#MI455X
MI455x is a fucking beast!
January 23, 2026 at 7:50 AM
NVIDIA upgrades Vera Rubin HBM4 bandwidth by 10% in order to stay ahead of AMD Instinct MI455X
NVIDIA upgrades Vera Rubin HBM4 bandwidth by 10% in order to stay ahead of AMD Instinct MI455X
NVIDIA raised Vera Rubin's HBM4 bandwidth by 10% to outpace AMD's MI455X, showing ongoing pressure in AI hardware development. ⚙️🤖
www.tweaktown.com
January 22, 2026 at 5:25 AM
Nvidia boosts Vera Rubin performance to challenge AMD Instinct AI accelerators: Nvidia is reportedly enhancing its Vera Rubin AI platform with increased power and memory bandwidth to outperform AMD's Instinct MI455X accelerators. The Rubin GPU's Thermal Design Power (… https://ranked.news/140892?u=b
January 22, 2026 at 12:04 AM
NVIDIAが10ヶ月で3回スペックを書き換えた理由

NVIDIAがVera Rubinの帯域幅を10ヶ月で3回書き換えました
13TB/s → 20.5TB/s → 22TB/s、70%の上方修正
計画的ロードマップではなく「追われる者の軌跡」に見えます

AMDのInstinct MI455Xが19.6TB/sで迫り、OpenAIとの6GW契約が締結された今、王座は揺れています
競争がもたらした恩恵と、市場構造の変化を解説しました

note.com/joho_no_toda...

#NVIDIA #AMD #AI半導体
NVIDIA rewrites specs 3 times
NVIDIAがスペックを"後出し"で引き上げた理由|情報の灯台
AIチップの王者が、静かに焦っている。CES 2026でNVIDIAが発表したVera Rubinのスペック改訂は、単なるアップグレードではない。AMDという名の影が、確実に迫っている証拠だ。 NVIDIAは今週、次世代AIプラットフォーム「Vera Rubin NVL72」のメモリ帯域幅を22TB/sに引き上げた。AMDのInstinct MI455Xが持つ19.6TB/sを約10%上回る数値...
note.com
January 21, 2026 at 11:20 PM
NVIDIA Is Feeling the Heat From AMD’s Instinct MI455X AI Chips, Triggering Unusual Vera Rubin Upgrades to Hold Its Competitive Edge

wccftech.com/nvidia-is-fe...
NVIDIA Is Feeling the Heat From AMD’s Instinct MI455X AI Chips, Triggering Unusual Vera Rubin Upgrades to Hold Its Competitive Edge
NVIDIA's Rubin AI chips have seen significant upgrades in key specifications over time, as Team Green aims to maintain its competitive lead.
wccftech.com
January 21, 2026 at 5:15 PM
=>
GPU: HBM + (LP) DDR DRAM
NVIDIA x.com/ogawa_tter/s...
Patent, Oct 2023, (Apr 2021)
ASSYST 2023, with GaTech
H100 HBM + Tensor Offloading-Memory

AMD
"Integrated Memory Controller Circuitry", (P: Sep 2023) patents.google.com/patent/US202...
MI455X, CES 2026 x.com/SemiAnalysis...
HBM4 + LPDDR5
January 20, 2026 at 5:46 AM
Meta is reportedly scaling back its in-house ASIC (MTIA) plans, and instead of buying TPUs, it’s buying AMD’s MI455X.
January 20, 2026 at 1:30 AM
Nvidia z důvodu parametrů konkurenčního Instinct MI455X již jednou zvyšovala specifikace Rubin, což vedlo k odkladu výroby z roku 2025 na začátek roku 2026. Došlo ještě na další úpravy a odklad…
https://diit.cz/clanek/nvidia-znovu-odklada-rubin-masova-vyroba-se-posouva-na-brezen
Nvidia znovu odkládá Rubin, masová výroba se posouvá na březen | Diit.cz
diit.cz
January 12, 2026 at 6:50 AM
AMD ukázalo Helios s 2nm čipy: EPYC na Zen 6 a MI455X

