/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Huawei unveils two AI chips, the Ascend 910 and Ascend 310, for data centers and internet-connected consumer devices, as it takes aim at Qualcomm and Nvidia

Arjun Kharpal / CNBC :

CNBC Arjun Kharpal

Context & Ripple Effects

The 2018 Ascend 910 and 310 launch is the origin point of a product line that has since become China's answer to Nvidia. What began as a two-chip announcement — one for data centers, one for connected consumer devices — matured into a full-stack challenge: by 2024, Chinese internet companies were testing the Ascend 910C, which Huawei told clients is comparable to Nvidia's H100.

The intervening years were not linear. Huawei hit a production ramp wall on its Ascend server chips under US export crackdowns, then responded with packaging workarounds like the dual-die 910C and, after Washington banned Nvidia's H20 from China, unveiled the Ascend 920 with mass production targeted for H2 2025. The 2018 unveil matters because it marks when Huawei committed to designing its own AI silicon rather than depending on American suppliers.

First-order effects

  • Huawei enters merchant AI silicon against two incumbents at once — Nvidia in data-center training and inference, Qualcomm at the device edge — making itself the default domestic option for Chinese buyers of both.

Second-order effects

  • US export controls on Nvidia's China-bound chips (A100 through H20) converted Huawei's alternative from a fallback into a primary procurement path, as seen in clients moving to mass shipments of the 910C and feasibility tests of the newer 910D.
  • Nvidia loses pricing and volume leverage in China specifically, pushing it toward workarounds while Huawei's roadmap cadence — 910B to 910C to 910D to 920 — accelerates to fill each gap the bans open.

Third-order effects

  • If the pattern holds, AI compute splits into two largely non-interchangeable stacks — an Nvidia-led global one and a Huawei-led Chinese one — with Chinese cloud and internet customers structurally locked into domestic silicon regardless of relative performance.
  • The edge-versus-datacenter split Huawei drew in 2018 anticipates where competition lands long-term: inference migrating onto devices, where Qualcomm's position, not just Nvidia's, comes under pressure.

The trend: AI compute is bifurcating into parallel national stacks, and Huawei's 2018 Ascend bet is the seed of the Chinese one now scaling behind export-control walls.