South Korean chip startup Xcena is betting that AI's real bottleneck is not compute, but memory.
This chip startup just raised $135M on a bet that AI's biggest bottleneck isn't compute -- it's memory | TechCrunch
South Korean chip startup Xcena is betting that AI's real bottleneck is not compute, but memory.
techcrunch.com
May 29, 2026 at 12:07 PM
South Korean chip startup Xcena is betting that AI's real bottleneck is not compute, but memory.
Xcena says it reduces data movement to tackle memory bottlenecks. The approach uses the Compute Express Link protocol.
Source: EE Times
Source: EE Times
Xcena Cuts Data Movement to Address Memory Bottlenecks
Xcena says it reduces data movement to tackle memory bottlenecks. The approach uses the Compute Express Link protocol.
www.eetimes.com
September 28, 2026 at 9:42 AM
Xcena says it reduces data movement to tackle memory bottlenecks. The approach uses the Compute Express Link protocol.
Source: EE Times
Source: EE Times
Chip startups are finally challenging NVIDIA's dominance by solving AI's real problem: memory, not compute. XCENA…
https://dev.to/richard_dillon_b9c238186e/ai-weekly-digest-memory-wars-model-upgrades-and-the-trading-benchmark-that-humbled-five-llms-2f66
#AI #MachineLearning
https://dev.to/richard_dillon_b9c238186e/ai-weekly-digest-memory-wars-model-upgrades-and-the-trading-benchmark-that-humbled-five-llms-2f66
#AI #MachineLearning
June 4, 2026 at 7:00 PM
Chip startups are finally challenging NVIDIA's dominance by solving AI's real problem: memory, not compute. XCENA…
https://dev.to/richard_dillon_b9c238186e/ai-weekly-digest-memory-wars-model-upgrades-and-the-trading-benchmark-that-humbled-five-llms-2f66
#AI #MachineLearning
https://dev.to/richard_dillon_b9c238186e/ai-weekly-digest-memory-wars-model-upgrades-and-the-trading-benchmark-that-humbled-five-llms-2f66
#AI #MachineLearning
South Korea's XCENA bets on CXL memory with compute baked in, CEO explains - www.blocksandfiles.com/ai-ml/2026/0...
XCENA bets on CXL memory with compute baked in, CEO explains
XCENA is a South Korean startup focusing on CXL memory pooling and sharing with computational memory ...
www.blocksandfiles.com
March 26, 2026 at 8:37 AM
South Korea's XCENA bets on CXL memory with compute baked in, CEO explains - www.blocksandfiles.com/ai-ml/2026/0...
XcenaとSamsungが、次世代メモリ規格CXLを活用した「Near Memory Compute」デバイスを開発。データ処理をメモリ側で行うことで、AI計算のボトルネックを解消し、演算効率を劇的に向上させます。次世代AIインフラの鍵となる技術に注目です。 #AI #TechNews
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
August 30, 2026 at 9:19 AM
XcenaとSamsungが、次世代メモリ規格CXLを活用した「Near Memory Compute」デバイスを開発。データ処理をメモリ側で行うことで、AI計算のボトルネックを解消し、演算効率を劇的に向上させます。次世代AIインフラの鍵となる技術に注目です。 #AI #TechNews
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
🔥 AI's Memory Bottleneck Solved?
This chip startup, Xcena, has raised $135M to tackle AI's memory issues, which they believe is the biggest bottleneck in AI development. Their innovative...
🌐 aitechcodex.uk/news/1934/?utm_source=bluesky&utm_medium=social&utm_campaign=daily_news 📡 t.me/AITechNewsUK
This chip startup, Xcena, has raised $135M to tackle AI's memory issues, which they believe is the biggest bottleneck in AI development. Their innovative...
🌐 aitechcodex.uk/news/1934/?utm_source=bluesky&utm_medium=social&utm_campaign=daily_news 📡 t.me/AITechNewsUK
June 13, 2026 at 1:00 AM
🔥 AI's Memory Bottleneck Solved?
This chip startup, Xcena, has raised $135M to tackle AI's memory issues, which they believe is the biggest bottleneck in AI development. Their innovative...
🌐 aitechcodex.uk/news/1934/?utm_source=bluesky&utm_medium=social&utm_campaign=daily_news 📡 t.me/AITechNewsUK
This chip startup, Xcena, has raised $135M to tackle AI's memory issues, which they believe is the biggest bottleneck in AI development. Their innovative...
