not asking whether this technology is possible. It's asking what a world looks like once something like this stops getting a second glance. The honest catch: none of this is real. This is AI generated concept art, not a robotics demo and not a person who exists. Nobody has built a body like this, and nobody is walking around a gala with their circuitry showing. But the image is doing something real anyway, it's rehearsing what casual acceptance of a half human, half machine body might eventually look like, long before the technology to build one actually exists. We keep debating robots as something separate from us, a machine on a construction site, a machine mopping a floor. This image skips that separation entirely and imagines the line already gone. If a body like this were real, would it still read as beautiful to you, or does the machinery underneath change how you'd see her?
1:00

⚠️NOBODY AT THIS PARTY IS REACTING TO THE EXPOSED CIRCUIT BOARD IN HER CHEST, AND THAT'S THE PART THAT SHOULD UNSETTLE YOU. She's crouched in the middle of a formal event, dressed and posed like anyone else there, except half her body is peeled back to show mechanical joints, wiring, and a green circuit board sitting exactly where her ribcage should be. The people around her are holding wine glasses and talking, like there's nothing unusual standing a few feet away. That's the actual point of the image, not the machinery itself. It's not asking whether this technology is possible. It's asking what a world looks like once something like this stops getting a second glance. The honest catch: none of this is real. This is AI generated concept art, not a robotics demo and not a person who exists. Nobody has built a body like this, and nobody is walking around a gala with their circuitry showing. But the image is doing something real anyway, it's rehearsing what casual acceptance of a half human, half machine body might eventually look like, long before the technology to build one actually exists. We keep debating robots as something separate from us, a machine on a construction site, a machine mopping a floor. This image skips that separation entirely and imagines the line already gone. If a body like this were real, would it still read as beautiful to you, or does the machinery underneath change how you'd see her?

V0LYX

2,350,792 views • 1 month ago

THE MS-A2 IS THE MINISFORUM BOX HERMES AGENT WAS NEVER MEANT TO RUN ON, AND THAT'S BY DESIGN The MS-S1 MAX earns its spot running Hermes Agent because of one thing. 128GB of unified memory, up to 96GB of it handed straight to the GPU. That's the only reason a 120B model fits and runs locally for $0 a month The MS-A2 solves a different problem. Ryzen 9 9955HX, 16 cores, 32 threads, up to 96GB of regular DDR5-5600, no unified pool Three M.2 PCIe 4.0 slots, one U.2, two 22110. Dual 10Gbps SFP+ LAN plus 2.5G. WiFi 6E. Bluetooth 5.3. A slide-out motherboard for fast upgrades. A real PCIe x16 slot that actually takes a low-profile GPU That last part is where the two machines split for good. The MS-S1 MAX's PCIe slot won't take a GPU at all, every bit of GPU power has to come from the unified chip itself. The MS-A2 trades that unified memory trick for raw expandability instead One box runs a local AI agent. The other runs a home lab that needs storage, networking, and room to grow. Minisforum built both on purpose, not as the same product wearing two names
1:06

THE MS-A2 IS THE MINISFORUM BOX HERMES AGENT WAS NEVER MEANT TO RUN ON, AND THAT'S BY DESIGN The MS-S1 MAX earns its spot running Hermes Agent because of one thing. 128GB of unified memory, up to 96GB of it handed straight to the GPU. That's the only reason a 120B model fits and runs locally for $0 a month The MS-A2 solves a different problem. Ryzen 9 9955HX, 16 cores, 32 threads, up to 96GB of regular DDR5-5600, no unified pool Three M.2 PCIe 4.0 slots, one U.2, two 22110. Dual 10Gbps SFP+ LAN plus 2.5G. WiFi 6E. Bluetooth 5.3. A slide-out motherboard for fast upgrades. A real PCIe x16 slot that actually takes a low-profile GPU That last part is where the two machines split for good. The MS-S1 MAX's PCIe slot won't take a GPU at all, every bit of GPU power has to come from the unified chip itself. The MS-A2 trades that unified memory trick for raw expandability instead One box runs a local AI agent. The other runs a home lab that needs storage, networking, and room to grow. Minisforum built both on purpose, not as the same product wearing two names

