Загрузка видео...

Не удалось загрузить видео

На главную

Team PG&E at THRML hackathon estimating that video denoising uses 179000× less energy on Z1 TSU vs H200 GPU 🤯⚡

GillVerd's profile picture

Gill Verdon

64,008 subscribers

15,867 просмотров • 9 месяцев назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

TSU MBB HC, Nolan Smith on facing #2 seed Iowa State in the NCAA Tournament on Friday. "At the end of the day, it's man vs. man, team vs. team, they're not invincible... We feel that our guys gotta go into this thing with the utmost confidence." Tennessee State Men's Basketball WKRN News 2 #RoarCity
0:36

TSU MBB HC, Nolan Smith on facing #2 seed Iowa State in the NCAA Tournament on Friday. "At the end of the day, it's man vs. man, team vs. team, they're not invincible... We feel that our guys gotta go into this thing with the utmost confidence." Tennessee State Men's Basketball WKRN News 2 #RoarCity

Kendall Lanier

38,051 просмотров • 5 месяцев назад

Dereck Lively (right knee) said “time will tell” if he’s feeling good enough to play vs. the Clippers tonight. When asked about Nico Harrison’s firing: “Just things that are out of my control. I gotta be focused on the team, getting this W today & bringing that effort & energy.”
0:48

Dereck Lively (right knee) said “time will tell” if he’s feeling good enough to play vs. the Clippers tonight. When asked about Nico Harrison’s firing: “Just things that are out of my control. I gotta be focused on the team, getting this W today & bringing that effort & energy.”

Mike Curtis

30,008 просмотров • 9 месяцев назад

on the Knicks losing Game 2 at MSG vs. Hawks: "i gotta call Mike Brown out. i did it to Thibodeau ... i gotta keep that same energy (...) there were lineups with no PG. lineups with zero regular season experience (...) you don't experiment in Game 84." CBS Sports
2:05

on the Knicks losing Game 2 at MSG vs. Hawks: "i gotta call Mike Brown out. i did it to Thibodeau ... i gotta keep that same energy (...) there were lineups with no PG. lineups with zero regular season experience (...) you don't experiment in Game 84." CBS Sports

Ashley Nicole Moss

111,963 просмотров • 4 месяцев назад

Extropic founder Beff (e/acc) explains why his new Z1 chip uses 10,000x fewer transistors than a GPU to run the same AI algorithms: "pbits are not traditional bits that are definitely zero, definitely one. They're oscillating between the two. It's like a light switch dimmer. You ever have it flicker when you're at the edge? It's kind of like that." "It's a fractional bit. I could be like, hey, 20% of the time I'm zero, 80% of the time I'm one. That would be a fractional flip." "The most interesting thing is that you can do all sorts of AI algorithms with about 10,000 times less transistors than if you were to do them digitally." "If you go back to the ancestry of neural networks, they came from something called energy-based models from Hinton." "We can do that specifically with Z1. We have literal Boltzmann machines in the hardware. They're very dense, and they use very few transistors, because each transistor is operating in the regime that would be usually considered janky. It's so low power, it's flipping all the time." Extropic
2:54

Extropic founder Beff (e/acc) explains why his new Z1 chip uses 10,000x fewer transistors than a GPU to run the same AI algorithms: "pbits are not traditional bits that are definitely zero, definitely one. They're oscillating between the two. It's like a light switch dimmer. You ever have it flicker when you're at the edge? It's kind of like that." "It's a fractional bit. I could be like, hey, 20% of the time I'm zero, 80% of the time I'm one. That would be a fractional flip." "The most interesting thing is that you can do all sorts of AI algorithms with about 10,000 times less transistors than if you were to do them digitally." "If you go back to the ancestry of neural networks, they came from something called energy-based models from Hinton." "We can do that specifically with Z1. We have literal Boltzmann machines in the hardware. They're very dense, and they use very few transistors, because each transistor is operating in the regime that would be usually considered janky. It's so low power, it's flipping all the time." Extropic