🌐 aitechcodex.uk/news/1934/?utm_source=bluesky&utm_medium=social&utm_campaign=daily_news 📡 t.me/AITechNewsUK
📰 South Korean chip startup Xcena raised $135M to address AI's memory bottleneck, betting that it is a more significant challenge than compute power.
🔗 https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
#Tech #Startups
🔗 https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
#Tech #Startups
This chip startup just raised $135M on a bet that AI's biggest bottleneck isn't compute -- it's memory | TechCrunch
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
techcrunch.com
May 29, 2026 at 1:34 PM
📰 South Korean chip startup Xcena raised $135M to address AI's memory bottleneck, betting that it is a more significant challenge than compute power.
🔗 https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
#Tech #Startups
🔗 https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
#Tech #Startups
South Korean chip startup XCENA just secured $135M in Series B funding, aiming to revolutionize AI memory solutions. As the tech world debates the future of AI, XCENA challenges the norm by focusing on smarter memory over just powerful processors. A game changer…
Read More: https://tscoop.news/Yzq2
Read More: https://tscoop.news/Yzq2
May 30, 2026 at 7:20 PM
South Korean chip startup XCENA just secured $135M in Series B funding, aiming to revolutionize AI memory solutions. As the tech world debates the future of AI, XCENA challenges the norm by focusing on smarter memory over just powerful processors. A game changer…
Read More: https://tscoop.news/Yzq2
Read More: https://tscoop.news/Yzq2
𝗧𝗵𝗶𝘀 𝗰𝗵𝗶𝗽 𝘀𝘁𝗮𝗿𝘁𝘂𝗽 𝗷𝘂𝘀𝘁 𝗿𝗮𝗶𝘀𝗲𝗱 $𝟭𝟯𝟱𝗠 𝗼𝗻 𝗮 𝗯𝗲𝘁 𝘁𝗵𝗮𝘁 𝗔𝗜’𝘀 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝗶𝘀𝗻’𝘁 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 — 𝗶𝘁’𝘀 𝗺𝗲𝗺𝗼𝗿𝘆
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
June 1, 2026 at 3:43 PM
𝗧𝗵𝗶𝘀 𝗰𝗵𝗶𝗽 𝘀𝘁𝗮𝗿𝘁𝘂𝗽 𝗷𝘂𝘀𝘁 𝗿𝗮𝗶𝘀𝗲𝗱 $𝟭𝟯𝟱𝗠 𝗼𝗻 𝗮 𝗯𝗲𝘁 𝘁𝗵𝗮𝘁 𝗔𝗜’𝘀 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝗶𝘀𝗻’𝘁 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 — 𝗶𝘁’𝘀 𝗺𝗲𝗺𝗼𝗿𝘆
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
Telegram AI Digest
#ai #news
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
Telegram AI Digest
#ai #news
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
techcrunch.com
May 30, 2026 at 7:30 AM
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
Telegram AI Digest
#ai #news
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
Telegram AI Digest
#ai #news
𝗧𝗵𝗶𝘀 𝗰𝗵𝗶𝗽 𝘀𝘁𝗮𝗿𝘁𝘂𝗽 𝗷𝘂𝘀𝘁 𝗿𝗮𝗶𝘀𝗲𝗱 $𝟭𝟯𝟱𝗠 𝗼𝗻 𝗮 𝗯𝗲𝘁 𝘁𝗵𝗮𝘁 𝗔𝗜’𝘀 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝗶𝘀𝗻’𝘁 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 — 𝗶𝘁’𝘀 𝗺𝗲𝗺𝗼𝗿𝘆
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
May 31, 2026 at 9:16 PM
𝗧𝗵𝗶𝘀 𝗰𝗵𝗶𝗽 𝘀𝘁𝗮𝗿𝘁𝘂𝗽 𝗷𝘂𝘀𝘁 𝗿𝗮𝗶𝘀𝗲𝗱 $𝟭𝟯𝟱𝗠 𝗼𝗻 𝗮 𝗯𝗲𝘁 𝘁𝗵𝗮𝘁 𝗔𝗜’𝘀 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝗶𝘀𝗻’𝘁 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 — 𝗶𝘁’𝘀 𝗺𝗲𝗺𝗼𝗿𝘆
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
South Korean chip startup XCENA is betting that AI's real bottleneck is not compute, but memory.
