What are the latest funding news in the AI chip market? (October 2026)
Every funding round, every investor, and who is really ahead in AI chips, in one document. Have a look ↓
The latest AI chip funding activity shows investors concentrating capital around data-center accelerators built to make AI training and inference faster, cheaper and more energy efficient.
The 12 qualifying deals represent about $5.59 billion in disclosed funding, with nine of the rounds announced during 2026.
Inference is the strongest theme in the group, while only a smaller number of companies are building accelerators that target both AI training and inference at scale.
And if you want to better understand this new industry, you can get our report on AI chip funding, valuations and who's ahead.
Insights
- The 12 selected AI chip companies raised about $5.59 billion across their latest qualifying rounds, showing how capital-intensive it has become to move accelerator architectures from design into data-center production.
- At least 10 of the 12 companies are primarily focused on AI inference or inference-led deployments, reflecting how serving large models is becoming a larger part of data-center compute demand.
- Positron, Etched, SambaNova and Cerebras alone raised roughly $3.58 billion, meaning four companies accounted for around 64% of the total disclosed capital in this selection.
- Five of the 12 rounds were worth at least $500 million, illustrating how AI accelerator startups increasingly need financing levels normally associated with much more mature infrastructure businesses.
- The market is not limited to traditional GPU architectures, with companies pursuing wafer-scale processors, NPUs, digital in-memory compute, model-specific ASICs and memory-focused inference architectures.
- Several companies are now selling complete racks or systems rather than standalone chips, suggesting that competition in the AI chip market is increasingly moving toward integrated data-center infrastructure.
- Most of the newest rounds are financing manufacturing, tapeouts, supply chains and production ramps rather than basic research, which indicates that several alternative AI accelerators are moving closer to commercial deployment.
- The United States remains the largest center of activity, but the UK, South Korea, Canada and the UAE are also producing well-funded AI chip companies with distinct regional or sovereign-compute strategies.
Written by the New Market Pitch teamWe track new markets so founders and investors can move faster. About us
We keep an up-to-date report on AI chips. Have a look ↓
Sample pages from an earlier edition. The report has been updated with new data since.
Summary table of the latest funding deals in the AI chip market as of October 2026
We define the AI chip market as data-center accelerators whose primary purpose is to run AI workloads, including training and inference.
We include GPUs, TPUs, and other AI accelerators and ASICs sold for deployment in servers used to train or serve machine-learning models.
We exclude general-purpose CPUs, networking and memory components, and endpoint or edge chips in phones, PCs, cars, and IoT devices.
| Name | When | Amount in $ | Round Type | Category |
|---|---|---|---|---|
| Positron AI | September 10, 2026 | $875M | Series C | AI inference ASIC & accelerator systems |
| Etched | August 18, 2026 | $700M | Series D | AI inference ASIC & rack-scale systems |
| OLIX | August 3, 2026 | $312M | Series B | Data-center inference ASIC |
| SambaNova Systems | July 8, 2026 | $1,000M | Series F first close | AI accelerator & RDU systems |
| Fractile | May 13, 2026 | $220M | Series B | In-memory inference accelerator |
| Rebellions | March 30, 2026 | $400M | Pre-IPO | Data-center NPU & inference systems |
| MatX | February 24, 2026 | $500M | Series B | Training & inference accelerator ASIC |
| Taalas | February 19, 2026 | $169M | Venture financing | Model-specific inference ASIC |
| Cerebras Systems | February 3, 2026 | $1,000M | Series H | Wafer-scale training & inference accelerator |
| Mastiṣka | November 25, 2025 | $10M | Seed | Sovereign AI inference accelerator |
| d-Matrix | November 12, 2025 | $275M | Series C | Digital in-memory inference accelerator |
| FuriosaAI | July 30, 2025 | $125M | Series C bridge | Data-center NPU & tensor contraction processor |
All the latest funding deals in the AI chip market as of October 2026
Positron AI raised $875M in September 2026 to scale its next generation of AI inference hardware.
When was it?
Positron AI announced the financing on September 10, 2026.
Who are they?
Positron AI builds memory-heavy AI inference accelerators and complete server systems designed to run large models with lower power use and infrastructure cost.
Geographical focus?
Positron AI is based in Reno, Nevada, and targets large cloud and data-center customers internationally.
Why do we include them in the AI chip market?
Positron AI belongs in the AI chip market because its core products are purpose-built data-center inference chips and accelerator systems.
What is the company stage?
Positron AI is at the growth and early large-scale commercialization stage, with deployed systems and a new generation of production silicon under development.
How much did they raise?
Positron AI raised $875 million in this financing.
What round is it?
Positron AI raised a Series C round.
Why did they raise?
Positron AI raised the capital to complete its Asimov chip tapeout and increase production of its Titan inference systems.
These pages are excerpts from our report. See the rest ↓
Sample pages from an earlier edition. The report has been updated with new data since.
Etched raised $700M in August 2026 as it prepared to scale specialized AI inference systems.
When was it?
Etched announced the new funding on August 18, 2026.
Who are they?
Etched develops specialized AI inference chips and rack-scale systems designed to serve very large frontier and mixture-of-experts models.
Geographical focus?
Etched is based in San Jose and targets global data-center operators, with manufacturing and supply-chain relationships extending into Taiwan.
Why do we include them in the AI chip market?
Etched belongs in the AI chip market because dedicated inference silicon is the foundation of its data-center computing platform.
What is the company stage?
Etched is at the early commercial and production-scaling stage, with its first rack already delivered to a disclosed customer.