MTS

22,277 просмотров • 27 дней назад

THIS CLUSTER OF 4 MAC STUDIOS USES LESS POWER THAN ONE NVIDIA CHIP - AND COSTS 5X LESS 17:30 the number hits hard - 600 watts for all four Mac Studios - less than one H200 that costs $30,000 on its own instead of one $30,000 GPU you get 4 machines with 1.5TB of memory, 3x more performance and models that simply cannot run on a single chip Apple quietly added RDMA over Thunderbolt in a beta update - latency dropped 99% and the same cluster became 3x faster without buying any new hardware before the update: 5 tokens per second. after: 15.5 - a free upgrade through a software update 8 Nvidia DGX Sparks for the same memory would cost $32,000 - this cluster costs a fraction and gets faster with every update 10x more productive at half the price - and the only thing you need is knowing where to look
20:03

THIS CLUSTER OF 4 MAC STUDIOS USES LESS POWER THAN ONE NVIDIA CHIP - AND COSTS 5X LESS 17:30 the number hits hard - 600 watts for all four Mac Studios - less than one H200 that costs $30,000 on its own instead of one $30,000 GPU you get 4 machines with 1.5TB of memory, 3x more performance and models that simply cannot run on a single chip Apple quietly added RDMA over Thunderbolt in a beta update - latency dropped 99% and the same cluster became 3x faster without buying any new hardware before the update: 5 tokens per second. after: 15.5 - a free upgrade through a software update 8 Nvidia DGX Sparks for the same memory would cost $32,000 - this cluster costs a fraction and gets faster with every update 10x more productive at half the price - and the only thing you need is knowing where to look

Noisy

106,054 просмотров • 2 месяцев назад

Mikel Arteta on the atmosphere at the Emirates Stadium now: “Our 12th man & the magic player. When the stadium provides that level of energy, passion, optimism, playing every single action with the team - the value of that is just infinite.” ❤️🏟️
0:30

Mikel Arteta on the atmosphere at the Emirates Stadium now: “Our 12th man & the magic player. When the stadium provides that level of energy, passion, optimism, playing every single action with the team - the value of that is just infinite.” ❤️🏟️

afcstuff

37,045 просмотров • 3 месяцев назад

#Iowa PG Bennett Stirtz (6-4, 190, SR) Stirtz put together one of his best shot-making displays yet vs. Ole Miss — such a gifted P&R handler (1.389 PPP) that can create at all levels (71.4% rim + 53.6% 3P) & be turnover efficient (3 A/TO). 19.5/3/5.5 on 53.7% FG on the season.
1:28

#Iowa PG Bennett Stirtz (6-4, 190, SR) Stirtz put together one of his best shot-making displays yet vs. Ole Miss — such a gifted P&R handler (1.389 PPP) that can create at all levels (71.4% rim + 53.6% 3P) & be turnover efficient (3 A/TO). 19.5/3/5.5 on 53.7% FG on the season.

Mohamed

31,187 просмотров • 9 месяцев назад

placed 1st, winning $5000 at the Google DeepMind x Cerebral Valley hackathon bangalore! 😭 > we built an RPG game generator that uses Nano Banana 2 Lite to create worlds and assets as you play and progress > you can create your own worlds that are rooted in Indian culture > every asset that you see is generated on the fly > the gameplay assets, dialogues, and clues all change with the theme > for instance, the first visual in the video is from the movie lagaan! insane clutch at the last moment shoutout to the team 💙: Apoorva Fahim Harsh Agarwal
0:32

placed 1st, winning $5000 at the Google DeepMind x Cerebral Valley hackathon bangalore! 😭 > we built an RPG game generator that uses Nano Banana 2 Lite to create worlds and assets as you play and progress > you can create your own worlds that are rooted in Indian culture > every asset that you see is generated on the fly > the gameplay assets, dialogues, and clues all change with the theme > for instance, the first visual in the video is from the movie lagaan! insane clutch at the last moment shoutout to the team 💙: Apoorva Fahim Harsh Agarwal