NO1ennn

25,645 views • 2 months ago

FOUR 600W GPUS. 384GB OF GDDR7. AND DELL MADE THE INSTALL LOOK EASIER THAN BUILDING A GAMING PC. this is a Dell PowerEdge XE7740 receiving four NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. each card brings: 96GB GDDR7 ECC 24,064 CUDA cores 4 PFLOPS FP4 up to 600W but the most satisfying part is the cable management. every GPU power lead is already routed, labeled and held beside its assigned slot. drop in the card. lock the retention hardware. connect one 16-pin plug. no loose bundle hanging across the airflow path and no guessing which cable belongs where. the four-card build totals 384GB of GPU memory and 16 PFLOPS FP4. the 4U XE7740 supports up to eight of these GPUs, doubling that to 768GB and 32 PFLOPS FP4. honest line: the memory remains split between separate cards, and four GPUs can still pull up to 2.4kW. Dell made a rack-scale AI server this dense without turning the next GPU swap into cable surgery. bookmark & watch today ↓
1:51

FOUR 600W GPUS. 384GB OF GDDR7. AND DELL MADE THE INSTALL LOOK EASIER THAN BUILDING A GAMING PC. this is a Dell PowerEdge XE7740 receiving four NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs. each card brings: 96GB GDDR7 ECC 24,064 CUDA cores 4 PFLOPS FP4 up to 600W but the most satisfying part is the cable management. every GPU power lead is already routed, labeled and held beside its assigned slot. drop in the card. lock the retention hardware. connect one 16-pin plug. no loose bundle hanging across the airflow path and no guessing which cable belongs where. the four-card build totals 384GB of GPU memory and 16 PFLOPS FP4. the 4U XE7740 supports up to eight of these GPUs, doubling that to 768GB and 32 PFLOPS FP4. honest line: the memory remains split between separate cards, and four GPUs can still pull up to 2.4kW. Dell made a rack-scale AI server this dense without turning the next GPU swap into cable surgery. bookmark & watch today ↓

Grimmer

26,693 views • 1 month ago

THIS IS WHERE "THE CLOUD" ACTUALLY LIVES - A ROOM OF DELL STORAGE ARRAYS YOU RENT A SLIVER OF EVERY MONTH. that clip is a real datacenter aisle, rack after rack of dell powerstore arrays, every hexagon panel packed with drives. this is "the cloud." not a metaphor, this room, plus the power, cooling and staff keeping it alive. when you pay for cloud storage, you rent a few gigabytes of these arrays. forever. the bill grows with every photo. here's the part they'd rather you not connect. these enterprise drives run on a schedule. data centers swap them out long before they fail, by the pallet. those decommissioned drives get tested and resold as recertified, the same 20tb enterprise disk for a fraction of new. so the exact hardware humming in this room ends up in a nas on your desk, minus the monthly rent. no cloud subscription, no data on someone else's array, no bill that grows with your library. the uncomfortable part isn't that the cloud is huge. it's that "the cloud" was always just this room, and you can buy a piece of it once instead of renting it forever. follow RetroChainer for what the cloud doesn't want you to picture, and bookmark this before your next storage renewal.
1:10

THIS IS WHERE "THE CLOUD" ACTUALLY LIVES - A ROOM OF DELL STORAGE ARRAYS YOU RENT A SLIVER OF EVERY MONTH. that clip is a real datacenter aisle, rack after rack of dell powerstore arrays, every hexagon panel packed with drives. this is "the cloud." not a metaphor, this room, plus the power, cooling and staff keeping it alive. when you pay for cloud storage, you rent a few gigabytes of these arrays. forever. the bill grows with every photo. here's the part they'd rather you not connect. these enterprise drives run on a schedule. data centers swap them out long before they fail, by the pallet. those decommissioned drives get tested and resold as recertified, the same 20tb enterprise disk for a fraction of new. so the exact hardware humming in this room ends up in a nas on your desk, minus the monthly rent. no cloud subscription, no data on someone else's array, no bill that grows with your library. the uncomfortable part isn't that the cloud is huge. it's that "the cloud" was always just this room, and you can buy a piece of it once instead of renting it forever. follow RetroChainer for what the cloud doesn't want you to picture, and bookmark this before your next storage renewal.

