Positioning
Rack-scale system: 36 Grace CPUs + 72 Blackwell GPUs
标签:
Specifications
厂商
NVIDIA
架构
Blackwell
类别
Rack system
制程
TSMC 4NP
发布时间
2024
显存容量
13440 GB
显存类型
HBM3e
显存带宽
576 TB/s
FP16 算力
360000 TFLOPS
FP8 算力
720000 TFLOPS
FP4 算力
1440 PFLOPS
互联带宽
130 TB/s
功耗 TDP
— W
状态
Current
Sources
数据来自厂商官方文档与公开行业报告(NVIDIA、AMD、华为、IDC 等)。算力数值请结合精度与稀疏/稠密口径阅读;本站统一标注,避免纸面参数误导。采集时间 2026-11。
Frequently Asked Questions
Clear answers about AI accelerators, specs, precision and buying. (10 questions)
What is the memory capacity of GB200 NVL72?+
13440 GB HBM3e with 576 TB/s bandwidth.
What is the FP16 throughput of GB200 NVL72?+
360000 TFLOPS (dense).
What does "dense FP4 PFLOPS" mean on a spec card?+
It is the peak throughput in FP4 precision without sparsity, measured in petaflops (10^15 ops/s). Blackwell cards list dense FP4 (e.g. B200 ~9 PFLOPS dense) and an even higher sparse figure. Use dense for honest comparisons.
What is Blackwell Ultra B300?+
B300 is the refreshed Blackwell Ultra GPU with 288GB HBM3e for long-context inference, higher FP4 throughput than B200 and is the core of GB300 NVL72 systems. It targets reasoning models that need large memory and high bandwidth.
Why is sparse compute often double the dense figure?+
Structured sparsity (e.g. 2:4) skips every other zero value in a pattern, letting hardware double useful throughput. Vendors quote sparse numbers because they look higher, but real-world gains depend on whether the model and framework exploit 2:4 sparsity. Always cite dense figures for fair comparison.
Why is memory bandwidth critical for large model inference?+
Autoregressive generation is memory-bandwidth bound: each token requires reading the entire model weights from HBM. Higher bandwidth (TB/s) directly increases tokens-per-second, often more than raw TFLOPS. This is why H100/H200/B200 emphasize HBM bandwidth.
What is the difference between GPGPU and a gaming GPU?+
A gaming GPU is tuned for raster graphics and framed around real-time rendering. A GPGPU (General-Purpose GPU) is the same silicon family but optimized and marketed for compute: larger memory, ECC support, better FP64/FP8 throughput, server cooling and virtualization. Cards like the NVIDIA A100/H100 are GPGPUs, while GeForce RTX is aimed at games.
What are Tensor Cores?+
Tensor Cores are specialized units inside NVIDIA GPUs that perform mixed-precision matrix multiply-and-accumulate in a single operation. They dramatically accelerate deep-learning math at FP16, BF16, FP8 and FP4 while using less power than CUDA cores doing the same work.
What is CoWoS and why is it a bottleneck?+
CoWoS is TSMCs 2.5D packaging technology that places a GPU die next to HBM stacks on a silicon interposer. Advanced AI chips depend on it, and limited CoWoS capacity has been a key supply constraint for NVIDIA/AMD accelerators.
What is NVLink used for?+
NVLink is NVIDIs high-speed GPU-to-GPU interconnect, allowing direct memory sharing and coherent access between GPUs at hundreds of GB/s. It is used to build multi-GPU servers (HGX/DGX) so models larger than one cards memory can be sharded efficiently.
No matching questions.