This chip startup just raised $135M on a bet that AI's biggest bottleneck isn't compute — it's memory
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
May 30, 2026 at 8:27 AM
This chip startup just raised $135M on a bet that AI's biggest bottleneck isn't compute — it's memory
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
This Chip Startup Just Raised $135M On A Bet That AI’s Biggest Bottleneck Isn’t Compute — It’s Memory
XCENA chip raises $135M to tackle AI memory bottlenecks with near-memory compute innovation reshaping AI infrastructure costs.
DETAILS: www.techticia.com/2026/05/this...
XCENA chip raises $135M to tackle AI memory bottlenecks with near-memory compute innovation reshaping AI infrastructure costs.
DETAILS: www.techticia.com/2026/05/this...
May 30, 2026 at 6:58 AM
This Chip Startup Just Raised $135M On A Bet That AI’s Biggest Bottleneck Isn’t Compute — It’s Memory
XCENA chip raises $135M to tackle AI memory bottlenecks with near-memory compute innovation reshaping AI infrastructure costs.
DETAILS: www.techticia.com/2026/05/this...
XCENA chip raises $135M to tackle AI memory bottlenecks with near-memory compute innovation reshaping AI infrastructure costs.
DETAILS: www.techticia.com/2026/05/this...
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
South Korean chip startup XCENA raised $135 million based on its premise that memory, not compute power, is the primary constraint limiting AI development. The company is…
Via TechCrunch
South Korean chip startup XCENA raised $135 million based on its premise that memory, not compute power, is the primary constraint limiting AI development. The company is…
Via TechCrunch
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
techcrunch.com
May 30, 2026 at 8:31 AM
This chip startup just raised $135M on a bet that AI’s biggest bottleneck isn’t compute — it’s memory
South Korean chip startup XCENA raised $135 million based on its premise that memory, not compute power, is the primary constraint limiting AI development. The company is…
Via TechCrunch
South Korean chip startup XCENA raised $135 million based on its premise that memory, not compute power, is the primary constraint limiting AI development. The company is…
Via TechCrunch
AIのボトルネックは計算能力ではなく「メモリ」にある。この仮説を掲げるAIチップ新興企業が1億3500万ドルを調達。データ転送効率を最大化する次世代アーキテクチャが、AI開発の勢力図を塗り替えるかもしれません。 #AI #TechNews
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
May 29, 2026 at 1:19 PM
AIのボトルネックは計算能力ではなく「メモリ」にある。この仮説を掲げるAIチップ新興企業が1億3500万ドルを調達。データ転送効率を最大化する次世代アーキテクチャが、AI開発の勢力図を塗り替えるかもしれません。 #AI #TechNews
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
https://techcrunch.com/2026/05/29/xcena-secures-135m-at-570m-valuation-betting-on-memory-as-ais-real-bottleneck/
대규모 모델 최적화에 대한 고민, KVCache Meetup에서 함께 나눠요~
대규모 모델 추론 시 병목은 어디에서 주로 발생할까요? 대규모 모델 추론 인프라의 메모리 병목에 대해서 고민해보셨던 실무자들라면, KVCache Meetup Korea와 함께 해주세요!
이번 KVCache Meetup은 추론 비용과 지연시간을 좌우하는 병목이 연산에서 메모리로 옮겨가고 있는 이 때, 어떻게 Cache를 관리할 것인지에 대한 고민과 경험을 나누는 자리입니다. 우리 파이토치 한국 사용자 모임(PyTorchKR)과 CXL 전문 스타트업 엑…
대규모 모델 추론 시 병목은 어디에서 주로 발생할까요? 대규모 모델 추론 인프라의 메모리 병목에 대해서 고민해보셨던 실무자들라면, KVCache Meetup Korea와 함께 해주세요!
이번 KVCache Meetup은 추론 비용과 지연시간을 좌우하는 병목이 연산에서 메모리로 옮겨가고 있는 이 때, 어떻게 Cache를 관리할 것인지에 대한 고민과 경험을 나누는 자리입니다. 우리 파이토치 한국 사용자 모임(PyTorchKR)과 CXL 전문 스타트업 엑…
대규모 모델 최적화에 대한 고민, KVCache Meetup에서 함께 나눠요~
대규모 모델 추론 시 병목은 어디에서 주로 발생할까요? 대규모 모델 추론 인프라의 메모리 병목에 대해서 고민해보셨던 실무자들라면, KVCache Meetup Korea와 함께 해주세요!