RetroChainer

33,825 views • 1 month ago

AMD JUST KILLED THE $200 - A - MONTH AI SUBSCRIPTION - A 65-WATT BOX WITH 128GB OF UNIFIED MEMORY THAT RUNS 70B MODELS LOCALLY. that clip is amd's ryzen ai max+ 395 mini box, held up on stage. the whole pitch is one spec: 128gb of high-speed unified memory. why it matters: cpu, gpu and npu share the same 128gb pool, so a model that used to need a rack of vram just fits. what it actually runs: llama 70b in full precision on the built-in gpu open models up to ~100b parameters, locally, no cloud all of it in a mini pc pulling around 65 watts that's the part that ends the subscription: your prompts, your data, your model, on a box on your desk, not metered by the token. honest line: this runs open models, not gpt-5 or claude opus. for chat, coding help and agents on your own data it's more than enough. for the absolute frontier, the cloud still wins. and it isn't free: the box runs ~$1,500 and up. but that's a few months of a $200 sub, then it's yours forever. no monthly ai bill, no tokens metered, no prompt leaving the room. the uncomfortable part isn't that amd made a fast box. it's that "you need the cloud for ai" was a pricing choice, and unified memory just made it optional. follow RetroChainer for the hardware that turns rent into ownership, and bookmark this before your next ai renewal.
0:41

AMD JUST KILLED THE $200 - A - MONTH AI SUBSCRIPTION - A 65-WATT BOX WITH 128GB OF UNIFIED MEMORY THAT RUNS 70B MODELS LOCALLY. that clip is amd's ryzen ai max+ 395 mini box, held up on stage. the whole pitch is one spec: 128gb of high-speed unified memory. why it matters: cpu, gpu and npu share the same 128gb pool, so a model that used to need a rack of vram just fits. what it actually runs: llama 70b in full precision on the built-in gpu open models up to ~100b parameters, locally, no cloud all of it in a mini pc pulling around 65 watts that's the part that ends the subscription: your prompts, your data, your model, on a box on your desk, not metered by the token. honest line: this runs open models, not gpt-5 or claude opus. for chat, coding help and agents on your own data it's more than enough. for the absolute frontier, the cloud still wins. and it isn't free: the box runs ~$1,500 and up. but that's a few months of a $200 sub, then it's yours forever. no monthly ai bill, no tokens metered, no prompt leaving the room. the uncomfortable part isn't that amd made a fast box. it's that "you need the cloud for ai" was a pricing choice, and unified memory just made it optional. follow RetroChainer for the hardware that turns rent into ownership, and bookmark this before your next ai renewal.

RetroChainer

16,995 views • 29 days ago

Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology right now. The United States currently consumes roughly 500 gigawatts of electricity on average. To double that capacity which is what continued AI expansion on the current terrestrial trajectory would eventually require would mean building as many power plants as currently exist in the entire country. He is not arguing that this is technically impossible, just that communities are not willing to accept it, that permitting timelines make it unrealistic, and that the hard ceiling on Earth based power generation means the expansion of AI compute will eventually hit a wall that no amount of capital can overcome on the ground. His observation is that in space, that wall does not exist. A solar panel in orbit produces roughly five times more power than the same panel on Earth, operates in continuous sunlight uninterrupted by weather or nighttime, and benefits from the vacuum of space as a completely passive cooling system meaning the two largest operating costs of any terrestrial data center, energy and cooling, are effectively eliminated. He then said that you could theoretically increase harnessed energy by a factor of one million and still be using less than a millionth of the sun's total energy output. This is the underlying physics of why SpaceX filed with the FCC to launch up to one million solar powered AI satellites, and why they described that constellation in their own filing as a first step toward becoming a Kardashev Type II civilization capable of harnessing the full power of the sun. To understand what makes this credible rather than visionary, you need to understand what SpaceX already controls that no other company on earth possesses. Starship, once operating at full cadence, can deliver 100 to 150 tons of payload to orbit per launch, at a target cost per kilogram that is an order of magnitude lower than any existing vehicle. Musk's stated ambition is to scale Starship to 10,000 to 30,000 launches per year, a frequency that would allow the deployment of orbital compute infrastructure at a pace that is currently unimaginable with any existing rocket. He told xAI staff earlier this year that achieving space-based AI at scale will eventually require manufacturing facilities on the moon, building solar panels and heat dissipation structures from lunar silicon and aluminum, and launching them into orbit from there rather than from Earth's surface because the moon's lower gravity makes the economics of launch dramatically more favorable. SpaceX's S-1 filing explicitly states that its launch capabilities could enable massive AI compute satellite constellations with the potential for millions of satellites for orbital data centers, with the first launch potentially occurring as soon as 2028. Google and Alphabet are already in advanced talks with SpaceX about deploying space-based data centers. Starcloud, a startup running Nvidia H100 GPUs in orbit, has already validated that high-performance AI inference workloads can operate in space, with plans to scale to five gigawatts of orbital compute power by 2035. This is why Musk believes the cost crossover happens in two to three years because SpaceX's launch cost trajectory intersects with the accelerating energy constraint on the ground in a way that makes space genuinely cheaper, faster, and less regulated at exactly the moment AI demand is hitting its hardest physical limits.
1:01

