Teraswitch

Our Products

Enterprise-grade infrastructure solutions built for performance, reliability, and scale

Bare Metal Servers

Dedicated servers with 100% of resources allocated to you. No virtualization overhead, no noisy neighbors—just raw computing power for your most demanding workloads.

Latest AMD EPYC & Intel Xeon Processors

Turin, Genoa, Milan, and Rome generation CPUs with up to 128 cores and 3.7GHz+ boost clocks

NVMe SSD Storage

Enterprise-grade PCIe 4.0 NVMe drives with up to 7GB/s throughput and 1M IOPS

Deploy in Minutes

Automated provisioning with your choice of OS and configuration - ready in under 5 minutes

Full Root Access

Complete control over your hardware with root/admin access and custom kernel support

16 Global Locations

Deploy servers in datacenters across North America, Europe, and Asia Pacific

10Gbps+ Networking

High-bandwidth connectivity with upgrade options to 25Gbps, 40Gbps, and 100Gbps

Popular Configurations

Compute Optimized

AMD EPYC 9454 • 96 cores • 256GB RAM • 2TB NVMe • 10Gbps Network

Starting at $899/mo

Memory Optimized

AMD EPYC 9374F • 32 cores • 768GB RAM • 4TB NVMe • 10Gbps Network

Starting at $1,299/mo

Storage Optimized

AMD EPYC 9334 • 32 cores • 256GB RAM • 16TB NVMe • 10Gbps Network

Starting at $1,499/mo

Balanced Performance

AMD EPYC 9254 • 48 cores • 384GB RAM • 8TB NVMe • 25Gbps Network

Starting at $1,099/mo

High-End Workstation

AMD EPYC 9754 • 128 cores • 512GB RAM • 4TB NVMe • 100Gbps Network

Starting at $2,499/mo

GPU Options

NVIDIA H100 PCIe

80GB HBM3 • Hopper Architecture • 456 Tensor Cores • 2TB/s Memory Bandwidth

Perfect for LLMs & Training

NVIDIA A100 PCIe

80GB HBM2e • Ampere Architecture • 432 Tensor Cores • 2TB/s Memory Bandwidth

Ideal for ML & Analytics

NVIDIA RTX A6000

48GB GDDR6 • Ampere Architecture • 336 Tensor Cores • Ray Tracing

Professional Visualization & AI

NVIDIA RTX A5000

24GB GDDR6 • Ampere Architecture • 256 Tensor Cores • Ray Tracing

Great for Rendering & CAD

NVIDIA L40

48GB GDDR6 • Ada Lovelace • 568 Tensor Cores • AV1 Encode/Decode

Video AI & Graphics Workloads

GPU Servers

Dedicated GPU servers for AI/ML training, inference, rendering, and compute-intensive workloads. Get the power of NVIDIA's latest GPUs without sharing resources.

Latest NVIDIA GPUs

H100, A100, A6000, RTX A5000, and more - optimized for AI/ML, rendering, and HPC

Multi-GPU Configurations

Up to 8 GPUs per server with NVLink for maximum throughput and GPU-to-GPU communication

Flexible Scaling

Scale GPU resources up or down based on your workload needs without long-term commitments

Pre-Configured ML Frameworks

Optional pre-installed CUDA, TensorFlow, PyTorch, and popular AI/ML frameworks

High-Speed Storage

NVMe SSDs for fast dataset loading and model checkpointing

100Gbps Networking

Ultra-fast network connectivity for distributed training and large dataset transfers

Block Storage

High-performance SSD block storage that can be attached to any server. Scale storage independently from compute with enterprise-grade reliability and performance.

NVMe Performance

Up to 100,000 IOPS with sub-millisecond latency for demanding database and application workloads

Triple Replication

Your data is replicated across 3 separate storage nodes for 99.999% durability

Instant Snapshots

Create point-in-time snapshots for backup and disaster recovery with zero downtime

Hot Attach/Detach

Add or remove volumes from servers without rebooting or service interruption

Volume Cloning

Instantly clone volumes for testing, development, or scaling workloads

Encryption at Rest

AES-256 encryption for all data stored on block storage volumes

Storage Tiers

Performance SSD

RECOMMENDED

100,000 IOPS • <1ms latency • NVMe • 7GB/s throughput

$0.15/GB/mo

Standard SSD

25,000 IOPS • 2-3ms latency • SATA • 550MB/s throughput

$0.10/GB/mo

Balanced SSD

50,000 IOPS • 1-2ms latency • NVMe • 3GB/s throughput

$0.12/GB/mo

Archival HDD

500 IOPS • 10-15ms latency • High capacity • 150MB/s throughput

$0.05/GB/mo

Use Cases

Media & Content Delivery

Store and deliver images, videos, and static assets globally with integrated CDN for optimal performance

Backup & Archive

Long-term data retention with lifecycle policies, versioning, and compliance features like WORM storage

Data Lakes & Analytics

Build scalable data lakes for analytics and machine learning pipelines with direct integration to compute clusters

Application Storage

Store user-generated content, logs, and application data with automatic scaling and high availability

Object Storage

S3-compatible object storage for storing and serving any amount of unstructured data. Perfect for backups, media, and data lakes with 99.999999999% durability.

S3-Compatible API

Drop-in replacement for AWS S3 with full API compatibility - works with existing tools and SDKs

11 Nines Durability

Erasure coding and multi-region replication for 99.999999999% durability guarantee

Lifecycle Policies

Automate data tiering and expiration to optimize costs with intelligent storage class transitions

CDN Integration

Built-in CDN for low-latency content delivery to users worldwide

Versioning & Immutability

Object versioning and WORM (Write Once Read Many) for compliance and data protection

Multi-Region Replication

Automatically replicate objects across multiple geographic regions for disaster recovery

Cloud Direct Connect

Private, high-speed connections to AWS, Google Cloud, Azure, and other cloud providers. Reduce latency and data transfer costs with dedicated links.

Multi-Cloud Connectivity

Connect to AWS, GCP, Azure, and more from a single port with flexible peering options

Up to 100Gbps

High-bandwidth private connections with consistent performance and low latency

Private & Secure

Traffic never touches the public internet for enhanced security and compliance

Instant Provisioning

Deploy private cloud connections in minutes with automated provisioning

Cost Savings

Reduce cloud egress fees by up to 90% compared to public internet transfers

Consistent Performance

Guaranteed bandwidth with SLA-backed latency and uptime commitments

Supported Providers

AWS Direct Connect

1Gbps, 10Gbps, 100Gbps options available with hosted VIF support

Google Cloud Interconnect

Dedicated and Partner Interconnect supported with VLAN attachments

Azure ExpressRoute

Private connections to Azure services with FastPath and Global Reach

Oracle Cloud FastConnect

Direct connectivity to Oracle Cloud Infrastructure (OCI)

IBM Cloud Direct Link

Private connections to IBM Cloud with 1-10Gbps options

Custom Peering

Direct peering with your infrastructure or private cloud environments

Ready to Get Started?

Talk to our team about the right infrastructure solution for your needs