Mingxin

Partnerships

Mingxin (Tianjin) Semiconductor Equipment Co., Ltd. starts every partnership by asking what value we can create for the other side. The four tracks below speak to different counterparties. Fees and tenors are diligence-pack material, reached via the contact form and an NDA.

Verifiable progress

  • First tranche: 60 × 8-GPU AMD Instinct MI308X chassis (480 cards), equipment value about CNY 90 million; mid-term plan 5,000-card class.
  • Pre-production acceptance on a single 8×MI308X node for a 1T-parameter INT4 MoE model: 3,210 samples across four accuracy benches all passed (official delta −0.8% to −1.5%, inside the 2% red line).
  • Deployment adaptation and performance tests with MiniMax are complete; token supply is in the commercial-signing process (not a signed contract).
  • Signed storage-acceleration result: throughput +29–40%, TTFT −26–32% (Qwen3-Coder-480B-FP8 long-context cold recovery vs local NVMe; reports R2/R3). See the evidence library.

Licensed financial lessors

Sale-and-leaseback

You take legal title, not a claim. During the lease the GPU assets belong to the lessor; on default you repossess rather than litigate a mortgage. The asset is a data-center GPU with an active secondary market. Downstream payment is real and checkable: MiniMax is in the token-supply commercial-signing process.

We hope to close a near-par sale-and-leaseback funding loop. Advance rate, tenor and pricing are yours to quote.

Banks, factors, SCF platforms

Supply-chain finance and factoring

A short-chain, high-frequency receivable: compute and token sales. The payer is a large-model company; confirmation is direct. First-tranche equipment value is about CNY 90 million, expandable to the 5,000-card class — a reusable asset pipeline, not a one-off ticket.

We can discuss token-sales factoring, import LCs, collection-account control, and a referral to your group leasing arm.

IDC, GPU cloud, bare-metal, OEMs

Long-term GPU chassis rental

We already hold 480 MI308X cards and are looking for incremental rental capacity, not a first cluster. We bring models, traffic and our own all-flash array; you keep title and can lift idle-hour utilisation. Mid-range cards are in scope.

Please quote available chassis, SKU, lead time and settlement options (monthly chassis, GPU-hour, off-peak). Two hard conditions: 2U rack space for our array, and staff access for deployment.

Strategy / IR / industrial capital / IT distribution

Listed-company and industrial collaboration

A continuing pipeline of compute assets, plus a signed storage-acceleration result: inference throughput +29–40% and TTFT −26–32% versus a local-NVMe single-drive baseline (reports R2/R3). If supply and finance sit in one group, friction is lowest.

First confirm the owning team and whether a 2026 plan exists. Materials on request, no strings.

How we work

  1. Tell us your role and preferred track via the contact form, or reply to an outreach email.
  2. After NDA, we send the diligence-pack index and reproducible model scripts within 48 hours.
  3. You quote under your own policy. We do not disclose our advance-rate floor.
  4. Start with a batch; first-tranche participants get priority on the mid-term expansion.

Contact 王利民 · 13161818898 · Send an inquiry

This page is a public description of one-to-one B2B cooperation. It does not constitute an offer, an investment offer, or any return promise. Figures are the 2026-08-22 external baseline; economic numbers are engineering-model estimates, not signed purchase or sales prices.