Pro akceleraci úloh umělé inteligence má AMD nový rack Helios. Ten obsahuje první 2nm čipy společnosti, a to jak CPU EPYC, tak i GPU Instinct MI455X....
AMD ukázalo Helios s 2nm čipy: EPYC na Zen 6 a MI455X
Pro akceleraci úloh umělé inteligence má AMD nový rack Helios. Ten obsahuje první 2nm čipy společnosti, a to jak CPU EPYC, tak i GPU Instinct MI455X.
svethardware.cz
January 11, 2026 at 12:00 PM
Co vidíte na snímku, je čip s největším kusem 2nm křemíku, jaký zatím byl k vidění nebo vůbec ohlášen. Na trh půjde ve druhém pololetí jako řada akcelerátorů Instinct MI400…
https://diit.cz/clanek/amd-ukazala-kremik-instinct-mi455x-s-320-miliardami-tranzistoru-432-gb-hbm4
AMD ukázala křemík Instinct MI455X s 320 miliardami tranzistorů a 432 GB HBM4 | Diit.cz
diit.cz
January 8, 2026 at 11:04 PM
AMD Instinct MI500: CDNA 6 erscheint 2027 mit HBM4E, MI455X debütiert mit Helios #AMD #Instinct #AI #CES2026
AMD Instinct MI500: CDNA 6 erscheint 2027 mit HBM4E, MI455X debütiert mit Helios
AMD gab auf der CES erneut einen Ausblick auf die kommende KI-Beschleuniger-Serie MI500. Diese nutzt die CDNA-6-Architektur und HBM4E.
www.computerbase.de
January 8, 2026 at 7:14 AM
🤖 CEO upbeat on AI — Lisa Su highlighted soaring computing demand and new MI455X data-center chip, saying AI adoption is just beginning.

📊 After a big rally last year, some traders are wary amid lofty expectations and concerns about an AI spending bubble.

Source: Yahoo Finance
January 7, 2026 at 7:31 PM
TSMC

AMD zeigt riesige Chipkonstrukte Epyc Venice
und Instinct MI455X (mma)
heise.de/-11133033
AMD zeigt riesige Chipkonstrukte Epyc Venice und Instinct MI455X
Erstmals zeigt AMD-Chefin Lisa Su nackte Versionen der nächsten Serverprozessoren und KI-Beschleuniger. Beide sind riesig.
heise.de
January 7, 2026 at 5:57 PM
But at #CES they’ve also unveiled Helios, their rack-scale platform for AI, consisting of the AMD Instinct MI455X accelerators and AMD EPYC Venice CPUs. Their vision is a blueprint for yotta-scale infrastructure, delivering up to 3 AI exaflops of performance in a single rack.
January 7, 2026 at 5:12 PM
AMD’s EPYC Venice, Instinct MI455X, & Helios Hardware On Display for First Time at CES 2026 - ServeTheHome www.servethehome.com/amds-epyc-ve...
AMD’s EPYC Venice, Instinct MI455X, & Helios Hardware On Display for First Time at CES 2026
At CES 2026, AMD publicly demonstrated for the first time their 2026 data center hardware stack: EPYC Venice, Instinct MI455X, and the Helios rack
www.servethehome.com
January 7, 2026 at 5:02 PM
AMD zeigt riesige Chipkonstrukte Epyc Venice und Instinct MI455X

Erstmals zeigt AMD-Chefin Lisa Su nackte Versionen der nächsten Serverprozessoren und KI-Beschleuniger. Beide sind riesig […]
Original post on social.heise.de
social.heise.de
January 7, 2026 at 2:11 PM
AI 400 시리즈, 그리고 9850X3D!! > 리포트 | 퀘이사존 QUASARZONE

quasarzone.com/bbs/qn_repor...

- AMD가 MI455X 탑재 AI 인프라 헬리오스, 라이젠 AI 400 시리즈, AI Max+ 모델, ROCm 7.2 업데이트, 라이젠 7 9850X3D CPU(9800X3D보다 400MHz 빠름)를 발표

- 모두 2026년 1~2분기 출시 예정
January 7, 2026 at 2:01 PM
AMD just sketched its 2027 roadmap: 2 nm MI500, 432 GB HBM4 MI455X, 7000-lb Helios rack. Who’s ready for the YottaFLOPS era?
#AMD
open.substack.com/pub/chippub/...
AMD’s Two Strongest Chips Unveiled
AMD unveils 2nm MI500 roadmap, 2.9 ExaFLOPS Helios rack, Ryzen AI 400 PCs.
open.substack.com
January 7, 2026 at 6:48 AM
#AMD launches MI440X for on-premise enterprise AI | MI455X data center flagship | MI500 preview: 1,000x faster than MI300X (2027) | OpenAI partnership strengthens | Challenges NVIDIA dominance
January 7, 2026 at 1:30 AM
This tweet appeared under this Techmeme headline:

Ryan Shrout / @ryanshrout:

@AMD The total compute available in each rack is really...astounding. To put this performance in context, @LisaSu says that MI455X will offer up to 10x better perf than the previous generation. [image]
January 6, 2026 at 11:23 PM
This tweet appeared under this Techmeme headline:

Patrick Moorhead / @patrickmoorhead:

"Helios, the world's best AI rack"... oooh this is 🌶️ MI455X is a giant chip. 4 per compute tray. Venice CPU with 256 cores. "Data at full speed". Plus Pensando Volcano and Salina. $AMD [image]
January 6, 2026 at 11:17 PM
This tweet appeared under this Techmeme headline:

@economyapp:

$AMD AMD Data Center at CES: • Introduces "Helios" AI Rack. • MI455X 10x Performance vs. MI355X. • MI500 coming in 2027 with a massive leap. [image]
January 6, 2026 at 11:03 PM