Sign in
Menu
Scientific computing / data science
Workload guide · Desktop & creative

Scientific computing / data science: hardware requirements guide

MATLAB, Python/R data science, simulations and numerical work.

On-demand — not 24/7

Multi-threaded and memory-hungry; a GPU only for GPU-accelerated libraries (CUDA, RAPIDS). FP64-heavy work belongs on workstation cards.

Hardware requirements by tier

Three tiers for every workload: the minimum that works, the recommended sweet spot, and the comfortable headroom level. These are the same tiers the WisePC decision engine uses when it plans a build around your goal.

TierCPU coresRAMStorageGPUVRAMNetwork
Minimum8 cores32 GB1 TB (ssd)any—1 GbE
Recommended16 cores64 GB2 TB (ssd)any—1 GbE
Comfortable32 cores128 GB4 TB (ssd)dedicated12 GB VRAM2.5 GbE
Overkill52 cores205 GB8 TB (ssd)dedicated12 GB VRAM2.5 GbE

Beyond Comfortable — 52 cores, 205 GB RAM, 8 TB (SSD): headroom you will never use for this workload alone. Put the difference into storage, backup or silence.

Recommended system shape

Derived by the decision engine from the recommended tier — the same logic that plans full builds.

Form factor
tower
Build style
diy
Drive plan
2× 2TB NVMe

Requirements (16 threads, 64GB RAM, 4TB storage) need a custom tower build.

One box or a separate machine?

Separate machine recommended

Runs on demand, not 24/7 — it conflicts with an always-on, energy-efficient architecture. A separate machine keeps the always-on box lean and quiet.

Whether this workload deserves its own system or should live on the main server.

Storage arrays, priced

Five physical strategies for the recommended capacity — the same planner, drive model and reference prices as the storage planner tool.

StrategyDrivesUsableCostIdle
Cheapest 1× 4TB hdd4 TB · none≈ €2145 W
Balanced 1× 16TB hdd16 TB · none≈ €7146 W
Redundant 3× 8TB hdd8 TB · parity2≈ €111915 W
Expansion-friendly 1× 20TB hdd20 TB · none≈ €9187 W
Performance Recommended2× 4TB nvme4 TB · parity1≈ €15862 W

For a fast working tier, go all-flash — the performance strategy protects speed without the HDD rebuild pain.

Backup is not storage

The extra copy that survives drive failure, deletion and ransomware — the same engine as the backup planner.

Protect 2TB with at least two copies on separate hardware, ideally one offsite. The recommended start is a dedicated NAS.

Recommended: Separate NAS
A dedicated box that only stores backups · ≈ €296
high
No offsite copy

Add a second copy outside the building — cloud or a drive at a friend's house.

medium
No versioning

Use a versioning-aware tool (restic, borg, Backblaze) that keeps history.

medium
Unscheduled backups

Schedule backups automatically and monitor them.

medium
Backup on the same machine

Keep the backup on separate hardware or offsite.

What will it cost?

Market-neutral EUR bands from the reference drive model (≈ = estimated street price). Live component prices resolve on the build page for your region.

Storage array
≈ €214 – €1586
Array energy / year
≈ €81/yr
∼37 W idle

Compute (CPU, board, RAM, GPU) is not included — see the recommended build for live prices.

Growth outlook

What happens when the data keeps growing — the same engine as the growth simulator.

This architecture becomes inadequate in year 3 (~793€ to expand to that point) — 3 bays needed. Plan the next step (bigger array, NAS or DAS, or a stronger platform) before then.

Exceeds max 4TB by 1TB — need larger array/DAS/NAS. Fresh Performance array for 5TB: 3 bays, ~2379€.

Which tier do you need?

  • Pick Minimum (8 cores, 32 GB RAM) only for a single-purpose machine on a tight budget — expect little headroom.
  • Recommended (16 cores, 64 GB RAM) is the sweet spot: enough for the workload plus the usual side-services, without overspending.
  • Pick Comfortable (32 cores, 128 GB RAM) when this workload shares the machine with others or will grow — you pay for headroom, not for anxiety.
  • GPU rule for this workload: any.GPU rule for this workload: any (no discrete VRAM needed).
Keep going

Frequently asked questions

How much RAM does scientific computing / data science need?
32 GB is the sensible minimum, 64 GB covers most real setups, and 128 GB gives comfortable headroom for growth and extra services.
How many CPU cores does scientific computing / data science need?
A 8-core CPU is the minimum, 16 cores is the recommended sweet spot, and 32 cores is comfortable when it shares the machine with other workloads.
Does scientific computing / data science need a dedicated GPU?
Any GPU helps (integrated graphics or a small dedicated card).
What storage and network does scientific computing / data science expect?
Storage: 2 TB of SSD is the recommended baseline (1 TB minimum, 4 TB comfortable). Network: 1 GbE is the recommended baseline.
What runs well alongside scientific computing / data science?
It pairs naturally with: Model training / fine-tuning, Software development.
How much power does scientific computing / data science use?
The storage array idles around 37 W (≈ 81 €/year at 0.25 €/kWh). The full system adds CPU, board and fans on top — see the cost section for the honest bands.

Related workload guides