Elon Musk just explained why the SpaceX IPO is an energy story and the energy constraint is why he believes space becomes the only viable path for AI to scale (Save this). The argument he is making is one of the most important and least understood things happening in technology right now. The United States currently consumes roughly 500 gigawatts of electricity on average. To double that capacity which is what continued AI expansion on the current terrestrial trajectory would eventually require would mean building as many power plants as currently exist in the entire country. He is not arguing that this is technically impossible, just that communities are not willing to accept it, that permitting timelines make it unrealistic, and that the hard ceiling on Earth based power generation means the expansion of AI compute will eventually hit a wall that no amount of capital can overcome on the ground. His observation is that in space, that wall does not exist. A solar panel in orbit produces roughly five times more power than the same panel on Earth, operates in continuous sunlight uninterrupted by weather or nighttime, and benefits from the vacuum of space as a completely passive cooling system meaning the two largest operating costs of any terrestrial data center, energy and cooling, are effectively eliminated. He then said that you could theoretically increase harnessed energy by a factor of one million and still be using less than a millionth of the sun's total energy output. This is the underlying physics of why SpaceX filed with the FCC to launch up to one million solar powered AI satellites, and why they described that constellation in their own filing as a first step toward becoming a Kardashev Type II civilization capable of harnessing the full power of the sun. To understand what makes this credible rather than visionary, you need to understand what SpaceX already controls that no other company on earth possesses. Starship, once operating at full cadence, can deliver 100 to 150 tons of payload to orbit per launch, at a target cost per kilogram that is an order of magnitude lower than any existing vehicle. Musk's stated ambition is to scale Starship to 10,000 to 30,000 launches per year, a frequency that would allow the deployment of orbital compute infrastructure at a pace that is currently unimaginable with any existing rocket. He told xAI staff earlier this year that achieving space-based AI at scale will eventually require manufacturing facilities on the moon, building solar panels and heat dissipation structures from lunar silicon and aluminum, and launching them into orbit from there rather than from Earth's surface because the moon's lower gravity makes the economics of launch dramatically more favorable. SpaceX's S-1 filing explicitly states that its launch capabilities could enable massive AI compute satellite constellations with the potential for millions of satellites for orbital data centers, with the first launch potentially occurring as soon as 2028. Google and Alphabet are already in advanced talks with SpaceX about deploying space-based data centers. Starcloud, a startup running Nvidia H100 GPUs in orbit, has already validated that high-performance AI inference workloads can operate in space, with plans to scale to five gigawatts of orbital compute power by 2035. This is why Musk believes the cost crossover happens in two to three years because SpaceX's launch cost trajectory intersects with the accelerating energy constraint on the ground in a way that makes space genuinely cheaper, faster, and less regulated at exactly the moment AI demand is hitting its hardest physical limits.