Parth

145,192 просмотров • 1 месяц назад

The East All-Star Team roster for the upcoming MHSFCA East vs. West All-Star game that will be played on June 24th at Lawrence Technological University. Brought to you by @athletico_pt, Hungry Howie's 🍕 & MHSFCA
1:25

The East All-Star Team roster for the upcoming MHSFCA East vs. West All-Star game that will be played on June 24th at Lawrence Technological University. Brought to you by @athletico_pt, Hungry Howie's 🍕 & MHSFCA

STATE CHAMPS! Michigan

37,956 просмотров • 3 лет назад

Power has been restored for most of the people in San Francisco who were impacted by a massive power outage that left more than 130,000 customers in the dark on Saturday. PG&E blamed the outages on a fire at one of its substations.
1:22

Power has been restored for most of the people in San Francisco who were impacted by a massive power outage that left more than 130,000 customers in the dark on Saturday. PG&E blamed the outages on a fire at one of its substations.

CBS Evening News

17,309 просмотров • 8 месяцев назад

WE'RE BAAAAACK ✌️ with the ONLY #TI12 preshow that matters: Team Liquid SAP | Dota Monster Mornings! Hang with Joey Thimian & Cap LIVE on the Monster Energy Twitch for daily hot takes, special guests & more! It all starts 10/20 at 8am PDT ⏰ FOLLOW NOW ▶️
2:12

WE'RE BAAAAACK ✌️ with the ONLY #TI12 preshow that matters: Team Liquid SAP | Dota Monster Mornings! Hang with Joey Thimian & Cap LIVE on the Monster Energy Twitch for daily hot takes, special guests & more! It all starts 10/20 at 8am PDT ⏰ FOLLOW NOW ▶️

Monster Energy Gaming

105,095 просмотров • 2 лет назад

The West All-Star Team roster for the upcoming Michigan High School Football Coaches Association's East vs. West All-Star game that will be played on June 24th at Lawrence Technological University. @athletico_pt, Hungry Howie's 🍕 & MHSFCA
0:38

The West All-Star Team roster for the upcoming Michigan High School Football Coaches Association's East vs. West All-Star game that will be played on June 24th at Lawrence Technological University. @athletico_pt, Hungry Howie's 🍕 & MHSFCA

STATE CHAMPS! Michigan

11,587 просмотров • 3 лет назад

This power outage is completely unacceptable. PG&E had a fire at this exact substation on the exact date 23 years ago, which left 120,000 San Franciscans in the dark. And before that in 1996! When do we say enough is enough?
1:33

This power outage is completely unacceptable. PG&E had a fire at this exact substation on the exact date 23 years ago, which left 120,000 San Franciscans in the dark. And before that in 1996! When do we say enough is enough?

Saikat Chakrabarti

16,804 просмотров • 8 месяцев назад

.Beff (e/acc) explains how thermodynamic computing fits into the current computing stack: "We don't do GPUs. They're called TSUs, thermodynamic sampling units. We think people are gonna put them next to GPUs or whatever their favorite accelerator is." "You don't need to run the whole workload on the TSU, but it could run certain parts of the workload, just like the GPU is complementary to the CPU initially." "The TSU is gonna help give you more performance per watt in your data center when combined with existing build-outs." "Let's say you're serving a large diffusion model. What we showed is that by using a TSU, you can use something called an energy-based model instead of a simple probabilistic kernel." "Overall you can have basically 100 times the performance per watt of just running on pure GPUs because our chips are much lower wattage."
3:48

.Beff (e/acc) explains how thermodynamic computing fits into the current computing stack: "We don't do GPUs. They're called TSUs, thermodynamic sampling units. We think people are gonna put them next to GPUs or whatever their favorite accelerator is." "You don't need to run the whole workload on the TSU, but it could run certain parts of the workload, just like the GPU is complementary to the CPU initially." "The TSU is gonna help give you more performance per watt in your data center when combined with existing build-outs." "Let's say you're serving a large diffusion model. What we showed is that by using a TSU, you can use something called an energy-based model instead of a simple probabilistic kernel." "Overall you can have basically 100 times the performance per watt of just running on pure GPUs because our chips are much lower wattage."

