Research-Grade Performance. Nonprofit Pricing.
Academic & Nonprofit Pricing — Q4, 2025
Built to support researchers, educators, and innovators, myresearchcloud.ca combines the transparency of open infrastructure with the reliability of enterprise-grade systems. Every instance is hosted entirely in Canada, powered by renewable-aware datacentres, and managed by a nonprofit organization — ensuring that performance, privacy, and purpose stay aligned.
1. Sovereignty — 100 % Canadian-owned and operated. Your data, models, and backups never leave the country.
2. Flexibility — Choose monthly, 1-year, or 3-year Reserved Instances — all billed monthly, with no prepayment required.
3. Energy Efficiency — Powered by renewable-aware infrastructure and optimized for maximum performance per watt.
4. Predictable Billing — No egress charges, IOPS fees, or hidden “usage” costs — ever.
5. Dedicated Performance — No noisy neighbours. Your instance is isolated and guaranteed — compute, memory, and storage are yours alone.
6. Transparent Pricing — Consistent CAD pricing, with 1-year and 3-year terms priced 40–60 % below hyperscalers.
7. Sustainable Impact — Every paid service funds free research resources for underserved Canadian scientists and students.
Eligibility
Academic & Nonprofit pricing is available to universities, colleges, hospitals, research institutes, government research agencies, and registered nonprofit organizations engaged in research, education, or innovation for public benefit.
Access is intended for workloads that advance science, learning, or social good — not commercial operations.
Projects must be non-commercial in nature, funded by public, institutional, or nonprofit sources.
Data and compute must remain within Canadian jurisdiction and comply with all privacy, ethics, and research-governance requirements.
Allocations may be free, subsidized, or cost-recovery, depending on project scale and resource availability.
All users agree to the myresearchcloud.ca Terms of Use, which prohibit cryptocurrency, scraping, surveillance, and other non-research activities.
myresearchcloud.ca reserves the right to verify eligibility or reclaim unused capacity to maximize research impact.
| Type | vCPU | RAM (GB) | SSD (GB) | 1 Y $/hr | 3 Y $/hr | Description |
|---|---|---|---|---|---|---|
| a.gp.large | 64 | 240 | 1600 | $1.25 | $0.96 | High-throughput VM for small labs running multiple jobs or simulations. |
| a.gp.2xlarge | 128 | 480 | 3200 | $2.50 | $1.92 | Mid-sized lab instance for parallel parameter sweeps and large batch workloads. |
| a.gp.4xlarge | 256 | 960 | 6400 | $4.99 | $3.84 | Institute-scale compute instance — ideal for collaborative or multi-user environments. |
A.GP-Series — General Purpose Compute
Designed for researchers who need scalable, parallel performance for simulations, parameter sweeps, genomics, or Monte Carlo workloads. These instances deliver high throughput and predictable pricing for multi-job or multi-user environments.
A.MO-Series — Memory-Optimized Compute
Engineered for memory-bound workloads and licensed research software that require extremely high RAM per vCPU. Ideal for computational chemistry, finite element analysis, and data-intensive modelling.
| Type | vCPU | RAM (GB) | SSD (GB) | 1 Y $/hr | 3 Y $/hr | Description |
|---|---|---|---|---|---|---|
| a.mo.large | 8 | 480 | 1600 | $1.32 | $1.01 | High-memory configuration with 64 GB per vCPU — ideal for memory-intensive solvers and licensed software. |
| a.mo.2xlarge | 16 | 480 | 1600 | $1.37 | $1.06 | Balanced memory-optimized instance for multi-user or multi-solver research workloads. |
| a.mo.4xlarge | 32 | 960 | 3200 | $2.74 | $2.11 | Large-scale memory instance for licensed modeling, FEA, or computational chemistry workloads. |
Additional Services & Discounts
myresearchcloud.ca offers flexible options and add-on services to support institutional budgeting and data management needs.
