#hbm4 — Public Fediverse posts

RT @SebAaltonen: Die AMD MI4555X scheint großartig zu sein: 40 Petaflops MXFP6 (gemeinsamer Exponent → Qualität nahe an fp8). 432 GB RAM mit ~20 TB/s. MXFP6-quantisiertes Kimi K3 sollte auf 6 GPUs passen. Das sind rund 300.000 US-Dollar für unbegrenzte Frontier-Modell-Token für etwa 10 Entwickler… videocardz.com/newz/amd-inst… Link AMD Instinct MI455X GPU: 432 GB HBM4 und 23,3 TB/s Speicherbandbreite - VideoCardz.com Die neue Instinct-GPU verwendet ein Chiplet-Paket mit zwei Fabric-Dies und zwei I/O-Dies. Die maximale Speicherbandbreite beträgt 23,3 TB/s. videocardz.com

mehr auf Arint.info

#AI #AMD #GPU #HBM4 #MI455X #MXFP6 #arint_info

https://x.com/SebAaltonen/status/2080496707406909455#m

RT @SebAaltonen: Die AMD MI4555X scheint großartig zu sein: 40 Petaflops MXFP6 (gemeinsamer Exponent → Qualität nahe an fp8). 432 GB RAM mit ~20 TB/s. MXFP6-quantisiertes Kimi K3 sollte auf 6 GPUs passen. Das sind rund 300.000 US-Dollar für unbegrenzte Frontier-Modell-Token für etwa 10 Entwickler… videocardz.com/newz/amd-inst… Link AMD Instinct MI455X GPU: 432 GB HBM4 und 23,3 TB/s Speicherbandbreite - VideoCardz.com Die neue Instinct-GPU verwendet ein Chiplet-Paket mit zwei Fabric-Dies und zwei I/O-Dies. Die maximale Speicherbandbreite beträgt 23,3 TB/s. videocardz.com

mehr auf Arint.info

#AI #AMD #GPU #HBM4 #MI455X #MXFP6 #arint_info

https://x.com/SebAaltonen/status/2080496707406909455#m

#Nvidia and #SKGroup announced a $500 billion #AIinitiative, including a #partnership with #SKHynix for next-generation memory and a 2-gigawatt AI data centre powered by Nvidia’s #VeraRubin #chips and SK Hynix’s #HBM4 #memory. https://www.reuters.com/business/media-telecom/nvidia-sk-group-unveil-500-billion-plus-ai-data-centers-initiative-memory-2026-07-24/?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI

#Nvidia and #SKGroup announced a $500 billion #AIinitiative, including a #partnership with #SKHynix for next-generation memory and a 2-gigawatt AI data centre powered by Nvidia’s #VeraRubin #chips and SK Hynix’s #HBM4 #memory. https://www.reuters.com/business/media-telecom/nvidia-sk-group-unveil-500-billion-plus-ai-data-centers-initiative-memory-2026-07-24/?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI

RT @wccftech: TRANSLASHT: AMD bringt die Instinct MI455X GPU auf den Markt, einen 320 Milliarden Transistoren umfassenden Riesen, der entwickelt wurde, um NVIDIA's Rubin zu bekämpfen, mit 50 % mehr HBM4-Speicher und bis zu 40 PFLOPs an KI-Computing-Leistung. wccftech.com/amd-instinct-mi… Link AMD veröffentlicht Instinct MI455X GPU, ein 320 Milliarden Transistoren großes Monster, das entwickelt wurde, um... AMDs Instinct MI455X-GPUs erweitern die KI-Roadmap des Unternehmens und bieten führende HBM4-Kapazitäten und über 40 PFLOPs an Rechenleistung für Agentic AI, im Wettbewerb mit NVIDIAs Rubin-Chip. AMD hat eine Antwort auf NVIDIA's... wccftech.com

mehr auf Arint.info

#AICompute #AMD #HBM4 #InstinctMI455X #NVIDIA #Rivalry #arint_info

https://x.com/wccftech/status/2080394171446067351#m

Dentro la gerarchia della memoria delle GPU: come i server AI spostano i dati dagli SSD alla HBM

