ProductTechnologySolutionsValidationCasesCustomersIPCompanyInvestorsNews Contact 中文

Customer cases

Centered on joint R&D, independent third-party testing and demonstration use — proving the value of disaggregated storage with reproducible data in real compute environments.

Quick answer

What deployments and validations does ZK-Storage have?

Joint R&D · 3rd-party test
Beijing Information Science and Technology University (Huawei Ascend, ~90.9% median reduction across 7)
Demonstration · validation site
Jingxi Zhigu intelligent-compute park (real-environment testing)
Framing
Joint R&D / third-party testing / demonstration; commercial scale-out subject to announcement
PROOF POINTS

Verifiable proof points

Objective and checkable: an independent third party, a stated platform, a stated baseline.

0
key metrics — all leading
vs NFS baseline
90.9%
median latency reduction
third-party testing
85×
peak model-load speedup
measured on DeepSeek etc.
0 units
live demo units
reproduce on your data
Joint R&D · independent validation

Beijing Information Sci-Tech University · Computing Center

National-level practice platform · independent validator

On a National Engineering Practice Education Center platform, the two sides jointly develop the hardware-accelerated, disaggregated compute-center architecture; the Center also performed the independent third-party validation of WS5000 on Huawei Ascend Atlas 910B against an NFS baseline — leading on all 7 key metrics with a ~90.9% median reduction, reproducible and verifiable. S38 · S42

7metrics leading
90.9%median reduction
910BAscend platform
Demonstration · testing / validation site

Jingxi Zhigu · intelligent-compute park

Intelligent-compute park · testing / validation site

In the live compute environment of the Jingxi Zhigu intelligent-compute park, ZK-Storage's all-flash accelerated storage completed performance testing and joint validation for large-model inference and training — demonstrating the deployability and acceleration of the disaggregated architecture in a real compute-park setting (stated as joint validation / testing; commercial scale-out subject to final announcement).

infer + traintest scope
disaggregatedarchitecture
live parkcompute env

Case statement (fact-based)

These cases are stated as joint R&D / independent third-party testing / demonstration use: Beijing Information Sci-Tech University is the independent validator of WS5000, and Jingxi Zhigu is a testing / joint-validation site. This page uses no fabricated commercial-scale figures and no unauthorized on-site photos; commercial scale-out is subject to final announcement. Full benchmark data is on the Validation page.

Benchmark it on your own workload

2 live demo units are ready for immediate PoC. Let the data do the talking.

Last updated: