@graykoder
Account based inKorea
About this account
- Account based in
- Korea
- Connected via
- Korea App Store
Account-level information from X, not a live location or the device used for a specific post.
AI & Blockchain Engineer: https://nitter.cf/t.co/MkJcrSYUac
Joined October 2021
- Tweets175
- Following958
- Followers284
- Likes1.3K
Tae Hwan Jung retweeted
언어모델의 계보에 대해 mlxcel 만들면서 삽질했던 내용들을 정리한 발표자료. 누가 어떻게 영향을 받으면서 발전해 오는지에 대해 아키텍처 관점에서 지난 5월에 이야기해 보았었다.
speakerdeck.com/inureyes/llm…
Tae Hwan Jung retweeted
KIMI K3의 역설: KV 캐시는 줄었지만 GPU와 HBM 수요는 더 커진다?
쉽게 말하면: Kimi K3는 데이터를 저장하는 부담은 줄였지만, 모델 자체가 워낙 커서 여러 GPU에 나눠 돌려야 한다. 결국 GPU와 HBM, 네트워크 수요는 줄기는커녕 더 늘어날 수 있다.
기술적으로 말하면: Kimi K3의 KDA는 KV 캐시 전송 부담을 최대 10분의 1로 줄이지만, 2.8조개가 넘는 파라미터와 896개 전문가 모델을 여러 GPU에 나눠 올려야 해 전체 시스템은 오히려 더 많은 GPU, 메모리, 네트워크를 요구한다.
K3의 가중치만 1.5TB 이상의 HBM을 차지한다. 남은 공간이 부족해 KV 캐시는 CPU DDR5와 NVMe로 밀려날 수 있고, WideEP는 수많은 GPU 사이에서 가중치를 주고받기 위해 막대한 네트워크 대역폭을 사용한다.
Kimi도 최적의 추론 환경으로 64개 이상의 가속기를 권장한다. 효율 개선이 하드웨어 수요를 줄이는 게 아니라 AI 사용량과 시스템 규모를 키우는 제본스의 역설이 다시 나타날 수 있다는 이야기다.
*KDA는 이전 내용을 모두 쌓아두는 대신 중요한 정보만 압축해 기억해 KV 캐시 부담을 줄이는 Kimi의 선형 어텐션(linear attention) 기술이지만, 거대한 모델 자체가 요구하는 GPU와 HBM 수요까지 없애지는 않는다.
#KimiK3 #KDA #선형어텐션 #LinearAttention #KV캐시 #KVCache #엔비디아 #NVIDIA #고대역폭메모리 #HBM #디램 #DRAM #AI인프라 #AIInfrastructure #네트워킹 #Networking $NVDA $SKHY $MU $DRAM $SOXX $SMH
Tae Hwan Jung retweeted
If you run this workflow, ask Fable to make codex the workhorse. github.com/steipete/agent-sc…
The most common mistake Asian immigrant founders make in SF is just sitting down and coding.
An accelerator I really respect told me this and I strongly agree. He said outgoing immigrant founders' output is much, much bigger than the introverted ones.
And I'm not telling you to always be active on X. You came here to meet your customers, to learn about a new environment, to meet potential partners. If you just sit down and code all day, you're losing all the opportunities you came here for.
이 시국에 Strategy을 옹호하는 사람이 있다면 항상 근거를 의심해보도록... ㅋㅋ
While you continue to sell your shares of Strategy $MSTR and poke fun at Saylor's "ponzi scheme" institutional investors have bought shares at the fastest pace ever in the last few months
Tae Hwan Jung retweeted
You can run Claude Fable 5 without Anthropic ever knowing it's you.
No email required to sign up.
Here's how to use Fable 5 anonymously through Venice:
I want to say a final thing about my Fable first reaction: I dedicated my life to programming and I'll use every innovation in the field, also to extract value and bring it to the local inference world, to Redis, and so forth. But:
😅
Breaking: FTX founder Sam Bankman-Fried formally applies for a pardon from President Trump
If Sam Bankman-Fried did nothing illegal, he might have been the best VC in history
What SBF bought vs. what it's worth today:
• Cursor: ~$200K → ~$3B (+1,499,900%)
• Anthropic: ~$499M → $82.3B (+16,400%)
• SpaceX: ~$200M → ~$15B (+7,400%)
• Solana: ~$189M → $5.1B (+2,600%)
• Robinhood: $612.5M → $4.9B (+700%)
• Genesis Digital: $1.17B → $3.5B (+200%)
Had he done nothing wrong, he'd have an estimated worth of $114,000,000,000 today
Tae Hwan Jung retweeted
1/3 PropAMM liquidity is now fully operational on Ethereum mainnet!
Three makers are live in every Titan block, and quotes are already consistently beating Binance VIP9 taker fees for retail orders (trades <$1k).
흠 요즘 pypi나 npmjs 매일 털리는거 같은데 왜 이런 오픈소스를 생각한 사람이 없을까요 ㅎㅎ..
github.com/graykode/packvet
after the axios incident, I started building packvet, an ai agent that vets dependencies before install and flags real risks. if you’re into dev tooling / security, let’s talk!
🚨 CRITICAL: Active supply chain attack on axios -- one of npm's most depended-on packages.
The latest [email protected] now pulls in [email protected], a package that did not exist before today. This is a live compromise.
This is textbook supply chain installer malware. axios has 100M+ weekly downloads. Every npm install pulling the latest version is potentially compromised right now.
Socket AI analysis confirms this is malware. plain-crypto-js is an obfuscated dropper/loader that:
• Deobfuscates embedded payloads and operational strings at runtime
• Dynamically loads fs, os, and execSync to evade static analysis
• Executes decoded shell commands
• Stages and copies payload files into OS temp and Windows ProgramData directories
• Deletes and renames artifacts post-execution to destroy forensic evidence
If you use axios, pin your version immediately and audit your lockfiles. Do not upgrade.