3/6 ๐ŸŸข Settimana di fuoco per l'infrastruttura AI:

โ€ข NVIDIA GTC: Jensen Huang presenta nuove piattaforme per agenti AI autonomi
โ€ข Cerebras CS-3 sbarca su AWS Bedrock con inferenza 5x piรน veloce

La corsa ai chip non si ferma. ๐Ÿš€

#Nvidia #GTC2026 #AWS #Cerebras #AI

Andrew Feldman (@andrewdfeldman)

GTC์—์„œ ์  ์Šจ ํ™ฉ์ด NVL 72๊ฐ€ ์„œ๋น„์Šค ์š”๊ตฌ(์˜ˆ: 400โ†’1000 TPS)์—์„œ ๋Œ€์—ญํญ ๋ถ€์กฑ์œผ๋กœ ์„ฑ๋Šฅ ํ•œ๊ณ„์— ๋„๋‹ฌํ•œ๋‹ค๊ณ  ์–ธ๊ธ‰ํ–ˆ์Šต๋‹ˆ๋‹ค. ์ž‘์„ฑ์ž๋Š” ์ด ๋ฐœ์–ธ์„ ์ธ์šฉํ•˜๋ฉฐ NVL 72์˜ ๋Œ€์—ญํญ ์ œ์•ฝ์„ ์ง€์ ํ•˜๊ณ  @cerebras๋ฅผ ์–ธ๊ธ‰ํ•ด ํ•˜๋“œ์›จ์–ด ์ˆ˜์ค€์˜ ๋ณ‘๋ชฉ๊ณผ ๋น„๊ต ๋…ผ์Ÿ์„ ํ™˜๊ธฐํ•ฉ๋‹ˆ๋‹ค.

https://x.com/andrewdfeldman/status/2033654304905785505

#nvidia #gtc #bandwidth #cerebras

Andrew Feldman (@andrewdfeldman) on X

Today at GTC Jensen said โ€œYou wanted to have services that deliver not just 400 TPS, but 1000 T/s. All of a sudden the NVL 72 runs out of steam-we just donโ€™t have enough bandwidth.โ€ Thatโ€™s right. It runs out of bandwidth. @cerebras we have been saying all along that if you buy

X (formerly Twitter)

Amazon ะฑัƒะดะตั‚ ะธัะฟะพะปัŒะทะพะฒะฐั‚ัŒ ะณะธะณะฐะฝั‚ัะบะธะต ั‡ะธะฟั‹ Cerebras ะดะปั ั€ะฐะฑะพั‚ั‹ ะผะพะดะตะปะตะน ะธัะบัƒััั‚ะฒะตะฝะฝะพะณะพ ะธะฝั‚ะตะปะปะตะบั‚ะฐ. https://webno.ru/novosti-i-stati/amazon-budet-ispolzovat-gigantskie-chipy-cerebras-dlya-raboty-modelej-iskusstvennogo-intellekta/

#amazon #cerebras

SRAM. Static RAM. The stuff used for CPU caches, including AMD's 3D chips.

There have been mumbles of CPU prices spiking like DRAM....this may be part of why.

"Companies like Cerebras, Groq, and d-Matrix are designing AI inference chips that use massive amounts of on-chip SRAM instead of relying on external DRAM (HBM), which significantly reduces latency and power consumption."

nVidia bought Groq. Amazon and Cerebras just signed a deal. Cerebrasโ€™ WSE-3 chip includes 900,000 cores and 44 gigabytes of on-chip SRAM.

Wait for it...............

#ai #dram #memory #sram #datacenters #gpu #cerebras #groq #amazon #nvidia

Cerebras (@cerebras)

์ฝ”๋”ฉ ์›Œํฌํ”Œ๋กœ์šฐ๊ฐ€ ๋А๋ฆฐ ์ถ”๋ก ์„ ์ „์ œ๋กœ ์„ค๊ณ„๋˜์–ด ์žˆ์—ˆ์œผ๋‚˜, OpenAI์˜ 'Codex Spark'๊ฐ€ Cerebras ํ•˜๋“œ์›จ์–ด๋กœ ๊ตฌ๋™๋˜๋ฉฐ ์ƒํ™ฉ์„ ๋ฐ”๊ฟจ๋‹ค๋Š” ์ฃผ์žฅ์ž…๋‹ˆ๋‹ค. ์ดˆ๋‹น 1,200 ํ† ํฐ์ด๋ผ๋Š” ๋†’์€ ์ฒ˜๋ฆฌ๋Ÿ‰์„ ํ™œ์šฉํ•ด ์‹ค๋ฌด ์ฝ”๋“œ ์ž‘์—…์—์„œ ์„ฑ๋Šฅ์„ ๋Œ์–ด๋‚ด๋Š” ๋ฐฉ๋ฒ•์„ ๊ณต์œ ํ•œ๋‹ค๊ณ  ํ•ฉ๋‹ˆ๋‹ค.

https://x.com/cerebras/status/2032173925577412657

#openai #codexspark #cerebras #inference #coding

Sarah Chieng (@SarahChieng)

