DP Technology (深势科技)

Beijing AI-for-science company built on "AI for molecular simulation" — open-source DeePMD potentials plus a scientific-computing cloud, with Weinan E as scientific co-founder.

1. Core Product / Service

DP Technology (深势科技) applies machine-learned interatomic potentials to molecular dynamics, with its research lineage in DeePMD-kit (deep potential molecular dynamics), an open-source framework widely used in computational chemistry and materials (see mattersim for the related MLIP lineage). Its commercial products include Bohrium (玻尔), a scientific-computing cloud/workbench, and Hermite, a drug-design platform built on the same simulation engine [local].

2. Target Users & Pain Points

Research institutions and industrial R&D (chemistry, materials, pharma) needing ab-initio-accuracy simulation at a fraction of first-principles cost.
Pain solved: classical molecular dynamics is fast but inaccurate; first-principles DFT is accurate but slow — machine-learned potentials offer DFT-level accuracy at MD speed, a step change for screening and simulation.

3. Competitive Landscape

Player	Positioning	Vs. DP Technology
xtalpi	CSP + AI drug discovery (China, listed)	XtalPi is drug-discovery-vertical; DP is simulation-infra horizontal
insilico-medicine	Generative-AI biotech	Different method (generative design vs. simulation)
mattersim	Open MLIP model (Microsoft)	DP is a company; Mattersim is an open model
Schrödinger / Dassault (BIOVIA)	Computational chemistry incumbents	DP is MLIP-native, cloud-first

4. Unique Observations

China's AI4S flagship: this week's research positioned 深势科技 as China's leading AI-for-science company, led by a 院士-grade scientific team — the domestic counterpart to the US isomorphic-labs/google-deepmind axis [local].
Method → infra → vertical: DP's path from open-source method (DeePMD) to cloud platform (Bohrium) to vertical app (Hermite drug design) mirrors how US AI4S players monetize research, and matches this week's "serve-AI4S-companies" platform-opportunity thesis [local].
Academic-founder advantage: Weinan E (鄂维南, applied mathematician, Princeton professor, CAS academician) gives it top-tier scientific credibility for talent and customer trust in China.

5. Financials / Funding

Funding: multiple VC rounds (Series A/B+) from Chinese tech investors; specific amounts not captured in local research this week [local].
Status: private, pre-IPO.

6. People & Relationships

Scientific co-founder: 鄂维南 (Weinan E) — applied mathematician, Princeton professor, CAS academician; co-founded with 张林峰 (Zhang Linfeng) and 孙伟杰 (Sun Weijie).
Peers: xtalpi, insilico-medicine, mattersim.