bitboy.ro:nvda<GO>

Datacenter GPU & AI accelerator catalog

16 accelerators from 4 vendors, with specifications taken from vendor product pages and datasheets. Pick any two to compare them side by side.

The catalog

NVIDIA datacenter GPUs are fabbed via Samsung Electronics, GlobalFoundries and Taiwan Semiconductor Mfg (ADR) — see the supply-chain graph for foundry → NVDA edges.

CDNA4 '25AMDMach-1 '25SAMSUNGBlackwell '24NVIDIACDNA3 '23AMDHopper '22NVIDIAAscend '19HUAWEI
CMPSKUVENDORARCHMEMBWTDPFP16DOCS
AMD Radeon RX 7900 XTX
RX-7900-XTX
Consumer GPU reference; not datacenter AI but competes with GeForce.
amdRDNA324 GB GDDR6960 GB/s355 W[PAGE]
AMD Instinct MI300X
MI300X-OAM-192GB
CDNA3 accelerator; 192GB HBM3; primary NVDA H100 competitor.
amdCDNA3192 GB HBM35,300 GB/s750 W[PAGE]
AMD Instinct MI325X
MI325X-OAM-256GB
HBM3e refresh of MI300X; 256GB capacity.
amdCDNA3256 GB HBM3e6,000 GB/s1000 W[PAGE]
AMD Instinct MI355X
MI355X-OAM-288GB
CDNA4 generation; targets Blackwell B200 class workloads.
amdCDNA4288 GB HBM3e8,000 GB/s1400 W[PAGE]
Huawei Ascend 910B
ASCEND-910B
Primary China-market AI training chip; specs from Huawei Ascend product documentation.
huaweiAscend64 GB HBM2e400 W320 TFLOPS[PAGE][ARCH]
Huawei Ascend 910C
ASCEND-910C
Successor Ascend generation; public spec sheet limited outside China — verify before production use.
huaweiAscend[PAGE][ARCH]
NVIDIA GeForce RTX 5090
RTX-5090
Consumer Blackwell flagship; 32GB GDDR7.
nvidiaBlackwell32 GB GDDR7575 W[PAGE][ARCH]
NVIDIA B200 SXM
B200-SXM-192GB
Blackwell datacenter GPU; FP4 native; 5th-gen NVLink.
nvidiaBlackwell192 GB HBM3e8,000 GB/s1000 W[PAGE][ARCH]
NVIDIA GB200 Grace Blackwell Superchip
GB200-SUPERCHIP
2x B200 GPU + Grace CPU; aggregate 384GB HBM3e per superchip module.
nvidiaBlackwell384 GB HBM3e2700 W[PAGE]
NVIDIA H100 PCIe
H100-PCIe-80GB
PCIe FHFL dual-slot; 350W default TGP mode.
nvidiaHopper80 GB HBM2e2,000 GB/s350 W[PAGE][PDF]
NVIDIA H100 SXM5
H100-SXM5-80GB
SXM5 variant; FP8/FP16 from NVIDIA product spec table (sparse peak).
nvidiaHopper80 GB HBM33,350 GB/s700 W1,979 TFLOPS[PAGE][PDF]
NVIDIA H200 SXM
H200-SXM-141GB
Hopper refresh with HBM3e; ~1.4x memory capacity vs H100.
nvidiaHopper141 GB HBM3e4,800 GB/s700 W[PAGE]
NVIDIA L40S
L40S-48GB
Inference/visualization GPU; Ada architecture, not Hopper/Blackwell.
nvidiaAda Lovelace48 GB GDDR6864 GB/s350 W[PAGE]
NVIDIA RTX PRO 6000 Blackwell
RTX-PRO-6000-BLACKWELL
Workstation Blackwell; 96GB GDDR7 ECC.
nvidiaBlackwell96 GB GDDR71,792 GB/s600 W[PAGE]
Samsung HBM3E 12-Hi Stack
SAMSUNG-HBM3E-12H
Memory component used by NVDA/AMD accelerators — supply-chain competition, not a compute GPU.
samsungHBM3E36 GB HBM3E[PAGE]
Samsung Mach-1 AI Accelerator
MACH-1-AI
Inference-focused logic chip roadmap; Samsung also competes as HBM supplier (see supply chain). Not a direct GeForce/datacenter GPU SKU.
samsungMach-1[PAGE][ARCH]

Background

Memory capacity and bandwidth usually decide which model fits on a single accelerator, and TDP decides what a rack can hold — which is why those are the columns shown first rather than peak throughput. Where a vendor doesn't publish a figure it is shown as absent rather than estimated.

