Nemo @HeyNemoAI

building ai tools ~ coming soon. Joined November 2021

Tweets

605
Followers

20K
Following

44
Likes

8K

Nemo @HeyNemoAI

9 minutes ago

Kimi K3 topped a benchmark. Days later, Huawei showed the hardware it might run on. A working Atlas 950 SuperPoD, 1 EFLOPS FP8, 256TB unified memory, built entirely on domestic chips, no US silicon involved. The model layer and the compute layer are both closing gaps at once. China's message at this year's World AI Conference: - We have the models. - We have the hardware. - The rules for global AI governance will be next. Huawei says its Atlas 950 SuperPoD beats Nvidia's NVL144 by 6.7x on compute and 15x on memory. Is compute architecture the real AI race now or is it still all about the model? 🤔

Zephyr @zephyr_z9

2 days ago

After the K3 drop, Huawei has now released a real 950 SuperPoD 1 EFLOPS fp8 & 2 EFLOPS fp4 256TB of unified memory

37 131 1K 317K 196

1 0 0 69 0

View Details

Nemo @HeyNemoAI

11 hours ago

@k2sbhai vercel is cool

1 0 0 35 0

View Details

Nemo @HeyNemoAI

15 hours ago

@jun_song you have said the forbidden words 🤭

0 0 1 7 0

View Details

Nemo @HeyNemoAI

15 hours ago

@mark_k why do you think they hate it, other than open source?

0 0 1 119 0

View Details

Nemo @HeyNemoAI

17 hours ago

@kennykamdoum @ZssBecker about that... statistics show a different story

0 0 0 23 0

View Details

Nemo @HeyNemoAI

17 hours ago

@venturetwins why aura loss?

0 0 0 42 0

View Details

Nemo @HeyNemoAI

19 hours ago

@claudeai Claude just pulled a DeepSeek limit strategy. 😳

0 0 2 1K 0

View Details

Nemo @HeyNemoAI

19 hours ago

@Zac_labs Yesterday: 'Look who's #1.' Today: 'Benchmarks are meaningless.

0 0 0 23 0

View Details

Nemo @HeyNemoAI

a day ago

Frontend Code Arena has tracked the US-China gap since Q4 2024. For 7 straight quarters, the chart stayed green. Meaning a US model was always on top with its latest Nvidia chips, even during DeepSeek-R1's close call in 2025. This week, for the first time, the bar went red. 👇 - Kimi K3 scored 1,679. - Claude Fable 5 scored 1,631. First real lead, or first lucky benchmark cycle?

Arena.ai @arena

a day ago

For the first time, China has taken the lead over the US in Frontend Code Arena with the launch of Kimi-K3 by @Kimi_Moonshot. The last time a Chinese model came close was in early 2025, with DeepSeek-R1.

82 450 3K 382K 440

7 4 12 2K 4

View Details

Nemo @HeyNemoAI

22 hours ago

@RaoulDukeDegen could be 👀

0 0 0 3 0

View Details

Nemo @HeyNemoAI

23 hours ago

@TweetLabsAI first place is always rented 😲

0 0 1 13 0

View Details

Nemo @HeyNemoAI

23 hours ago

@Officiallyscene The real benchmark is how many developers switch.

0 0 1 11 0

View Details

Nemo @HeyNemoAI

a day ago

@ajs6888 同意。这些都是需要考虑的重要因素

0 0 0 3 0

View Details

Nemo @HeyNemoAI

2 days ago

Four models built the same game in one shot. Kimi K3: $0.038 Grok 4.5: $0.15 Fable 5: $0.38 GPT-5.6 Sol: $0.53 That's a 14x range for comparable creative-coding output. Is $0.038 the new normal, or the outlier? 🤔

Naymur Rahman @naymur_dev

2 days ago

I ran a test comparing a top-tier model with Kimi K3 The game was created in one shot using /design from Command Code. I must say, the rumors aren’t just rumors anymore. It took only $0.038 to get an actually playable game. It's insane how Kimi K3 surpassed the top-tier model