์ž‘์„ฑ์ž๋Š” Codex์™€ Claude Code์— ์ž์‹ ์˜ ํŠธ์œ„ํ„ฐ ๊ณ„์ • ์ „์ฒด ์ ‘๊ทผ๊ถŒ์„ ์ค˜ ๋น„๊ตํ–ˆ๋‹ค. ํŠธ์œ—์— ๋”ฐ๋ฅด๋ฉด 'Codex spark'(Cerebras ๊ธฐ๋ฐ˜)๋Š” Claude Code๋ณด๋‹ค GTM ์•„์›ƒ๋ฆฌ์น˜ ์›Œํฌํ”Œ๋กœ๋ฅผ 3๋ฐฐ ๋น ๋ฅด๊ฒŒ ์™„๋ฃŒํ–ˆ๊ณ  ํˆด ํ˜ธ์ถœ ํšŸ์ˆ˜๋„ ํ›จ์”ฌ ์ ์—ˆ๋‹ค๊ณ  ์ฃผ์žฅํ•˜์—ฌ, ์‹ค๋ฌด ์„ฑ๋Šฅยทํšจ์œจ์„ฑ ๋น„๊ต ์‚ฌ๋ก€๋ฅผ ์ œ์‹œํ–ˆ๋‹ค.

https://x.com/SarahChieng/status/2031464651519832209

#codex #claudecode #cerebras #ai #benchmark

Sarah Chieng (@SarahChieng) on X

I gave both Codex and Claude Code full access to my twitter account. Codex spark (powered by Cerebras) did a full GTM outreach workflow 3x FASTER than Claude code, and used way less tool calls.

X (formerly Twitter)

Tommy Nguyen (@tuannvm)

์ž˜ ์„ค๊ณ„๋œ AI ์›Œํฌํ”Œ๋กœ๋ฅผ ์ตœ๋Œ€ํ•œ ํ™œ์šฉํ•˜๋Š” ๊ฒƒ์„ ๋†’์ด ํ‰๊ฐ€ํ•˜๋ฉฐ, @cerebras(๋ฆฌ๋”์‹ญ/์ œํ’ˆ)๊ฐ€ ์ƒˆ๋กœ์šด ์ง€ํ‰์„ ์—ด์—ˆ๋‹ค๊ณ  ์นญ์ฐฌโ€”Cerebras์˜ ํ•˜๋“œ์›จ์–ดยท์‹œ์Šคํ…œ์ด AI ์›Œํฌํ”Œ๋กœ์˜ ์„ฑ๋Šฅยทํšจ์œจ ๊ฐœ์„ ์— ๊ธฐ์—ฌํ•˜๊ณ  ์žˆ์Œ์„ ์‹œ์‚ฌํ•ฉ๋‹ˆ๋‹ค.

https://x.com/tuannvm/status/2029923941427605525

#cerebras #aiworkflows #hardware #accelerator

Tommy Nguyen (@tuannvm) on X

I'm a big fan of making the most out of solid & well-designed AI workflows, and @cerebras has opened a new frontier!

X (formerly Twitter)

OpenAI has launched GPT-5.3-Codex-Spark - its first production #AI model deployed on #Cerebras wafer-scale chips instead of traditional Nvidia GPUs.

The resultโ“
โ€ข Higher throughput
โ€ข Lower latency
โ€ข A real-time, interactive coding experience

๐Ÿ”— Read more on #InfoQ โ‡จ https://bit.ly/3MKZQ17

#OpenAI #ChatGPT #LLMs #AIagents

Ray Fernando (@RayFernando1337)

Cerebras๊ฐ€ GTC ๊ธฐ๊ฐ„์— ์ง„ํ–‰ํ•˜๋Š” 'Cafe Compute' ์ฐธ์—ฌ๋ฅผ ๊ถŒ์œ ํ•˜๋Š” ํŠธ์œ—์ž…๋‹ˆ๋‹ค. Cerebras๊ฐ€ ์ฃผ์ตœํ•˜๋Š” ํ–‰์‚ฌ/์„ธ์…˜์— ์ดˆ๋Œ€ํ•˜๋ฉฐ GTC์—์„œ ์ค‘์š”ํ•œ ํ•œ ์ฃผ๊ฐ€ ๋  ๊ฒƒ์ด๋ผ๊ณ  ์•Œ๋ฆฌ๊ณ  ์žˆ์–ด ํ–‰์‚ฌ ์ฐธ๊ฐ€์™€ ๋„คํŠธ์›Œํ‚น ๊ธฐํšŒ๋ฅผ ํ™๋ณดํ•ฉ๋‹ˆ๋‹ค.

https://x.com/RayFernando1337/status/2028631382264836260

#cerebras #cafecompute #gtc #ai

Ray Fernando (@RayFernando1337) on X

Who wants in to Cafe Compute by Cerebras? It is going to be a huge week at GTC and would be fun for you to join me! https://t.co/Q84we4F4B0

X (formerly Twitter)