Latest on silicon

AUG 30Nvidia Q2 Earnings 2026: Record $96B Revenue & AI Chip Analysis - Intellectia AIIntellectia AIAUG 30AI Chip Stocks Investment 2026: Nvidia vs AMD vs Broadcom Analysis - Intellectia AIIntellectia AIAUG 28AMD Stock Price: Why AMD Is Moving After Nvidia Earnings and What Comes Next - MEXCMEXCAUG 28AMD Stock Price: Why AMD Is Moving After Nvidia Earnings and What Comes Next - MEXCMEXCAUG 28AMD Stock Price: Why AMD Is Moving After Nvidia Earnings and What Comes Next - mexc.commexc.comAUG 28AMD Stock Price: Why AMD Is Moving After Nvidia Earnings and What Comes Next - MEXCMEXCAUG 25Nvidia Earnings August 26 2026: AI Chip Leader Q2 Preview & Stock Analysis - Intellectia AIIntellectia AIJUL 31AMD Stock Analysis 2026: Is $600 Per Share Within Reach? - Intellectia AIIntellectia AIJUL 28AMD’s $14B Data Center Bet on Core Scientific Targets Nvidia’s AI Infrastructure Lead - Tech TimesTech TimesJUL 27AMD-Anthropic Chip Partnership Signals Growing AI Competition - EPS Surprise History - VinanetVinanetJUL 25Nvidia Locks $500B Memory Deal With SK Hynix in AI Arms Race - The Tech BuzzThe Tech BuzzJUL 23AMD attacks the rack with Helios systems that rival Nvidia's - The RegisterThe RegisterJUL 20AMD Launches Helios Rack System, Lands Microsoft as Customer - The Tech BuzzThe Tech BuzzJUL 18Nvidia Maintains AI Dominance Amid Growing Competitive Pressures - Book Value Growth - VinanetVinanetJUL 15AMD Leads Semiconductor Rally With 6.5% Gain As AI Demand Reshapes Chip Landscape - alphabetastock.comalphabetastockJUL 06AI Investment Bubble 2026: Is the Semiconductor Rally Sustainable or Headed for a Correction? - Intellectia AIIntellectia AIJUL 04AI Semiconductor Stocks July 2026: NVDA, AMD, INTC Investment Analysis - Intellectia AIIntellectia AIJUL 04AI Chip Stocks Investment July 2026: NVDA vs AMD Analysis & Forecasts - Intellectia AIIntellectia AIJUN 29AMD refines AI chip roadmap, shares track Nvidia and Intel peers - ad-hoc-news.dead-hoc-news.deJUN 25Top 25+ AI Chip Makers: NVIDIA & Its Competitors - AIMultipleAIMultipleJUN 24AI Infrastructure Investment Boom 2026: $700B Hyperscaler Spending Race - Intellectia AIIntellectia AIJUN 18Lisa Su's AMD Posts $5.8B AI Data Center Quarter as Meta Signs 6GW Deal - StartupHub.aiStartupHub.aiMAY 27Investors Eye AMD, Arm, Marvell as Potential AI Chip Contenders Beyond Nvidia - Quarterly Earnings - po-news-eg.netpo-news-eg.netMAY 17AI Chip Stocks Investment 2026: Nvidia vs AMD Analysis & Forecasts - Intellectia AIIntellectia AIMAY 15Data Center GPU Market Size, Share | Research Report [2035] - Market Research FutureMarket Researc