Ti sei mai chiesto come i modelli di intelligenza artificiale trasferiscono i dati alla GPU? Questo articolo spiega in modo semplice i diversi livelli di memoria all'interno di un server AI, perché la memoria della GPU è diventata un collo di bottiglia e quali nuove tecnologie potrebbero migliorare le prestazioni dell'intelligenza artificiale.

buysellram.com/blog/inside-the…

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD

Dentro la gerarchia della memoria delle GPU: come i server AI spostano i dati dagli SSD alla HBM

Ti sei mai chiesto come i modelli di intelligenza artificiale trasferiscono i dati alla GPU? Questo articolo spiega in modo semplice i diversi livelli di memoria all'interno di un server AI, perché la memoria della GPU è diventata un collo di bottiglia e quali nuove tecnologie potrebbero migliorare le prestazioni dell'intelligenza artificiale.

buysellram.com/blog/inside-the…

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD

https://www.wacoca.com/life/423980/ Samsung Electronics、7日に4〜6月期速報値　営業利益予想は80兆〜92兆ウォン #DRAM #HBM4 #investment #NAND #SamsungElectronics #toshi #半導体 #成果給引当金 #投資 #業績

A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.

https://www.buysellram.com/blog/inside-the-gpu-memory-hierarchy-how-ai-servers-move-data-from-ssd-to-hbm/

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #buysellram

Why can't a GPU just carry more HBM? Interposers max out in size, stacking 12 to 16 DRAM dies compounds yield losses, and every stack sits beside a kilowatt-class package that hates sharing heat. So capacity climbs in careful steps — 80 GB on the H100, 141 on the H200, 192 on the B200, 288 on Blackwell Ultra — while KV caches for long-context inference balloon past 40 GB per request.

That gap between what models demand and what packaging permits is reshaping server design. NVIDIA's Rubin platform treats CPU memory and HBM as one coherent pool. SanDisk and SK hynix are standardizing High Bandwidth Flash as a capacity tier under HBM, with first samples due this half. CXL 4.0 pooling hardware is landing in racks now.

This article walks the whole memory hierarchy, from on-chip SRAM to NVMe, and explains what each emerging technology actually solves — and what it doesn't.

https://www.buysellram.com/blog/inside-the-gpu-memory-hierarchy-how-ai-servers-move-data-from-ssd-to-hbm/

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #technology

Why can't a GPU just carry more HBM? Interposers max out in size, stacking 12 to 16 DRAM dies compounds yield losses, and every stack sits beside a kilowatt-class package that hates sharing heat. So capacity climbs in careful steps — 80 GB on the H100, 141 on the H200, 192 on the B200, 288 on Blackwell Ultra — while KV caches for long-context inference balloon past 40 GB per request.

That gap between what models demand and what packaging permits is reshaping server design. NVIDIA's Rubin platform treats CPU memory and HBM as one coherent pool. SanDisk and SK hynix are standardizing High Bandwidth Flash as a capacity tier under HBM, with first samples due this half. CXL 4.0 pooling hardware is landing in racks now.

This article walks the whole memory hierarchy, from on-chip SRAM to NVMe, and explains what each emerging technology actually solves — and what it doesn't.

https://www.buysellram.com/blog/inside-the-gpu-memory-hierarchy-how-ai-servers-move-data-from-ssd-to-hbm/

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #ITAD #technology

A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.

https://www.buysellram.com/blog/inside-the-gpu-memory-hierarchy-how-ai-servers-move-data-from-ssd-to-hbm/

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #tech

A modern AI server runs five layers of memory, from nanosecond on-chip SRAM to petabyte-scale SSDs, and keeping the GPU fed is the whole engineering game. This piece walks the full hierarchy: why HBM became the bottleneck, why adding more isn't simple, and where HBF, CXL, and PIM fit into the next generation.

