Memory & RAM Buyback

OEMs covered

Samsung · Micron · SK Hynix · Kingston · Crucial · Server-pulls of any major OEM

Pricing notes

Per-stick pricing for matched kits; bulk pricing for de-commissioned-server pulls. Reuse-First refurb economics applied.

What we see most of in Malaysia

How we process your engagement

Send your asset list. We respond with a written MYR quote in per engagement SLA (5 business days for AI accelerators). Pickup against signed manifest within Pickup scheduled per engagement, nationwide Malaysia. Wipe + functional test included as standard — drives sanitised to NIST SP 800-88 Rev. 1 Purge for spinning HDDs and IEEE 2883-2022 firmware Sanitize for SSDs/NVMe. Per-asset Certificate of Destruction issued, line-item invoicing per asset, settlement against your PO. Multi-site programme engagements consolidate to a single MYR settlement line; single-event engagements settle within 7 business days of manifest reconciliation.

Condition grading rubric — how we set the offer

Memory grades on origin (matched-kit vs decommissioned-server pulls), generation (DDR4 vs DDR5), capacity, and registered/buffered status. Matched kit (sealed): highest recovery — 30–50% of original list. Decommissioned-server pulls (matched): 18–32%. Decommissioned-server pulls (unmatched, mixed-OEM): 10–22%. RDIMM / LRDIMM enterprise-grade ECC: typically clears higher than UDIMM. DDR5 vs DDR4: DDR5 holds value longer; DDR4 has fallen meaningfully since DDR5 became broadly available, but DDR4 retains residual demand for legacy server refresh cycles. Failed sticks: tested at receipt; failed sticks recovered at material-recovery economics.

Typical recovery ranges — current secondary market

Sanitisation specifics — what the certificate cites

Memory does not retain customer data — DRAM is volatile and clears at power-down. The sanitisation requirement is procedural (proof of de-installation) and physical (testing, sorting). De-installation record: cross-referenced to the host-server certificate. Per-stick testing: every stick is tested in a memory-validator rig — capacity, speed, ECC integrity, thermal stability over a 30-minute soak. Failed sticks flagged. SPD (Serial Presence Detect) check: verifies the stick reports correctly to the BIOS and has no SPD-tampering artefacts (counterfeit indicator). Heat-spreader inspection: bent or detached heat-spreaders noted on the certificate. Sort by JEDEC profile: sticks sorted by speed grade and CL latency for matched-kit reassembly where the customer wants matched-kit pricing on the pulls.

Recent engagement scenarios (anonymised)

Scenario 1 — Server fleet retirement, 7,680 DIMMs. A retired data-centre estate of 320 servers (each with 24 DIMM slots populated) generated 7,680 DDR4 RDIMM pulls (mostly 32GB and 64GB sticks). Per-stick testing through the validator rig over 6 weeks. Failed-stick rate: 4% (within normal). Reuse-First reuse rate: 91%. Settlement in MYR aggregated on a single invoice line. Customer's reuse rate KPI was lifted by this engagement.

Scenario 2 — DDR5 surplus, 480 sticks (sealed kit). A customer who over-ordered DDR5 RDIMM 96GB for a deployment retired 480 sticks unopened. Sealed-kit pricing (highest in the memory grade structure). Settlement in MYR at the upper end of the band; customer recovered meaningful percentage of original purchase price.

Scenario 3 — Mixed-OEM unmatched pulls, 1,200 DIMMs. A retired data-centre with mixed-OEM server estate generated 1,200 unmatched DIMMs across DDR3 and DDR4 generations. Sorted at receipt; programme-level pricing applied. Failed-stick rate: 7% (slightly above normal due to legacy mix). Settlement in MYR.

Pickup, chain of custody, and logistics

Memory ships in anti-static foam trays or in original DIMM-pack containers. Per-stick serialisation where the SPD-readable serial is captured; bulk pickup for high-volume programmes. The validator-rig lead time adds 3–10 business days depending on volume; settlement is against validated counts. Heat-spreader status: flagged at receipt. Counterfeit screening: SPD verification + visual inspection of the chip markings; counterfeit memory has been an industry issue particularly in the DDR3/DDR4 secondary channel. Sort to JEDEC profile: high-volume pickups (5,000+ sticks) are sorted by JEDEC speed grade (DDR4-2400 / 2666 / 2933 / 3200; DDR5-4800 / 5600 / 6400) so matched-kit reassembly can be priced at matched-kit rates rather than unmatched-pulls rates. SPD lock state: some OEM-specific SPD-locked sticks (e.g., Dell PowerEdge-locked DIMMs) are validated for cross-OEM compatibility before refurb routing; locked sticks are flagged on the certificate. Storage between receipt and validation: memory inventory is held in climate-controlled, anti-static storage between receipt and validation to prevent humidity-related contact corrosion that would otherwise drop the validation pass rate.

Common pitfalls that erode recovery

Pitfall 1 — Bent pins on RDIMM contacts. Reduces grade significantly; bent pins on enterprise RDIMM are difficult to straighten reliably and the secondary buyer will not accept the risk. We inspect every stick at receipt and flag bent-pin units. Pitfall 2 — Counterfeit sticks (mostly legacy DDR3, some DDR4). SPD-tampered sticks fail validation; we flag and dispose. The counterfeit-DDR market relabels lower-grade chips with higher-grade markings; SPD verification + visual chip-marking inspection catches most. Pitfall 3 — Mixed-OEM sticks shipped as a "matched kit". Matched-kit pricing requires same OEM, same speed, same CL, same revision — we re-sort at receipt and re-price the unmatched pulls at unmatched-pull rates. Pitfall 4 — High-failure-rate batches. Some legacy data-centre estates have batches with elevated failure rates due to environmental conditions (heat, humidity, capacitor aging); we test every stick and the MYR settlement is against validated count, not received count. Pitfall 5 — Customer expects retail-DIMM pricing on enterprise RDIMM. Retail UDIMM and enterprise RDIMM clear at different rates; we explain the distinction at scoping. Pitfall 6 — DIMM serials not in customer asset list. Memory replaced under hot-swap maintenance is often not added to the fixed-asset register; we re-scan and reconcile any discrepancy at receipt and the certificate documents both received and asset-list counts.

Why customers in Malaysia choose Maxicom for memory / ram buyback

Continuous operation since 1996 — the Maxicom group was founded in India in 1996 and operates across Malaysia, plus four other Maxicom regions. Per-asset certificate format admissible against , PDPA Malaysia, NIST SP 800-88 Rev. 1, IEEE 2883-2022, and equivalent regulators in every jurisdiction we have served. Reuse-First reuse rate of 65–75% blended across our 2024–2025 cohort — typically 1.5–3× the residual value of destruction-first OEM trade-in programmes. Settlement in MYR against your purchase order, line-item per asset, payment terms 7 business days from manifest reconciliation. Cross-border resale routing under NDA preserves channel-respect — your retired gear never re-appears in your local market unless you specifically authorise local resale. Programme-level pricing for multi-event commitments; quarterly business reviews for ongoing engagements. Single SOW, single ledger, single regulator-facing report — your fixed-asset team reconciles to one document.