Loading video...

Video Failed to Load

Go Home

a single 4090 that's insane

kaiokendev1's profile picture

kaio ken

3,380 subscribers

418,574 views • 2 years ago •via X (Twitter)

GamingScience & TechnologyNews & Politics
LIVE

Anya Rossi• Live Now

Private livecam show

10 Comments

kaio ken's profile picture
kaio ken2 years ago

this is the last time we see this

kayhe's profile picture
kayhe2 years ago

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

kaio ken's profile picture
kaio ken2 years ago

grandma's fb feed about to get even more wild

nion's profile picture
nion2 years ago

them romantic scams boutta get insane

Lotto's profile picture
Lotto2 years ago

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

Jennifer RM's profile picture
Jennifer RM2 years ago

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.

Lelouch 👾's profile picture
Lelouch 👾2 years ago

That’s insane lol

major katsurAGI's profile picture
major katsurAGI2 years ago

is this shit open source?

race's profile picture
race2 years ago

why is the detection of deepfakes so hard?

Thomas Millar's profile picture
Thomas Millar2 years ago

@AiFlux new video template?

Related Videos

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 views • 20 days ago

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 views • 1 year ago

this is a single cut? fucking insane
0:25

this is a single cut? fucking insane

Aman CW: Shippuden

482,401 views • 1 year ago

Who wants a brand new 4090?
0:38

Who wants a brand new 4090?

Doug "Censor" Martin

286,404 views • 3 years ago

That's insane
0:20

That's insane

Enjoy Wild Moments

16,193 views • 1 month ago

That's insane
1:06

That's insane

PEAK

17,993,511 views • 1 year ago

That's insane
1:15

That's insane

The Net Daily

2,328,994 views • 1 year ago

That's insane
0:54

That's insane

PEAK

820,585 views • 1 year ago

That's insane
0:30

That's insane

Crazy Moments

17,396 views • 2 months ago

That's insane
0:37

That's insane

Crazy Moments

209,697 views • 6 months ago

THAT'S INSANE
0:20

THAT'S INSANE

⭒

29,917 views • 2 days ago

That's insane!
1:00

That's insane!

Vince Langman

2,622,710 views • 11 months ago

That's insane
1:41

That's insane

Movies and tvshows

10,778,639 views • 1 year ago

That's pretty insane
1:00

That's pretty insane

Interesting AF

5,271,060 views • 10 months ago

THAT'S INSANE😍
0:36

THAT'S INSANE😍

Interesting things

435,297 views • 4 months ago

That's pretty INSANE
0:56

That's pretty INSANE

Modern History

51,626 views • 1 year ago

That's pretty insane
1:00

That's pretty insane

HOW THINGS WORK

25,519,973 views • 1 year ago

That's insane 👇
0:26

Sensitive content

That's insane 👇

Vince Langman

1,902,227 views • 1 year ago

That's just insane
1:38

That's just insane

PEAK

10,186,225 views • 1 year ago

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 views • 1 year ago