Milk Road AI

12,738 views • 3 months ago

THIS BUILDER JUST DROPPED 256GB OF RAM INTO A MONSTER THREADRIPPER WORKSTATION TO RUN UNFILTERED LOCAL AI Imagine trying to fit a computer setup into a chassis that is basically the size of a mini fridge. That is the Corsair 1000D tower. The builder crammed an ASUS Pro WS WRX80E-SAGE motherboard inside and slapped a 64-core AMD Threadripper PRO 5995WX right into the socket Why go this heavy? Simple. To make sure local open weights models don't instantly choke standard desktop hardware during heavy reasoning tasks Then things get downright ridiculous with the memory configuration. He unboxes eight separate Kingston DDR4 modules, pinning a massive 256 gigabytes of system RAM directly to the board. Having that kind of local memory headroom is an absolute necessity if a team wants to handle massive datasets or run dense training loops without constantly swapping data to the storage drives Speaking of storage, the system relies on two lightning fast 2TB Samsung 990 PRO NVMe drives For the graphics pipeline, he drops in a top-tier ASUS ROG RTX 4090 boasting 24 gigabytes of VRAM. That is pretty much the gold standard right now if you want to run quick local inference cycles and completely stop paying corporate cloud token fees to OpenAI or Anthropic Powering this whole grid requires a monstrous ASUS ROG 1600W Thor Gen 2 power supply. And to prevent the entire workstation from turning into a space heater under full load, the builder went all out with a 360mm AIO liquid cooling setup and an insane cluster of sixteen Lian Li SL-Infinity RGB fans It looks incredibly flashy, probably sounds like a jet engine when the cores push maximum load What to buy for local AI? => my guide below Bookmark this so you don't lose it
1:48

THIS BUILDER JUST DROPPED 256GB OF RAM INTO A MONSTER THREADRIPPER WORKSTATION TO RUN UNFILTERED LOCAL AI Imagine trying to fit a computer setup into a chassis that is basically the size of a mini fridge. That is the Corsair 1000D tower. The builder crammed an ASUS Pro WS WRX80E-SAGE motherboard inside and slapped a 64-core AMD Threadripper PRO 5995WX right into the socket Why go this heavy? Simple. To make sure local open weights models don't instantly choke standard desktop hardware during heavy reasoning tasks Then things get downright ridiculous with the memory configuration. He unboxes eight separate Kingston DDR4 modules, pinning a massive 256 gigabytes of system RAM directly to the board. Having that kind of local memory headroom is an absolute necessity if a team wants to handle massive datasets or run dense training loops without constantly swapping data to the storage drives Speaking of storage, the system relies on two lightning fast 2TB Samsung 990 PRO NVMe drives For the graphics pipeline, he drops in a top-tier ASUS ROG RTX 4090 boasting 24 gigabytes of VRAM. That is pretty much the gold standard right now if you want to run quick local inference cycles and completely stop paying corporate cloud token fees to OpenAI or Anthropic Powering this whole grid requires a monstrous ASUS ROG 1600W Thor Gen 2 power supply. And to prevent the entire workstation from turning into a space heater under full load, the builder went all out with a 360mm AIO liquid cooling setup and an insane cluster of sixteen Lian Li SL-Infinity RGB fans It looks incredibly flashy, probably sounds like a jet engine when the cores push maximum load What to buy for local AI? => my guide below Bookmark this so you don't lose it

beamnxw ./

47,494 views • 1 month ago

"Yields are so low across DeFi right now that they're actually not attractive to traditional investors or Wall Street. You need higher than three or four percent for it to be compelling" Ryan Rasmussen, Head of Research at Bitwise, on why DeFi yields aren't compelling to Wall Street right now "The reality is yields are so low across DeFi right now in most places that they're actually not attractive to traditional investors or to Wall Street. The perceived risk, and there is risk, of lending in DeFi is much higher than what you're being compensated for at two or three or maybe four percent. You need higher than that for it to be compelling" "That's why a lot of focus is on yield. Because there's not a lot of euphoria in onchain activity right now, you don't see the double digit stablecoin yields you typically do in bull markets. I think that's good in the sense that it's driving creativity around strategies that properly manage risk to provide enhanced yield" "But that enhanced yield isn't like 28 percent, which is the type that ultimately blows up post bull markets. It's like eight or nine percent, and that's great for a lot of investors"
1:06