https://www.buysellram.com/blog/inside-the-gpu-memory-hierarchy-how-ai-servers-move-data-from-ssd-to-hbm/

#HBM #HBM4 #GPUMemory #AIInfrastructure #DataCenter #CXL #HighBandwidthFlash #AIHardware #MemoryHierarchy #AIInference #NVMe #tech

⚙️ JEDEC approva SPHBM4: HBM4 più accessibile per accelerare l’AI, ridurre i costi e aprire nuove prospettive per chip e data center. #AI #HBM4

🔗 https://www.tomshw.it/hardware/hbm4-piu-economica-per-lai-jedec-approva-sphbm4

Optical Links Rethink AI's Memory Bind

How will optical links change AI memory in May 2026? Learn why moving from physical stacking to light signals helps AI processors run faster.

#aimemory, #opticalinterconnect, #hbm4, #techinnovation, #aihardware

https://newsletter.tf/optical-links-ai-memory-may-2026/

Engineers are moving away from stacking memory chips to using light. This new method could increase AI memory capacity by several times compared to current limits.

#aimemory, #opticalinterconnect, #hbm4, #techinnovation, #aihardware
https://newsletter.tf/optical-links-ai-memory-may-2026/

SK 海力士市值破 1 兆美元 三星,台積電後成亞洲第三 記憶體需求成動力
南韓晶片大廠 SK Hynix 於 5 月 27 日股價單日急升 9.31% 收報 2,243,000 韓圜， […]
#人工智能 #城中熱話 #新科創業 #HBM4
https://unwire.hk/2026/05/29/sk-hynix-1-trillion-market-cap-hbm4/unwire-space/?utm_source=rss&utm_medium=rss&utm_campaign=sk-hynix-1-trillion-market-cap-hbm4

https://www.wacoca.com/life/395767/ 世界の半導体工場、活発な投資へと動く | サイエンス リポート | TELESCOPE magazine #1γnmプロセス #1δnm製品 #3次元IC #AIチップ #AIデータセンター #Apple #ASETechnology #CHIPＳ法案 #CoWoS #DRAM #FinFET #GAA #GPUHBMメモリ #HBM4 #investment #magazine #MicronTechnology #n2 #N2P #NhancedSemiconductors #NVIDIA #PowertechTechnology #Rubin #samsung #SiliconBox #SKHynix #SOCAMM2 #STMicroelectronics #Telescope #toshi #TSMC #ギガファブ #チップレット #テレスコープ #ファウンドリ20 #マガジン #メモリ半導体 #世界の半導体工場 #先端パッケージ #半導体 #半導体工場 #投資

The Samsung Electronics union has announced a general strike from May 21 to June 7. The dispute originated from differences between labor and ...

Samsung Electronics files injunction against union strike citing potential 30 trillion won losses, risks to semiconductor facilities, and threats to Korea's economy as company accounts for 41% of KOSPI.#SamsungElectronics #unionstrike #semiconductorindustry #injunction #KOSPI #HBM4 #supplychaindisruption #Koreaeconomy
Samsung Seeks Injunction Against Union Strike, Warns of 30 Trillion Won-Plus Losses

Samsung se convierte en el «cerebro» de OpenAI: Suministrará en exclusiva los chips y la memoria para ChatGPT

En un movimiento que redefine el equilibrio de poder en la industria tecnológica, Samsung ha cerrado un acuerdo histórico con OpenAI para convertirse en el proveedor exclusivo de la infraestructura de semiconductores que dará vida a las futuras versiones de ChatGPT. Se ha confirmado que la división de fundición de Samsung y sus memorias de última generación (HBM4) serán el motor detrás de la inteligencia artificial más utilizada del mundo (Fuente Hankyung).

