Global Recovery
A worldwide network of collection points with free on-site pickup keeps idle assets moving.
Recycling · Refurbishment · Rental · Maintenance · GPU Repair
From server recycling and certified refurbishment to used servers, rental and maintenance, GPU and AI servers with NVIDIA H100, H200, B200 and GB300, and NVIDIA GPU card repair - Daoshen delivers secure, sustainable enterprise hardware solutions worldwide.

assets processed to date
clients served to date
COMPANY PROFILE
Daoshen Technology is an integrated IT company focused on the enterprise market built around servers. We work by three principles - customer first, service as our foundation, and pragmatic innovation - and deliver server recycling, server refurbishment, used server supply, equipment rental and maintenance.
Starting in Zhongguancun, Beijing, we have grown a service network across North China, South China and Hong Kong, China. Recovery, diagnostics, refurbishment, rental and maintenance now form one continuous chain that helps enterprises run hardware longer, leaner and more reliably.
We start with your requirements, absorb the complexity and deliver certainty.
24/7 response and 48-hour delivery keep every promise we make.
Continuous improvement in remanufacturing makes reuse genuinely dependable.
Responsibility earns respect. Service earns trust. - The Daoshen way
OUR BUSINESS
From decommissioning to redeployment, six connected services give every device the right next destination.

IDC and enterprise hardware recovery with testing, repair, dismantling, media shredding and data erasure.
View capabilities
Upgrades, reconfiguration, stability testing and remanufacturing tailored to each workload.
View capabilities
H100, H200 and B200 GPU and AI server systems - flexible to buy or rent, ready to ship.
View capabilities
System and GPU module-level repair with 48-hour response, genuine parts, per-incident or annual plans.
View capabilities
Extensive server inventory and assembly capability with flexible short- and long-term rental.
View capabilities
OEM or third-party warranty renewal, spares, 24/7 response and scheduled inspection across the hall.
View capabilitiesSPECIALIST SERVICES
Server recycling, refurbishment and used servers, GPU and AI servers, server or GPU repair, rental and maintenance - each with its own scope, process and delivery standard.
Recycling for servers, storage, networking gear and components from data centers and enterprise server rooms, with free on-site valuation.
View service 02Reconfiguration, component upgrades and remanufacturing, backed by 100,000+ refurbished spares across 3,000+ SKUs.
View service 03H100, H200, H800, H20, A100 and A800 GPU servers and AI server systems - flexible rental or purchase, ready to ship.
View service 04System and GPU module repair for H100, H200, H800, H20, A100, A800, B200 and GB300, with 48-hour response and genuine parts.
View service 05From one unit and one day: rack and GPU compute servers for short or long terms, with free maintenance and 48-hour response.
View service 06OEM and third-party warranty renewal, spares, 24/7 response and scheduled inspection for servers, storage and network gear.
View serviceCAPABILITIES
Select a solution to see its scope, key figures and supported hardware.
One-stop service for data centers and enterprise server rooms - from on-site collection and valuation to certified data erasure and compliant destruction. Over 850,000 servers and network devices processed to date.
SCOPE

Moving beyond conventional second-hand resale: hardware upgrades, reconfiguration and stability testing bring decommissioned systems back into production with confidence.
SYSTEMS & COMPONENTS

NVIDIA H100, H200, H800, H20, A100 and A800 plus B200, B300 and GB300 GPU and AI server systems in 4- and 8-GPU configurations, available for purchase, project rental and cluster design, delivered after 72-hour burn-in testing.
MODELS

Fault repair, RAID recovery and data recovery for businesses and individuals worldwide. We specialize in GPU server and GPU module repair, covering NVIDIA H100, H200, H800, H20, A100 and A800, with genuine OEM parts.
SUPPORTED MODELS

Rent instead of buy to free up cash flow. From one unit, from one day, with free maintenance during the term and 48-hour response - ideal for projects and elastic compute needs.
EQUIPMENT & USE CASES