이번 KVCache Meetup은 추론 비용과 지연시간을 좌우하는 병목이 연산에서 메모리로 옮겨가고 있는 이 때, 어떻게 Cache를 관리할 것인지에 대한 고민과 경험을 나누는 자리입니다. 우리 파이토치 한국 사용자 모임(PyTorchKR)과 CXL 전문 스타트업 엑시나(XCENA)가 함께 주최하는 이번 행사에 함께 해주세요!
SK Hynix, FuriosaAI, LMCache Lab에서 연사로 함께 참여하는 이번 Meetup에 대한 자세한 정보는 다음과 같습니다:
행사 개요
• 행사명: KV-Cache Meetup Korea
• 일시: 2026년 9월 30일(수) 19:00 \~ 22:00
• 장소: 강남 드림플러스 (서울특별시...
discuss.pytorch.kr
September 15, 2026 at 10:35 AM
대규모 모델 최적화에 대한 고민, KVCache Meetup에서 함께 나눠요~
대규모 모델 추론 시 병목은 어디에서 주로 발생할까요? 대규모 모델 추론 인프라의 메모리 병목에 대해서 고민해보셨던 실무자들라면, KVCache Meetup Korea와 함께 해주세요!
이번 KVCache Meetup은 추론 비용과 지연시간을 좌우하는 병목이 연산에서 메모리로 옮겨가고 있는 이 때, 어떻게 Cache를 관리할 것인지에 대한 고민과 경험을 나누는 자리입니다. 우리 파이토치 한국 사용자 모임(PyTorchKR)과 CXL 전문 스타트업 엑…
대규모 모델 추론 시 병목은 어디에서 주로 발생할까요? 대규모 모델 추론 인프라의 메모리 병목에 대해서 고민해보셨던 실무자들라면, KVCache Meetup Korea와 함께 해주세요!
이번 KVCache Meetup은 추론 비용과 지연시간을 좌우하는 병목이 연산에서 메모리로 옮겨가고 있는 이 때, 어떻게 Cache를 관리할 것인지에 대한 고민과 경험을 나누는 자리입니다. 우리 파이토치 한국 사용자 모임(PyTorchKR)과 CXL 전문 스타트업 엑…
Xcena and Samsung's Near Memory Compute CXL Device https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs (https://news.ycombinator.com/item?id=49496543)
Hot Chips 2026: XCENA and Samsung’s Near-Memory Compute CXL Device
CXL memory expansion, with a side of compute
chipsandcheese.com
August 30, 2026 at 7:18 PM
Xcena and Samsung's Near Memory Compute CXL Device https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs (https://news.ycombinator.com/item?id=49496543)
There's a weird thing happening in compute right now where people keep reinventing blade servers as pcie devices, and I don't get it. I suppose it's cheaper than anything HP ever made.
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
Hot Chips 2026: XCENA and Samsung’s Near-Memory Compute CXL Device
CXL memory expansion, with a side of compute
chipsandcheese.com
August 31, 2026 at 3:34 AM
There's a weird thing happening in compute right now where people keep reinventing blade servers as pcie devices, and I don't get it. I suppose it's cheaper than anything HP ever made.
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
Xcena and Samsung's Near Memory Compute CXL Device
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
August 30, 2026 at 10:35 AM
Xcena and Samsung's Near Memory Compute CXL Device
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
Xcena and Samsung's Near Memory Compute CXL Device
#HackerNews
<a href="https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
#HackerNews
<a href="https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
August 30, 2026 at 9:48 AM
Xcena and Samsung's Near Memory Compute CXL Device
#HackerNews
<a href="https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
#HackerNews
<a href="https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs" class="hover:underline text-blue-600 dark:text-sky-400 no-card-link" target="_blank" rel="noopener" data-link="bsky">https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
Xcena and Samsung's Near Memory Compute CXL Device https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs (https://news.ycombinator.com/item?id=49496543)
Hot Chips 2026: XCENA and Samsung’s Near-Memory Compute CXL Device
CXL memory expansion, with a side of compute
chipsandcheese.com
August 30, 2026 at 12:14 PM
Xcena and Samsung's Near Memory Compute CXL Device https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs (https://news.ycombinator.com/item?id=49496543)
Xcena and Samsung's Near Memory Compute CXL Device
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
August 30, 2026 at 9:00 AM
Xcena and Samsung's Near Memory Compute CXL Device
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
https://chipsandcheese.com/p/hot-chips-2026-xcena-and-samsungs