"Yields are so low across DeFi right now that they're actually not attractive to traditional investors or Wall Street. You need higher than three or four percent for it to be compelling" Ryan Rasmussen, Head of Research at Bitwise, on why DeFi yields aren't compelling to Wall Street right now "The reality is yields are so low across DeFi right now in most places that they're actually not attractive to traditional investors or to Wall Street. The perceived risk, and there is risk, of lending in DeFi is much higher than what you're being compensated for at two or three or maybe four percent. You need higher than that for it to be compelling" "That's why a lot of focus is on yield. Because there's not a lot of euphoria in onchain activity right now, you don't see the double digit stablecoin yields you typically do in bull markets. I think that's good in the sense that it's driving creativity around strategies that properly manage risk to provide enhanced yield" "But that enhanced yield isn't like 28 percent, which is the type that ultimately blows up post bull markets. It's like eight or nine percent, and that's great for a lot of investors"

The Wolf Of All Streets

23,598 views • 1 month ago

$15,000 for a Mac sounds insane. Until you realize what Apple is actually selling. 256GB of unified memory. Not a huge GPU farm. Not an NVIDIA rack. Not 4 different machines. Just one box that can keep multiple large LLMs in memory and run them locally. That changes the equation for local AI. NVIDIA wins when you need raw GPU compute. Apple gets interesting when the bottleneck is memory capacity. Because once your model doesn't fit into 24GB, 32GB or 48GB of VRAM, buying a faster GPU doesn't solve the problem. You need more memory. And that's where machines like the Mac Studio start looking less crazy. The interesting question isn't: “Is a $15K Mac faster than an NVIDIA GPU?” It isn't. The real question is: “How much are you willing to pay to make huge local models actually fit?” That’s where unified memory gets interesting.
0:44

$15,000 for a Mac sounds insane. Until you realize what Apple is actually selling. 256GB of unified memory. Not a huge GPU farm. Not an NVIDIA rack. Not 4 different machines. Just one box that can keep multiple large LLMs in memory and run them locally. That changes the equation for local AI. NVIDIA wins when you need raw GPU compute. Apple gets interesting when the bottleneck is memory capacity. Because once your model doesn't fit into 24GB, 32GB or 48GB of VRAM, buying a faster GPU doesn't solve the problem. You need more memory. And that's where machines like the Mac Studio start looking less crazy. The interesting question isn't: “Is a $15K Mac faster than an NVIDIA GPU?” It isn't. The real question is: “How much are you willing to pay to make huge local models actually fit?” That’s where unified memory gets interesting.

KyzoroX

19,536 views • 27 days ago

Arteta on Martin Odegaard: “I think he had a very difficult season in terms of the days that he's missed, the amount of games, which is very unusual. “He's a player that needs that rhythm, that needs to be very connected with the team. And not only that he's missed that opportunity, but the team as well has missed him. “So much for what he brings, that some things that are probably unnoticed to the public or in general are not in those stats, but it's much more profound within the team and the identity of this team. “We need him at his best, we need him constantly with us because we are a much better team with Martin in and around the team. Isaan Khan: “What is it the public don’t notice?” Arteta: “He’s the one that connects everybody, that glides everybody together, that makes the team flow in a way that nobody else can do that, and that's a huge quality to have.”
1:23

Arteta on Martin Odegaard: “I think he had a very difficult season in terms of the days that he's missed, the amount of games, which is very unusual. “He's a player that needs that rhythm, that needs to be very connected with the team. And not only that he's missed that opportunity, but the team as well has missed him. “So much for what he brings, that some things that are probably unnoticed to the public or in general are not in those stats, but it's much more profound within the team and the identity of this team. “We need him at his best, we need him constantly with us because we are a much better team with Martin in and around the team. Isaan Khan: “What is it the public don’t notice?” Arteta: “He’s the one that connects everybody, that glides everybody together, that makes the team flow in a way that nobody else can do that, and that's a huge quality to have.”

Connor Humm

173,568 views • 5 months ago