MTS

42,570 просмотров • 3 месяцев назад

DO TRY THIS AT HOME Salt. Water. Metal =>Light. ⚡🌊💡 The grid taught us to think energy only comes from a plug. Nature reminds us that energy is already moving all around us. This simple saltwater lamp experiment uses an electrolyte solution and two different metals to create an electrochemical reaction—turning stored chemical energy into electricity that can power a small light. Not “free energy.” Not magic. Chemistry. Ingenuity. TreeMindedness. 🌱 The real green pill is learning how energy works, then asking: how can we design systems that are simpler, regenerative, decentralized, and accessible to everybody? Sometimes innovation starts with a jar on the table. 🥦 Stay curious. Stay experimental.
0:37

DO TRY THIS AT HOME Salt. Water. Metal =>Light. ⚡🌊💡 The grid taught us to think energy only comes from a plug. Nature reminds us that energy is already moving all around us. This simple saltwater lamp experiment uses an electrolyte solution and two different metals to create an electrochemical reaction—turning stored chemical energy into electricity that can power a small light. Not “free energy.” Not magic. Chemistry. Ingenuity. TreeMindedness. 🌱 The real green pill is learning how energy works, then asking: how can we design systems that are simpler, regenerative, decentralized, and accessible to everybody? Sometimes innovation starts with a jar on the table. 🥦 Stay curious. Stay experimental.

GP Q

16,090 просмотров • 15 дней назад

Boomers vs Millennials 💰 New research on #GenerationalWealth from Prof James Sefton unpicks the narrative that Boomers have benefited at the expense of Millennials & finds there may be more solidarity at play 🤝 Watch the full video: #IBKnowledge
0:59

Boomers vs Millennials 💰 New research on #GenerationalWealth from Prof James Sefton unpicks the narrative that Boomers have benefited at the expense of Millennials & finds there may be more solidarity at play 🤝 Watch the full video: #IBKnowledge

Imperial Business School

17,232 просмотров • 2 лет назад

Microsoft just made expensive GPUs useless. And it runs on your 5-year-old laptop. BitNet B1.58 crushes models 10x bigger using 96% less energy. Here's what changed: → Uses ternary weights (only -1, 0, +1) instead of full precision floats → 0.4GB memory usage (runs on a phone, no GPU needed) → 29 milliseconds per response (real-time fast) → Beats Llama 3 and Gemma on benchmarks while using almost no power → Free on HuggingFace, MIT license (download and run today) No API costs. No cloud bills. No data leaving your machine. ChatGPT charges per token. This is free forever. Save this video. Want the SOP? DM me. 💬
8:57

Microsoft just made expensive GPUs useless. And it runs on your 5-year-old laptop. BitNet B1.58 crushes models 10x bigger using 96% less energy. Here's what changed: → Uses ternary weights (only -1, 0, +1) instead of full precision floats → 0.4GB memory usage (runs on a phone, no GPU needed) → 29 milliseconds per response (real-time fast) → Beats Llama 3 and Gemma on benchmarks while using almost no power → Free on HuggingFace, MIT license (download and run today) No API costs. No cloud bills. No data leaving your machine. ChatGPT charges per token. This is free forever. Save this video. Want the SOP? DM me. 💬

Julian Goldie SEO

76,615 просмотров • 7 месяцев назад

Testing LCM LORAs in an AnimateDiff & multi-controlnet workflow in ComfyUI. I was able to process this entire Black Pink music video as a single .mp4 input. The LCM lets me render at 6 steps (vs 20+) on my 4090 and uses up only 10.5 GB of VRAM. Here's a breakdown 🧵[1/11]
3:25

Testing LCM LORAs in an AnimateDiff & multi-controlnet workflow in ComfyUI. I was able to process this entire Black Pink music video as a single .mp4 input. The LCM lets me render at 6 steps (vs 20+) on my 4090 and uses up only 10.5 GB of VRAM. Here's a breakdown 🧵[1/11]

CoffeeVectors

182,428 просмотров • 2 лет назад