Esta alianza llega en un momento crítico para Samsung, compensando el sabor agridulce de la cancelación del Galaxy Z TriFold y los problemas de fatiga ocular en el S26 Ultra con un contrato multimillonario que asegura su dominio en el sector de los centros de datos.

¿Qué implica este acuerdo para el futuro de la IA?

Hasta ahora, OpenAI dependía en gran medida de chips de terceros (principalmente NVIDIA) y diversas cadenas de suministro. Con este pacto:

• Chips a medida (ASIC): Samsung fabricará procesadores diseñados específicamente por y para los modelos de lenguaje de OpenAI, optimizando el consumo energético y la velocidad de respuesta.
• Memoria HBM4 de 16 capas: Samsung suministrará su memoria de banda ancha más avanzada, fundamental para que la IA procese trillones de parámetros en milisegundos.
• Independencia de la cadena de suministro: Sam Altman, CEO de OpenAI, busca con esto reducir la dependencia de un solo proveedor y asegurar el stock necesario para el despliegue global de sus próximos modelos.

El impacto en el ecosistema Galaxy

Aunque el acuerdo se centra en la infraestructura de servidores, los usuarios finales de Samsung verán beneficios directos:

1. Integración Profunda: Se espera que los futuros dispositivos Galaxy (como el rumoreado S27) incluyan funciones de IA «on-device» optimizadas por este hardware conjunto, superando en velocidad a lo que Apple ofrece actualmente en iOS.
2. Exclusividad de Funciones: OpenAI podría lanzar funciones de ChatGPT de forma prioritaria para los usuarios de Samsung, similar a lo que hemos visto esta semana con la llegada de WhatsApp nativo a los relojes Garmin.

Un marzo de alianzas y tensiones

El anuncio de Samsung y OpenAI es la pieza que faltaba en el rompecabezas de marzo de 2026:

• Privacidad: Mientras el FBI admite la compra de datos de ubicación, OpenAI y Samsung prometen que su infraestructura conjunta permitirá un procesamiento de datos más seguro y localizado.
• Competencia: Este movimiento responde directamente a la expansión de Claude de Anthropic, que recientemente duplicó sus límites de uso, y al lanzamiento del navegador Comet de Perplexity.
• Economía: Tras el «Impuesto TikTok» de 10.000 millones, los grandes contratos de infraestructura como este demuestran que el verdadero valor de la tecnología hoy reside en quién controla el hardware que procesa la inteligencia.

¿Cuándo veremos los resultados?

La producción a gran escala de los nuevos chips comenzará en la segunda mitad de 2026. Para los usuarios de ChatGPT, esto se traducirá en una menor latencia y capacidades multimodales más fluidas (voz y vídeo en tiempo real sin esperas) antes de que termine el año.

https://winbuzzer.com/2026/03/17/samsung-hbm4e-nvidia-gtc-2026-sk-hynix-supply-race-xcxwbn/

Samsung Unveils HBM4E AI Memory Chips at GTC 2026 in SK Hynix Supply Race

#AI #NVIDIA #Samsung #SKHynix #Semiconductors #Memory Groq #BigTech #Chipmakers #Inference #VeraRubin #HBM4 #HBM4E #GTC2026

https://winbuzzer.com/2026/03/17/samsung-hbm4e-nvidia-gtc-2026-sk-hynix-supply-race-xcxwbn/

Samsung Unveils HBM4E AI Memory Chips at GTC 2026 in SK Hynix Supply Race

#AI #NVIDIA #Samsung #SKHynix #Semiconductors #Memory Groq #BigTech #Chipmakers #Inference #VeraRubin #HBM4 #HBM4E #GTC2026

삼성·엔비디아 HBM4 동맹, AI 반도체 새 역사

엔비디아 GTC에서 삼성전자가 차세대 HBM4를 공개하며 AI 반도체 동맹을 한층 강화했습니다. 더 빠른 속도와 높은 전력 효율을 갖춘 HBM4는 글로벌 AI 인프라의 핵심 부품으로 주목받고 있습니다.

