TRD Network Cloud ⚡
Decentralized GPU compute — VMs, Containers, K8s, Bare Metal & more.
⚡ Active Instances
0
across all compute types
💰 Month Spend
$0.00
↑ Within budget
🌐 Network GPUs
48,291
↑ 2,140 online now
💎 $IO Balance
0
Add credits to start
📈 Network Usage (7d)
🔋 Resource Split
VMs —38%
Containers —25%
K8s —22%
Other —15%
GPU Utilization64%
Memory38%
Network I/O72%
⚡ Quick Deploy
Virtual Machine
From $0.39/hr
Container
From $0.29/hr
Kubernetes
Request Access
Bare Metal
Request Access
Confidential
Request Access
🕐 Recent Activity
All logs →📋
No recent activity
Deploy your first instance to see activity here
—
Compute
Manage all your GPU instances across compute types
Virtual Machines
GPU-accelerated VMs — select a processor to get started
🖥️ Total VMs
0
across all regions
✅ Running
0
⏹️ Stopped
0
💰 Hourly Cost
$0.00
⚡ Available GPU Processors
25 processors available
🖥️ Your VM Instances
🖥️
No VMs deployed yet
Select a GPU processor above and click Deploy to launch your first VM.
Containers
Docker containers on GPU compute — per-second billing
🐳 Containers
0
▶️ Running
0
📦 Images Pulled
0
💰 Hourly Cost
$0.00
⚡ Available GPU Processors
22 processors available
🐳 Your Containers
🐳
No containers deployed yet
Select a GPU processor above and click Deploy to launch your first container.
Kubernetes
Managed K8s clusters with GPU node pools
Kubernetes — Request Access
Managed Kubernetes with GPU node pools, auto-scaling, and built-in monitoring. Currently in private beta — request access to join the waitlist.
GPU-optimized node pools with NVIDIA device plugin
Auto-scaling based on GPU utilization metrics
Persistent volumes with high-speed NVMe storage
Integrated monitoring with Prometheus & Grafana
Helm chart support and GitOps workflows
Multi-region cluster federation
Bare Metal
Dedicated physical GPU servers — zero hypervisor overhead
Bare Metal Servers — Request Access
Get full access to physical GPU hardware. No virtualization, no shared tenancy — maximum performance for the most demanding compute workloads.
Dedicated NVIDIA HGX A100 / H100 systems
NVLink / NVSwitch fabric for multi-GPU communication
Up to 8x H100 SXM5 per node (640GB HBM3)
400GbE InfiniBand networking
IPMI / BMC remote management access
Custom OS images and firmware control
Minimum 72-hour reservation windows
Confidential Compute
Hardware-encrypted GPU enclaves for sensitive workloads
Confidential Compute — Request Access
Run sensitive AI workloads with hardware-level encryption. Your data and models remain encrypted even during computation — invisible to the infrastructure provider.
NVIDIA H100 Confidential Computing mode
Trusted Execution Environments (TEE) via AMD SEV-SNP
Encrypted memory with hardware attestation
Zero-knowledge proof of compute integrity
HIPAA, GDPR, and SOC 2 Type II compliant
Remote attestation APIs for verification
Ideal for medical AI, financial models, sensitive IP
GPU Marketplace
Rent decentralized GPU compute at the best prices
Clusters
Multi-node distributed GPU workloads across all compute types
🖥️ VM Clusters
0
✅ Running
0
🖥️ Total Nodes
0
💰 Hourly Cost
$0.00
🖥️ VM Cluster — Create New
Active VM Clusters
🖥️
No VM clusters
Create a VM cluster for distributed training across multiple GPU nodes.
🐳 Container Clusters
0
▶️ Running
0
📦 Total Replicas
0
💰 Hourly Cost
$0.00
🐳 Container Cluster — Create New
Active Container Clusters
🐳
No container clusters
Create a container cluster for scalable distributed inference or training.
Kubernetes Clusters — Request Access
Managed Kubernetes with GPU node pools, auto-scaling, and built-in monitoring. Currently in private beta — request access to join the waitlist.
GPU-optimized node pools with NVIDIA device plugin pre-installed
Auto-scaling based on GPU utilization metrics via KEDA
Persistent volumes with high-speed NVMe storage classes
Integrated Prometheus + Grafana GPU dashboards
Helm chart support, GitOps via ArgoCD / Flux
Multi-region cluster federation with global load balancing
Spot & preemptible GPU node pools for cost savings
Bare Metal Clusters — Request Access
Dedicated physical GPU servers clustered together. No virtualization overhead — raw hardware performance for the most demanding distributed workloads.
Dedicated NVIDIA HGX H100 / H200 / B200 server nodes
NVLink / NVSwitch fabric for intra-node GPU communication
400GbE InfiniBand NDR interconnect between nodes
Up to 8× H100 SXM5 per node — 640GB HBM3 per server
IPMI / BMC remote management and console access
Custom OS images, firmware control, BIOS-level tuning
GPUDirect RDMA for zero-copy inter-node GPU communication
Minimum 72-hour reservation windows, up to 12-month reserved
Confidential Compute Clusters — Request Access
Hardware-level encrypted GPU clusters where your data, models, and computation remain fully private — even from the infrastructure operator. Built for regulated industries and sensitive AI workloads.
NVIDIA H100 Confidential Computing mode with hardware TEE
AMD SEV-SNP encrypted VM memory across all cluster nodes
Remote attestation API — cryptographic proof of compute integrity
Zero-knowledge proof of model execution without data exposure
Encrypted inter-node communication with mutual TLS + SGX attestation
HIPAA, GDPR, SOC 2 Type II, and FedRAMP compliant infrastructure
Ideal for medical AI, financial models, legal data, and sensitive IP
Audit logs with tamper-evident ledger for compliance reporting
Storage
Datasets, checkpoints, and persistent volumes
💾 Used
0 GB
📦 Volumes
0
💵 Cost
$0.00
💾
No storage volumes
Upload datasets or create persistent volumes for your instances.
Analytics
Resource utilization, cost, and performance metrics
⚡ Avg GPU Util
0%
💰 Total Spend
$0.00
⏱️ Compute Hours
0h
🌐 Network Out
0 TB
📈 Spend
🖥️ GPU Hours
📊
No data yet
Deploy instances to see performance analytics.
Logs
System events and deployment logs
No logs yet
Billing
Plans, credits, and payment history
💰 Balance
Current Balance
$0.00
This Month
$0.00
$IO Tokens
0
Starter
Free
Current Plan
$10 free credits/mo
All GPU types
Up to 5 instances
Community support
Pro
$49
per month
$100 credit bonus
Priority support
50 instances
Reserved capacity
Enterprise
Custom
contact sales
Unlimited instances
Dedicated support
SLA guarantee
Custom billing
🧾
No invoices
Your billing history will appear here.
Team
Manage members and access
| Member | Role | Joined | Status | Actions | |
|---|---|---|---|---|---|
👥 No team members | |||||
Notifications
5 unread
Settings
Account, security & preferences
Profile
Security
Notifications
API Keys
Appearance
Profile
Update your public profile info
Danger Zone
Irreversible account actions
Sign Out
Sign out of TRD Network
Delete Account
Permanently delete account and data