1. Size for the refresh window, not the daily average
A data centre degausser's job is not steady-state. It is the weekend or the quarter when a refresh programme retires a rack of storage and a thousand drives. Average daily volume will tell you the machine sits idle; peak volume tells you whether the project finishes.
| Machine | Cycle | Drives per hour | Drives per 8-hour shift | Peak project of 1,000 drives |
|---|---|---|---|---|
| T-1.5 | < 5 sec | ~720 | ~5,700 | ~2 hours |
| T-5 (NSA) | < 20 sec | ~180 | ~1,440 | ~7 hours (under 1 shift) |
| T-1 / T-1.2 | ~20 sec | ~180 | ~1,440 | ~7 hours (under 1 shift) |
| T-4 (NSA) | < 1 min | ~60–90 | ~500–700 | ~1.5–2 shifts |
| 1100 wand | Operator-paced | 10–30 | 80–240 | Multi-day, operator bound |
2. The flash gap in a modern data centre
A large share of a current data centre's storage is flash: NVMe, M.2, SAS SSDs in appliances, and NVMe-oF targets. None of it is magnetic. A policy that says "all media degaussed" is simply wrong for a modern estate.
- Magnetic (HDD, tape): degauss with the T-5 or T-1.5, then crush with the PDS-100.
- Flash (NVMe, M.2, SSD, USB, SD): shred with the PDS-88, or the PDS-SSD destroyer at industrial volume.
- Optical: shred with the PDS-88.
- Appliances with encrypted storage: crypto-erase with key destruction where supported, and shred where the device is being disposed of.
Add a step to the decommissioning runbook: enumerate the media by type before the run. Most failures come from treating the flash as though it were in scope for the degausser.
3. Building the run
- Scope the project. List every asset with its media type, capacity and data classification. Separate magnetic from flash.
- Freeze the asset list. No additions mid-run; any new asset starts a new batch.
- Stage the media. Bring drives to the destruction area in closed, labelled containers. Never stage media in a general corridor or a lift lobby.
- Degauss in batches. Load the drawer, close, confirm interlock, run, confirm the field reading.
- Crush. PDS-100 with the rear output slide so drives eject straight into a container, minimising handling.
- Shred flash. PDS-88, with batteries removed from any device.
- Log every asset. One row per asset: tag, serial, media, machine, verification value, operator, custodian.
- Close the CMDB. Update asset status so the same media is not re-queued next quarter.
- Dispose of scrap through an authorised e-waste recycler with documentation.
4. What a data centre audit actually checks
| Question | What a good answer looks like |
|---|---|
| "Show me your media disposal procedure for this site." | A written, version-controlled procedure with a named owner and a review date. |
| "Show me destruction evidence for last quarter." | A register with one row per asset, per-cycle verification values, and matched CMDB closure. |
| "What about failed drives?" | A documented procedure that handles non-spinning and non-enumerating media, with an example register row. |
| "Where do your NVMe drives go?" | A named flash destruction step, not a degausser reference. |
| "Who witnessed this?" | Named operator and custodian signatures, and an internal audit sample report. |
| "How is scrap disposed?" | Authorised e-waste recycler with documentation retained. |
| "Is the degausser still in spec?" | A current field-strength verification certificate for the machine. |
5. Recommended configurations
| Profile | Degausser | Crusher | Flash | Get a quote |
|---|---|---|---|---|
| Colocation / small DC | T-1.5 | PDS-75 | PDS-88 | Request a quotation |
| Mid-size data centre | T-5 | PDS-100 + rear slide | PDS-88 | Request a quotation |
| Large / hyperscale-adjacent | T-5 + T-1.5 (redundant) | PDS-100 + PDS-100 | PDS-88 ×2 | Request a quotation |
| Government / PSU data centre | T-4 or T-5 (NSA) | PDS-100 (NSA/CSS) | PDS-88 | Request a quotation |
| ITAD / bulk processing | T-5 (highest throughput) | PDS-100 with output slide | PDS-SSD destroyer | Volume quote on request |