#삼성전자 #엔비디아 #HBM4 #AI반도체 #GTC2025 #블로그 #ODOB

Inference is becoming the primary cost center of AI, and NVIDIA’s Feynman roadmap suggests a shift from training-centric GPUs toward latency-optimized, inference-scale systems.

As real-time agents, copilots, and edge deployments grow, inference sovereignty—where compute is located, how fast it responds, and who controls the hardware—will define the next phase of AI infrastructure.

With NVIDIA GTC 2026 approaching, the key question is whether NVIDIA will formally introduce a new class of inference-focused silicon and fabric to complement its training platforms.

https://www.buysellram.com/blog/nvidia-next-gen-feynman-beyond-training-toward-inference-sovereignty/

#InferenceSovereignty #LLMInference #AgenticAI #NVIDIA #Feynman #HBM4 #SRAM #AdvancedPackaging #SiliconPhotonics #AIInfrastructure #GPU #GTC2026 #Rubin #Blackwell #DeterministicCompute #LPX #GroqLPU #technology

With NVIDIA GTC 2026 approaching, the key question is whether NVIDIA will formally introduce a new class of inference-focused silicon and fabric to complement its training platforms.

https://www.buysellram.com/blog/nvidia-next-gen-feynman-beyond-training-toward-inference-sovereignty/

#InferenceSovereignty #LLMInference #AgenticAI #NVIDIA #Feynman #HBM4 #SRAM #AdvancedPackaging #SiliconPhotonics #AIInfrastructure #GPU #GTC2026 #Rubin #Blackwell #DeterministicCompute #LPX #GroqLPU #technology

#Samsung will die #KI-Branche zur Kasse bitten: Für #HBM4-#Speicher sollen #Nvidia und Co. rund 700 Dollar pro Modul zahlen - 30% mehr als ursprünglich erwartet. https://winfuture.de/news,156984.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

#Samsung will die #KI-Branche zur Kasse bitten: Für #HBM4-#Speicher sollen #Nvidia und Co. rund 700 Dollar pro Modul zahlen - 30% mehr als ursprünglich erwartet. https://winfuture.de/news,156984.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

#Nvidia-Chef Jensen Huang verspricht zur GTC 2026 einen Chip, der "die Welt überraschen" wird. Nach einem Treffen mit SK Hynix deutet vieles auf die neue Rubin-Architektur mit #HBM4-Speicher hin. https://winfuture.de/news,156973.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

#Nvidia-Chef Jensen Huang verspricht zur GTC 2026 einen Chip, der "die Welt überraschen" wird. Nach einem Treffen mit SK Hynix deutet vieles auf die neue Rubin-Architektur mit #HBM4-Speicher hin. https://winfuture.de/news,156973.html?utm_source=Mastodon&utm_medium=ManualStatus&utm_campaign=SocialMedia

HBM4 für Nvidia Vera Rubin: Samsung startet Massenproduktion und Auslieferung (Update) https://www.computerbase.de/news/wirtschaft/hbm4-fuer-nvidia-vera-rubin-samsung-startet-massenproduktion-und-auslieferung.96121/#update-2026-02-13T08:38 #Samsung #HBM4

HBM4 für Nvidia Vera Rubin: Samsung startet Massenproduktion und Auslieferung (Update) https://www.computerbase.de/news/wirtschaft/hbm4-fuer-nvidia-vera-rubin-samsung-startet-massenproduktion-und-auslieferung.96121/#update-2026-02-13T08:38 #Samsung #HBM4

Nvidia, Rubin AI GPU'lar için HBM4 bellek hız beklentilerini düşürdü. Amaç, tedarik hacmini güvence altına almak. Yapay zeka donanım tedarik zincirinde stratejik bir hamle!

