$HBM is live on Pons v2, Robinhood Chain.$HBM is live on Pons v2. Buy $HBM

HBM
WebGPUEnterprise

GPUs / NVIDIA

RTX 4070 Laptop

Memory
GDDR6
Bus
128-bit
Published peak
256 GB/s
Capacity
from 8 GB

Measured

Read
217.1GB/s
Write
228.3GB/s
Copy
216.7GB/s
Random
41.2GB/s
Latency
282ns
Sustain
217.3GB/s

Medians of 2 verified runs from 1 operator.

66.72 verified runs, median 67.567.5

Local AI speed

ModelParameters4-bit, tokens/s8-bit, tokens/s
Llama 3.1 8B8.03B≤ 54Does not fit
Qwen 2.5 14B14.7B≤ 30Does not fit
gpt-oss-20b21B, 3.6B activeDoes not fitDoes not fit
Mistral Small 24B24BDoes not fitDoes not fit
Qwen 2.5 32B32.5BDoes not fitDoes not fit
Llama 3.3 70B70.6BDoes not fitDoes not fit
gpt-oss-120b117B, 5.1B activeDoes not fitDoes not fit
Upper bounds: median read ÷ bytes read per token. Mixture-of-experts models read only their active experts. Fit checks weights against the smallest 8 GB configuration; the KV cache needs room on top.

Top operators

#OperatorScoreReadDate
1OP-0A8C1F67.5210.1 GB/s2026-09-27

Full RTX 4070 Laptop leaderboard

Test your GPU