Skip to main content

NVIDIA GB200 (Grace Blackwell 200, 2024-Q4)

Product Overview

NVIDIA GB200 (Grace Blackwell 200) is NVIDIA's second-generation Grace Blackwell Superchip, entering mass production Q4 2024. It pairs a B200 GPU with an Arm Grace 72-core CPU via NV-HBI 900 GB/s high-speed interconnect — 1 Superchip = 1 GPU + 1 CPU. 72 GB200 units form an NVL72 rack delivering 1 EFLOPS FP4 sparse, NVLink 5 130 TB/s interconnect, and ConnectX-8 800G networking — NVIDIA's 2024–2025 data center AI flagship.

Generational comparison:

  • GH200 (2023-Q3): Grace + H100, FP8 1 PF sparse, NVLink 4 60 TB/s
  • GB200 (2024-Q4): Grace + B200, FP4 10 PF sparse, NVLink 5 130 TB/s, 1 EFLOPS / NVL72
  • GB300 (2025 H2): Vera 88-core + B300 Ultra + ConnectX-9 1.6T, 1.08 EFLOPS / NVL72 (page available)

Core Specifications

ParameterValue
ArchitectureGrace Blackwell 200 Superchip
GPU Chip1× B200 (Blackwell)
CPU Chip1× Arm Grace (72-core Neoverse V2)
NV-HBI900 GB/s bidirectional CPU-GPU interconnect
CPU-GPU Coherent MemoryUnified addressing
GPU Memory192GB HBM3E
GPU Bandwidth8 TB/s
CPU Memory480GB LPDDR5X (on Grace)
CPU Bandwidth512 GB/s
FP4 Sparse20 PFLOPS (B200 single GPU ×2 Superchip)
FP8 Dense4.5 PFLOPS
BF16 Dense2.25 PFLOPS
TDP (per Superchip)1000W
Form FactorBoard-integrated (non-removable)
Mass Production2024-Q4
Unit Price~$60,000–70,000 (Superchip module)

GB200 NVL72 Rack

ParameterConfiguration
Superchip Count72× GB200
GPU Count72× B200
CPU Count72× Arm Grace (72 cores × 72 = 5,184 cores)
Total HBM13.8 TB HBM3E
Total LPDDR5X34.6 TB
NVLink 5 Domain130 TB/s fully connected
ConnectX-8 Egress72× 800G = 57.6 Tb/s
FP4 Sparse Total1,440 PFLOPS (72× 20 PFLOPS)
FP8 Dense Total324 PFLOPS (72× 4.5 PFLOPS)
Rack TDP~120 kW
Rack Count8 (standard data center row)
Price~$3M / rack (estimated)

GB200 NVL576 (8-Rack)

ParameterConfiguration
Superchip Count576× GB200
GPU Count576× B200
NVLink 5 DomainCross-rack 130 TB/s
Total HBM110 TB
FP4 Sparse Total11.52 EFLOPS (576× 20 PFLOPS)
FP8 Dense Total2.6 EFLOPS (576× 4.5 PFLOPS)
Rack TDP960 kW
Price~$24M

GB200 NVL576 advantage: 8 racks, 576 GPUs sharing a single 130 TB/s NVLink domain — the largest single AI compute domain in the industry as of 2024, critical for trillion-parameter LLM training.

GH200 → GB200 → GB300 Comparison

MetricGH200 (2023-Q3)GB200 (2024-Q4)GB300 (2025 H2)
GPUH100B200B300 Ultra
CPUGrace 72-coreGrace 72-coreVera 88-core
GPU Memory96GB HBM3192GB HBM3E288GB HBM3E
GPU Bandwidth3.35 TB/s8 TB/s10 TB/s
NVLink Domain60 TB/s130 TB/s130 TB/s
NetworkingConnectX-7 400GConnectX-8 800GConnectX-9 1.6T
FP4 SparseN/A (FP8 2 PF)10 PF15 PF
FP8 Dense1 PF2.25 PF3.75 PF
TDP (Superchip)1000W1000W1200W

ConnectX-8 800G Networking

DimensionSpecification
Speed800 Gb/s single port (2× ConnectX-7)
Port Count2–4 per Superchip
ProtocolInfiniBand NDR / RoCE v2
Latency< 0.5 μs
GPUDirectGPU-NIC direct DMA
Congestion ControlSHARP v3
2024 DeploymentsORNL Aurora successor, CSCS Alps, EuroHPC

ConnectX-8 upgrade: 2× speed (400G → 800G), GPUDirect RDMA 3.0, supports NVLink over IB (cross-rack NVLink).

Arm Grace 72-Core

DimensionSpecification
ArchitectureArm Neoverse V2
Core Count72 cores
L3 CacheShared 192MB
LPDDR5X480GB
Bandwidth512 GB/s
TDP200W (CPU only)
PCIeGen5 ×32
FeaturesSVE2 enhanced

Grace vs Vera upgrade: Vera is Grace's successor (88-core + 256MB L3 + 480GB LPDDR5X). GB200 still uses Grace 72-core; only GB300 upgrades to Vera.

GB200 Use Cases

  • Trillion-parameter LLM training (NVL576 domain, 130 TB/s NVLink)
  • MoE model training (expert parallel + tensor parallel)
  • Ultra-large-scale RLHF (576 GPU synchronized)
  • Multimodal large models (video + text + image)
  • AI for Science (climate, materials, life sciences)
  • Cloud AI services (CoreWeave, Lambda, OVHcloud)
  • ❌ Small-scale inference (cost-prohibitive)
  • ❌ China market (export controls)

GB200 Customers

  • Meta: Llama 4 / 5 training (>$10B order)
  • Microsoft Azure: OpenAI GPT-5 + Copilot
  • Google Cloud: Gemini 1.5 / 2.0
  • AWS: Anthropic Claude 4 + Bedrock
  • CoreWeave: 30K+ GB200 deployed (H1 2025)
  • xAI Grok 3: Colossus cluster 100K+ GB200
  • Oracle Cloud: OCI deployment
  • Lambda Labs: Lambda 1-Click Cluster

Vendor Information

ParameterValue
CompanyNVIDIA Corporation
Product Pagehttps://www.nvidia.com/en-us/data-center/grace-blackwell/
CEOJensen Huang
FoundryTSMC 4NP (B200) + TSMC N3 (Grace)
2024-Q4 Mass ProductionYes
PriceSuperchip ~$60–70K, NVL72 ~$3M

GB200 vs GB300

MetricGB200 (2024-Q4)GB300 (2025 H2)
GPUB200B300 Ultra
CPUGrace 72-coreVera 88-core
GPU Memory192GB HBM3E288GB HBM3E
GPU Bandwidth8 TB/s10 TB/s
NetworkingConnectX-8 800GConnectX-9 1.6T
FP4 Sparse10 PF15 PF
FP8 Dense2.25 PF3.75 PF
TDP (Superchip)1000W1200W

GB300 upgrade: GPU memory +50% (192→288GB), compute +50% (FP8 2.25→3.75 PF), networking 2× (800G→1.6T), comparable price.

Key Features

  • NVL72 domain: 72 GPUs sharing 130 TB/s NVLink
  • NVL576 domain: 576 GPUs across 8 racks NVLink
  • ConnectX-8 800G: 800G single port
  • Arm Grace 72-core: CPU + GPU unified memory addressing
  • FP4 10 PFLOPS: Inference optimized
  • Drawbacks: 1000W TDP, export controls, CUDA-only software