🚩 #Nvidia #YapayZeka #HBM4 #GPU #Teknoloji #AI

HBM4 für Nvidia Vera Rubin: Samsung startet Massenproduktion und Auslieferung https://www.computerbase.de/news/wirtschaft/hbm4-fuer-nvidia-vera-rubin-samsung-startet-massenproduktion-und-auslieferung.96121/ #Samsung #HBM4

HBM4 für Nvidia Vera Rubin: Samsung startet Massenproduktion und Auslieferung https://www.computerbase.de/news/wirtschaft/hbm4-fuer-nvidia-vera-rubin-samsung-startet-massenproduktion-und-auslieferung.96121/ #Samsung #HBM4

Samsung announces the start of mass production of its next-generation HBM4 memory chips for AI. https://english.mathrubhumi.com/technology/samsung-mass-produces-next-gen-ai-hbm4-memory-chip-da1pa7um?utm_source=dlvr.it&utm_medium=mastodon #Samsung #AI #Chips #HBM4 #TechNews

Samsung begins mass production of HBM4 chips this month, supplying Nvidia for AI GPUs. Get ready for the next generation of AI technology. https://english.mathrubhumi.com/technology/samsung-hbm4-mass-production-nvidia-ai-n4645uzd?utm_source=dlvr.it&utm_medium=mastodon #Samsung #HBM4 #Semiconductors #AIChips #TechNews

Demis Hassabis warnt vor einem massiven Ungleichgewicht: Software skaliert, Hardware limitiert. HBM4-Speicher ist bis Ende 2026 komplett ausverkauft, da Hyperscaler die globale Produktion binden. Für Neueinsteiger schließt sich das Fenster fast vollständig. Energiekosten und Nvidias Preispolitik verschärfen die Lage drastisch. Innovation scheitert gerade an der Supply Chain. #Nvidia #OpenAI #HBM4
https://www.all-ai.de/news/beitrage2026/ki-paradoxon-hassabis

Demis Hassabis warnt vor einem massiven Ungleichgewicht: Software skaliert, Hardware limitiert. HBM4-Speicher ist bis Ende 2026 komplett ausverkauft, da Hyperscaler die globale Produktion binden. Für Neueinsteiger schließt sich das Fenster fast vollständig. Energiekosten und Nvidias Preispolitik verschärfen die Lage drastisch. Innovation scheitert gerade an der Supply Chain. #Nvidia #OpenAI #HBM4
https://www.all-ai.de/news/beitrage2026/ki-paradoxon-hassabis

Samsung is Back. The AI giants have spoken, and the HBM4 is a beast!

Samsung Electronics has officially kicked off 2026 with a massive "mic drop" in the semiconductor world. In a New Year address, co-CEO Jun Young-hyun revealed that the company's next-gen HBM4 (6th-gen High Bandwidth Memory) has received glowing reviews from major AI customers.

#samsungs #hbm4 #samsungisback #aitech #semiconductors #nvidia

Samsung is Back. The AI giants have spoken, and the HBM4 is a beast!

Samsung Electronics has officially kicked off 2026 with a massive "mic drop" in the semiconductor world. In a New Year address, co-CEO Jun Young-hyun revealed that the company's next-gen HBM4 (6th-gen High Bandwidth Memory) has received glowing reviews from major AI customers.

#samsungs #hbm4 #samsungisback #aitech #semiconductors #nvidia

Ah yes, the elusive #HBM4 standard is being developed to revolutionize our world by... reducing pin counts. 🙄 Too bad the details are as accessible as an invite to the Illuminati's annual gala—403 Forbidden, indeed. 🔒 Kudos to #JEDEC for teasing us with the tech version of a "no entry" sign. 🚫
https://blocksandfiles.com/2025/12/17/jedec-sphbm4/ #technology #pin-counts #innovation #403Forbidden #HackerNews #ngated

