IBM rental providers calculate pricing using a resource-based + time-based billing model, meaning you are charged for what you allocate, how long you use it, and the service level attached to it.

Here’s the clear breakdown of how pricing is actually calculated:

🧮 💰 1. Core pricing formula (simple view)

💡 Total IBM cost = (Compute + Memory + Storage + Network + Licensing) × Time

Where time is usually:

• Hourly billing (primary)
• Converted to monthly (~730 hours)

🖥️ 2. Compute pricing (CPU-based)

🧠 How it’s calculated:

• Charged per vCPU-hour or Power core-hour
• Each core has a fixed hourly rate

Example:

• 8 cores × $0.50/hour × 730 hours
  = $2,920/month

📌 Scaling rule:

More cores = linear increase in cost

🧠 3. Memory pricing (RAM)

How it’s calculated:

• Charged per GB-hour

Example:

• 128 GB × $0.02/hour × 730 hours
  = $1,868/month

📌 Important:

RAM often costs more than CPU in databases (SAP, Oracle, Db2)

💾 4. Storage pricing

How it’s calculated:

• Charged per GB-hour of provisioned storage

Example:

• 1 TB storage × hourly rate × 730 hours

Factors affecting price:

• SSD vs standard disk
• Performance tier (IOPS level)
• Snapshots + backups

📌 Rule:

You pay for allocated storage, not used storage

🌐 5. Network (bandwidth) pricing

How it’s calculated:

• Charged per GB of outbound data (egress)

Typical model:

• Inbound = free
• Internal IBM traffic = free
• Outbound internet = paid

Example:

• 10 TB/month outbound × $0.08/GB
  = $800/month

🧾 6. Licensing costs (software layer)

IBM adds or bundles software costs:

• AIX / IBM i / Linux subscriptions
• Db2 / WebSphere / MQ
• SAP / Oracle licenses (BYOL or included)

Pricing logic:

• Per core or per instance
• Or fixed subscription

📌 Example:

• $50 – $500 per core/month (middleware)
• SAP/Oracle can exceed infrastructure cost

📦 7. Pricing modifiers (VERY important)

IBM adjusts pricing based on:

🔹 Commitment level

• Pay-as-you-go → highest price
• 1–3 year contract → 15%–60% discount

🔹 Instance type

• Shared VM → cheaper
• Dedicated host → more expensive
• Power / LinuxONE → premium pricing

🔹 Region / data center

• High-cost regions = higher rates
• Enterprise locations = premium SLA pricing

🔹 SLA / managed services

• 24/7 support, HA, DR = +30% to +200% cost

📊 8. Real example calculation (enterprise DB server)

Configuration:

• 16 cores
• 256 GB RAM
• 2 TB storage
• 5 TB outbound traffic

Cost breakdown:

• CPU: ~$5,840/month
• RAM: ~$3,750/month
• Storage: ~$400/month
• Bandwidth: ~$400/month
• Licensing: ~$1,000–$3,000/month

👉 💰 Total:
~$11,000 – $14,000/month

⚙️ 9. Key pricing principle (important insight)

IBM pricing is not based on “server size” — it is based on resource consumption over time

So scaling happens like this:

• More CPU → linear cost increase
• More RAM → heavy cost increase
• More storage → predictable linear increase
• More traffic → usage-based increase
• More services (DB, SAP, managed) → major cost multiplier

📌 10. Simple summary

IBM rental pricing is calculated using:

• 🧠 CPU (core-hour billing)
• 🧠 Memory (GB-hour billing)
• 💾 Storage (GB-hour billing)
• 🌐 Bandwidth (GB egress)
• 🧾 Software licenses (per core or subscription)
• ⏱️ Time (hourly → monthly aggregation)

💡 Bottom line

IBM providers calculate pricing by combining resource allocation × time × service level, which means cost scales directly with CPU, RAM, storage, and enterprise features—not just server count.