| Type | vCPU | RAM (GB) | SSD (GB) | 1 Y $/hr | 3 Y $/hr | Description |
|---|---|---|---|---|---|---|
| a.hp.large | 8 | 60 | 400 | $0.27 | $0.20 | Compact high-IPC instance with 8 vCPUs and 8 GB per vCPU — ideal for smaller simulation or compute-bound workloads. |
| a.hp.2xlarge | 16 | 120 | 800 | $0.53 | $0.41 | Medium high-IPC configuration for parallel modeling, design optimization, and computational workflows. |
| a.hp.4xlarge | 32 | 240 | 1600 | $1.06 | $0.82 | Large high-IPC instance for demanding simulation, CFD, or molecular dynamics workloads requiring AVX-512 acceleration. |
| a.hp.8xlarge | 64 | 480 | 3200 | $2.12 | $1.64 | Double-sized high-IPC configuration for large-scale compute clusters and heavily vectorized workloads. |
A.HP-Series — High Performance Compute
Optimized for CPU-intensive applications that demand high IPC and AVX-512 acceleration — ideal for simulations, engineering, computational physics, and modelling workloads.
A.GI-Series — GPU Compute: Inference & Language Models
Engineered for memory-bound workloads and licensed research software that require extremely high RAM per vCPU. Ideal for computational chemistry, finite element analysis, and data-intensive modelling.
| Type | vCPU | RAM (GB) | SSD (GB) | 1 Y $/hr | 3 Y $/hr | Description |
|---|---|---|---|---|---|---|
| a.gi.4xlarge | 64 | 480 | 3200 | $14.92 | $8.17 | 4 × L40S GPUs (192 GB total VRAM). Runs full-precision models up to ≈30 B parameters (e.g., Llama 2 34B BF16) or quantized LLMs up to ≈70 B (Llama 2 70B INT4). Excellent for digital twins, inference clusters, and research-grade fine-tuning. |
| a.gi.8xlarge | 128 | 960 | 6400 | $29.84 | $16.34 | 8 × L40S GPUs (384 GB total VRAM). Supports full-precision LLMs up to ≈70 B parameters (Llama 2 70B BF16) and quantized or tensor-parallel inference for models exceeding 100 B parameters (Mixtral 12×7B, Llama 3 70B+). Ideal for multi-agent digital twins, medical imaging AI, and institute-scale R&D clusters. |
Volume & Commitment Discounts
- Reserved Instances (RIs) Paid Up-Front — Additional discount available for 1-year and 3-year terms paid in advance.
- Multiple RIs — Volume-based discounts available for research groups or institutions provisioning multiple reserved instances.
Additional S3 Storage (per month)
| Storage Tier | Monthly Cost (CAD / TB) |
|---|---|
| Tape-Backed S3 — Archival class; ideal for long-term cold storage. | $2 / TB |
| Standard Spindle S3 — Balanced cost-to-performance tier. | $5 / TB |
| NVMe-Backed S3 — High-throughput tier for active datasets or AI training workloads. | $10 / TB |
Add-on storage may be provisioned in addition to instance allocations upon request.
All storage tiers are Canadian-hosted and compliant with full data-residency requirements.
Please contact us for storage requirements exceeding one petabyte (PB) to discuss additional discounting.
Notes
The following terms apply to all academic and nonprofit instance types, ensuring transparency, predictability, and full data sovereignty.
- All pricing is listed in CAD per hour (approximately 730 hours per month).
- Pricing includes full compute, storage, and bandwidth usage — no egress, IOPS, or hidden fees.
- CPU instances (A.GP, A.HP, A.MO) include 10 GbE networking by default, with 25 GbE available on request.
- GPU instances (A.GI) feature dedicated 25 GbE networking for sustained throughput and low-latency inference.
- Each instance runs on isolated hardware — no shared cores, no noisy neighbors, and no throttling.
- All services are hosted in Canadian sovereign datacentres operated in partnership with ThinkOn.
- Infrastructure is optimized for energy efficiency, predictable performance, and full data residency within Canada.