Ah yes, the elusive #HBM4 standard is being developed to revolutionize our world by... reducing pin counts. 🙄 Too bad the details are as accessible as an invite to the Illuminati's annual gala—403 Forbidden, indeed. 🔒 Kudos to #JEDEC for teasing us with the tech version of a "no entry" sign. 🚫
https://blocksandfiles.com/2025/12/17/jedec-sphbm4/ #technology #pin-counts #innovation #403Forbidden #HackerNews #ngated

JEDEC developing reduced pin count HBM4 standard to enable higher capacity

https://blocksandfiles.com/2025/12/17/jedec-sphbm4/

#HackerNews #JEDEC #HBM4 #ReducedPinCount #HigherCapacity #MemoryTechnology

https://winbuzzer.com/2025/11/21/amd-unveils-instinct-mi430x-with-432gb-hbm4-memory-targeting-western-supercomputers-xcxwbn

AMD Unveils Instinct MI430X with 432GB HBM4 Memory, Targeting Western Supercomputers

#AMD #AI #Supercomputing #HBM4 #TechNews #SovereignAI #HPC #DataCenter #ChipWar #Semiconductors #GenerativeAI #MI430X #BigTech

Intel EMIB封裝技術會成為AI晶片的未來嗎？ - 電子工程專輯

Link

NextGen #MI450 #AI Lineup, #AMD Says There Will Be ‘No Excuses, No Hesitation’ in Choosing Team Red Over NVIDIA In AI Workloads
In terms of what to expect with #MI400 lineup, you are looking at integration of #HBM4, up to 432 GB, this will bring massive bandwidth improvements. AMD plans to aggressively expand #rackscale options with MI400 series as well, introducing highly anticipated #Helios rack as well, which is claimed to rival Vera Rubin's top configuration on-paper.
https://wccftech.com/with-the-mi450-amd-says-there-will-be-no-excuses-no-hesitation-in-choosing-team-red-over-nvidia/

Micron Begins Shipping HBM4 Memory To Key Customers #hardware #hbm4 #memory #micron

https://www.lowyat.net/2025/355731/micron-hbm4-shipping/

HBM4 wird bemustert: Micron folgt SK Hynix und kommt Samsung zuvor https://www.computerbase.de/news/arbeitsspeicher/hbm4-wird-bemustert-micron-folgt-sk-hynix-und-kommt-samsung-zuvor.93087/ #Micron #HBM4

Teurer Speicher: HBM4 soll 30 Prozent mehr als HBM3E kosten https://www.computerbase.de/news/arbeitsspeicher/teurer-speicher-hbm4-soll-30-prozent-mehr-als-hbm3e-kosten.92875/ #HBM4

Standard veröffentlicht: HBM4 ist fertig und noch etwas schneller geworden https://www.computerbase.de/news/arbeitsspeicher/standard-veroeffentlicht-hbm4-ist-fertig-und-noch-etwas-schneller-geworden.92293/ #HBM4 #JEDEC

💡 SK hynix lancia la memoria HBM4 a 12 livelli: velocità e capacità senza precedenti per l’AI

https://gomoot.com/sk-hynix-lancia-la-memoria-hbm4-a-12-livelli-velocita-e-capacita-senza-precedenti-per-lai/

#ai #blog #hbm #hbm3 #hbm3e #hbm4 #hbm4e #micron #news #picks #samsung #skhynix #tech #tecnologia #tsmc

Anfrage bei SKHynix: Nvidia will früher auf HBM4-Chips mit 1,6 TB/s setzen können https://www.computerbase.de/news/arbeitsspeicher/anfrage-bei-skhynix-nvidia-will-frueher-auf-hbm4-chips-mit-1-6-tb-s-setzen-koennen.90176/ #Nvidia #HBM4 #SKHynix

Samsung To Launch 3D Stacked HBM Memory By 2025 #3dstackedpackaging #airobotics #hardware #hbm #hbm4 #memory #saintd #samsung

https://www.lowyat.net/2024/324831/samsung-to-launch-3d-stacked-hbm-memory-by-2025/