Загрузка видео...

Не удалось загрузить видео

На главную

a single 4090 that's insane

kaiokendev1's profile picture

kaio ken

3,380 subscribers

418,574 просмотров • 2 лет назад •via X (Twitter)

ИгрыНаука и технологииНовости и политика
LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 10

Фото профиля kaio ken
kaio ken2 лет назад

this is the last time we see this

Фото профиля kayhe
kayhe2 лет назад

not turning on camera on zoom anymore, not taking the risk of someone making a deepfake of me committing to fake ETAs or tickets

Фото профиля kaio ken
kaio ken2 лет назад

grandma's fb feed about to get even more wild

Фото профиля nion
nion2 лет назад

them romantic scams boutta get insane

Фото профиля Lotto
Lotto2 лет назад

Literally have the tech for waifus but won’t ship it, worse than wwiii

Фото профиля Jennifer RM
Jennifer RM2 лет назад

Their eyes are dead. Arguably, this is comforting, because getting the eyes right basically requires a pretty deep theory of mind, and would imply that the image generator was simulating "real"** people for a bit.

Фото профиля Lelouch 👾
Lelouch 👾2 лет назад

That’s insane lol

Фото профиля major katsurAGI
major katsurAGI2 лет назад

is this shit open source?

Фото профиля race
race2 лет назад

why is the detection of deepfakes so hard?

Фото профиля Thomas Millar
Thomas Millar2 лет назад

@AiFlux new video template?

Похожие видео

MoVerse turns an image into a navigable 3D world. Runs at 8 FPS on a single RTX 4090.
0:32

MoVerse turns an image into a navigable 3D world. Runs at 8 FPS on a single RTX 4090.

⚡AI Search⚡

18,555 просмотров • 16 дней назад

GPU tradeoff series: A100 is not much more powerful than 4090 🫠 GPU Perf and Price: - 4090: 330 fp16 TFLOPs, $1,749 - A100 (80GB): 312 fp16 TFLOPs, $20,000 > A100 is 11.4X more pricy Training speed for GPT-2(124M) with llm.c: - 4090: 153K tokens/s - A100 (80GB): 195K tokens/s > A100 is only 1.3X faster (both trained using a single card, A100 llm.c training is shown in the video, 4090 video is in the quoted tweet) Conclusion: 4090 has a much better cost vs performance ratio Why: As in the H100 vs. 4090 comparison, the biggest difference between A100 and 4090 is their GPU memory size/bandwidth and cross-GPU communication bandwidth, which does not matter too much if your model can fit into a single 4090. Specs: 4090: - GPU memory size: 24GB - memory bandwidth: 1 TB/s - communication bandwidth: 64 GB/s A100: - GPU memory size: 80GB - memory bandwidth: 2 TB/s - communication bandwidth: 900 GB/s Nvidia killed off NVLink (a high-speed communication link that connects GPUs) on 4090. (Jensen Huang smiling face) If multiple 4090s could be interconnected via NVLink, their performance would be closer to datacenter-grade A100 GPUs, even for training larger models. Additionally, 4090 isn't allowed in datacenters, that's how Nvidia makes 💰💰💰
0:32

GPU tradeoff series: A100 is not much more powerful than 4090 🫠 GPU Perf and Price: - 4090: 330 fp16 TFLOPs, $1,749 - A100 (80GB): 312 fp16 TFLOPs, $20,000 > A100 is 11.4X more pricy Training speed for GPT-2(124M) with llm.c: - 4090: 153K tokens/s - A100 (80GB): 195K tokens/s > A100 is only 1.3X faster (both trained using a single card, A100 llm.c training is shown in the video, 4090 video is in the quoted tweet) Conclusion: 4090 has a much better cost vs performance ratio Why: As in the H100 vs. 4090 comparison, the biggest difference between A100 and 4090 is their GPU memory size/bandwidth and cross-GPU communication bandwidth, which does not matter too much if your model can fit into a single 4090. Specs: 4090: - GPU memory size: 24GB - memory bandwidth: 1 TB/s - communication bandwidth: 64 GB/s A100: - GPU memory size: 80GB - memory bandwidth: 2 TB/s - communication bandwidth: 900 GB/s Nvidia killed off NVLink (a high-speed communication link that connects GPUs) on 4090. (Jensen Huang smiling face) If multiple 4090s could be interconnected via NVLink, their performance would be closer to datacenter-grade A100 GPUs, even for training larger models. Additionally, 4090 isn't allowed in datacenters, that's how Nvidia makes 💰💰💰

Yuchen Jin

234,951 просмотров • 1 год назад

this is a single cut? fucking insane
0:25

this is a single cut? fucking insane

Aman CW: Shippuden

482,304 просмотров • 1 год назад

Who wants a brand new 4090?
0:38

Who wants a brand new 4090?

Doug "Censor" Martin

286,404 просмотров • 3 лет назад

That's insane
0:20

That's insane

Enjoy Wild Moments

16,193 просмотров • 1 месяц назад

That's insane
1:06

That's insane

PEAK

17,993,511 просмотров • 1 год назад

That's insane
0:37

That's insane

Crazy Moments

209,685 просмотров • 6 месяцев назад

That's insane
0:30

That's insane

Crazy Moments

17,396 просмотров • 2 месяцев назад

That's insane
1:15

That's insane

The Net Daily

2,328,994 просмотров • 1 год назад

That's insane
0:54

That's insane

PEAK

820,585 просмотров • 1 год назад

That's insane!
1:00

That's insane!

Vince Langman

2,622,689 просмотров • 11 месяцев назад

That's insane
1:41

That's insane

Movies and tvshows

10,778,639 просмотров • 1 год назад

That's pretty insane
1:00

That's pretty insane

Interesting AF

5,271,060 просмотров • 10 месяцев назад

THAT'S INSANE😍
0:36

THAT'S INSANE😍

Interesting things

435,297 просмотров • 4 месяцев назад

That's pretty insane
1:00

That's pretty insane

HOW THINGS WORK

25,519,356 просмотров • 1 год назад

That's pretty INSANE
0:56

That's pretty INSANE

Modern History

51,626 просмотров • 1 год назад

That's insane 👇
0:26

Sensitive content

That's insane 👇

Vince Langman

1,902,223 просмотров • 1 год назад

That's just insane
1:38

That's just insane

PEAK

10,186,118 просмотров • 1 год назад

The 9800x3D crushed expectations- the 4090 became the bottleneck. Achieves an unbelievable 400-500 FPS on every single BTB map in Halo Infinite.
0:26

The 9800x3D crushed expectations- the 4090 became the bottleneck. Achieves an unbelievable 400-500 FPS on every single BTB map in Halo Infinite.

XLR8

78,088 просмотров • 1 год назад

That's an insane talent
0:30

That's an insane talent

Community Notes & Violations

24,456 просмотров • 3 месяцев назад