$299/mo
Substrate Computing Device
Device Status Dashboard Online
Status Operational
Uptime 14d 7h 23m
Temperature 67C
Fan Speed 2,400 RPM
GPU Compute NVIDIA A100
GPU Utilization 78%
CUDA Cores Active 5,392 / 6,912
Tensor Core Load 64%
Network Connected
Bandwidth 12.4 Gbps
Latency 0.8ms
Power Draw 287W
System Gauges Real-time
84%
remaining
Battery
1.2T
of 2TB
Storage
72%
58/80 GB
Memory
GPU Execution Metrics Live
Operations/Second 847.2K
FLOPS (FP16) 312 TFLOPS
Memory Bandwidth 1,847 GB/s
Active Warps 1,024
Occupancy 91.4%
Cache Hit Rate 94.7%
Kernel Queue Depth 12
App Decomposition Viewer Neural Engine v3
NE
Neural Engine v3
14 MOSM modules | 2,847 opcodes
42% GPU
Decomposed Modules:
conv_layer_1 conv_layer_2 attention_block ffn_block norm_layer embed_table output_head loss_fn
Opcode Frequency Chart Top 10
MATMUL
4,281
LOAD
3,647
STORE
3,358
RELU
2,519
CONV2D
2,243
BATCHNORM
1,681
ATTN
1,447
SOFTMAX
1,028
REDUCE
841
TRANSPOSE
653
Installed Applications 5 apps | 3 running
NE
Neural Engine v3
Deep learning inference | 42% GPU
Running
IP
Image Pipeline
Real-time image processing | 18% GPU
Running
LI
LLM Inference
Language model serving | 31% GPU
Running
DT
Data Transform
ETL pipeline engine | 0% GPU
Idle
CV
Crypto Validator
Blockchain verification | 0% GPU
Idle