From a single warranty renewal to full data center coverage: OEM and third-party maintenance, forward-stocked spares, scheduled inspection and unified firmware across servers, storage, networking and the hall, with four response tiers written into the contract.
COVERAGE

WHY DAOSHEN
Expertise built into every detail, so asset disposition stays both secure and effortless.
A worldwide network of collection points with free on-site pickup keeps idle assets moving.
Dedicated appraisers provide itemised quotes so every asset realises its full value.
On-site within 24 hours locally and 48 hours elsewhere - less downtime, less waiting.
One-stop recovery and supply for servers, networking equipment and peripherals.
Military-grade erasure and physical media destruction make recovered data unrecoverable.
A closed loop of recovery, refurbishment and redeployment that supports greener IT.
OUR PROCESS
A six-step, fully traceable and verifiable asset processing workflow.
IDC and enterprise hardware intake
Full diagnostics and health assessment
Teardown with parts sorted and catalogued
Physical destruction of storage media
Irreversible data wiping
Repair and re-certification of reusable units
GLOBAL REACH
Warehousing and service teams across regions deliver reliable hardware solutions to customers in 100+ countries and regions.
countries and regions served
Dedicated specialists providing end-to-end hardware solutions.
Multiple warehouses worldwide for fast, predictable delivery.
Experts on standby around the clock, seven days a week.
OUR HISTORY
Two decades focused on one thing: helping enterprise IT hardware work better, for longer.
Starting out in the enterprise server market
Extending coverage across South China
A gateway to international customers
Expanded warehousing and refurbishment capacity
OUR PARTNERS
Responsibility, communication, execution and results - the principles we work by.










FAQ
Common questions about recycling valuation, data security, used server warranty and GPU repair turnaround.
Valuation considers brand and model, year of manufacture, CPU / memory / GPU configuration, condition and volume. Systems are quoted by configuration, components by model, and end-of-life units by recoverable metal content. A preliminary quote usually follows within one business day of receiving your asset list.
Yes. We collect within 24 hours locally and 48 hours elsewhere. On-site assessment and removal are free in major cities, and dedicated teams can be stationed on site for full data center decommissioning.
Three tiers are available: DoD 5220.22-M multi-pass software erasure, physical shredding of storage media, and degaussing of magnetic media. Every unit is logged and video-monitored, with destruction certificates and environmental disposal reports issued on request.
Yes. Refurbished systems carry a 3-12 month warranty by default, after 72-hour burn-in and stability testing with a test report. Extended warranty and spare-part packs are available.
Common faults such as dropped cards, memory errors and power issues are typically resolved in 3-7 business days. BGA rework or module replacement takes around 7-15 business days, with an expedited option for urgent projects.
Yes - servers, storage, disk arrays, switches, routers, firewalls, workstations, components and cabling, plus full room decommissioning and asset reconciliation.
Yes. Rentals start from one unit and one day, for terms from 3 days to 3 years, including free maintenance and 48-hour response during the term.
Rent when the project has a defined lifespan, the silicon moves fast or the capital outlay is large: no depreciation or disposal cost, and you can return or renew at the end. Buy when the workload is steady and asset ownership matters. Rent-to-own bridges the two.
Response commitments, spare parts, scheduled inspection, firmware and microcode updates, free replacement of faulty components and 24/7 cover, under four response tiers (4-hour, next business day, 48-hour, parts first) across servers, storage, networking and data center halls.
Yes. B200, B300 and GB300 are NVIDIA Blackwell and Blackwell Ultra platforms, available for purchase, project rental and cluster design with batch-based allocation. GPU cards and modules are also supplied individually, with module-level repair.
Tell us about your hardware and workload. Our team will tailor a recovery, refurbishment, rental or maintenance plan for you.
CONTACT US
Four offices across North China, South China and Hong Kong, China - ready when you need us.
BEIJING
SHENZHEN
DONGGUAN
HONG KONG, CHINA

Scan to follow Daoshen for industry insights and equipment updates.

Scan to see equipment walkthroughs and service cases.