How much did they raise?
Etched raised $700 million in new funding.
What round is it?
Etched described the transaction as new funding, while external funding databases generally classify it as a Series D.
Why did they raise?
Etched raised the capital to expand manufacturing, strengthen its global supply chain and move toward much larger inference deployments.
OLIX raised $312M in August 2026 to move its specialized AI inference chips toward customer deployment.
When was it?
OLIX announced the Series B financing on August 3, 2026.
Who are they?
OLIX develops specialized AI inference processors that connect into rack-scale systems through a custom architecture and photonic interconnect.
Geographical focus?
OLIX is headquartered in the UK and is building a British semiconductor company for customers in the global AI infrastructure market.
Why do we include them in the AI chip market?
OLIX belongs in the AI chip market because its main product is a dedicated data-center inference processor and associated rack architecture.
What is the company stage?
OLIX is pre-commercial and in late product development, with its first customer chips targeted for the second half of 2027.
How much did they raise?
OLIX raised $312 million in this financing.
What round is it?
OLIX raised a Series B round.
Why did they raise?
OLIX raised the capital to deliver its first chip, develop its custom-silicon platform and fund manufacturing and supply-chain commitments.
SambaNova Systems raised $1B in July 2026 to expand its commercial AI accelerator infrastructure.
When was it?
SambaNova Systems announced the first close of the financing on July 8, 2026.
Who are they?
SambaNova Systems builds proprietary Reconfigurable Dataflow Unit accelerators and integrated systems for fast enterprise AI inference.
Geographical focus?
SambaNova Systems is headquartered in the United States and sells AI infrastructure to enterprise and sovereign customers globally.
Why do we include them in the AI chip market?
SambaNova Systems belongs in the AI chip market because its proprietary RDU is a purpose-built data-center accelerator for running AI models.
What is the company stage?
SambaNova Systems is at the commercial growth stage, with production systems already selected for enterprise and on-premise AI workloads.
How much did they raise?
SambaNova Systems raised $1 billion in the first close of the financing.
What round is it?
SambaNova Systems raised the first close of a Series F round.
Why did they raise?
SambaNova Systems raised the capital to scale production and deployment of its latest RDU infrastructure as demand for inference grows.
Want the whole picture? Check our report ↓
Sample pages from an earlier edition. The report has been updated with new data since.
Fractile raised $220M in May 2026 to bring its next generation of AI inference hardware toward commercialization.
When was it?
Fractile announced the financing on May 13, 2026.
Who are they?
Fractile develops AI inference chips and systems that place compute and memory closer together to generate tokens more efficiently.
Geographical focus?
Fractile is headquartered in the UK but is building hardware for hyperscale and other global data-center customers.
Why do we include them in the AI chip market?
Fractile belongs in the AI chip market because its core business is purpose-built inference processors for running large AI models in data centers.
What is the company stage?
Fractile is at the pre-commercial product-development stage and is working toward its first chips and customer systems.
How much did they raise?
Fractile raised $220 million in this financing.
What round is it?
Fractile raised a Series B round.
Why did they raise?
Fractile raised the capital to complete development of its first-generation inference silicon and move toward commercial customer deployments.
Rebellions raised $400M in March 2026 to expand production and international deployment of its AI accelerators.
When was it?
Rebellions announced the financing on March 30, 2026.
Who are they?
Rebellions develops NPUs and complete server and rack systems designed for production-scale AI inference.
Geographical focus?
Rebellions originated in South Korea and is expanding into the United States, Middle East and other international AI infrastructure markets.
Why do we include them in the AI chip market?
Rebellions belongs in the AI chip market because its core products are server-side NPUs and accelerator systems built specifically for AI inference.
What is the company stage?
Rebellions is at the growth and commercial deployment stage and was preparing for a future IPO when the financing was announced.
How much did they raise?
Rebellions raised $400 million in this financing.
What round is it?
Rebellions described the financing as a pre-IPO round.
Why did they raise?
Rebellions raised the capital to expand accelerator production, grow in the United States, win larger customers and support its pre-IPO development.
MatX raised $500M in February 2026 to move its large-model AI accelerator toward production.
When was it?
MatX announced the financing on February 24, 2026.
Who are they?
MatX designs accelerators for large AI models, with one architecture intended to support training, reinforcement learning, prefill and token generation.
Geographical focus?
MatX is based in Mountain View, California, and targets global hyperscalers and frontier-model developers.
Why do we include them in the AI chip market?
MatX belongs in the AI chip market because it is building dedicated data-center accelerator silicon for both AI training and inference.
What is the company stage?
MatX is at the pre-production stage, with manufacturing preparations underway ahead of planned chip shipments.
How much did they raise?
MatX raised $500 million in this financing.
What round is it?
MatX raised a Series B round.
Why did they raise?
MatX raised the capital to move its accelerator into production at TSMC and prepare for customer shipments.
Prefer verified numbers to headlines? Our report is here ↓
Sample pages from an earlier edition. The report has been updated with new data since.
Taalas raised $169M in February 2026 to develop model-specific chips for high-efficiency AI inference.
When was it?
Taalas announced the financing on February 19, 2026.
Who are they?
Taalas develops processors that hardwire parts of individual AI models directly into silicon to run inference faster and with lower power consumption.
Geographical focus?
Taalas is based in Toronto and targets the global data-center inference market.
Why do we include them in the AI chip market?
Taalas belongs in the AI chip market because its core technology is model-specific accelerator silicon designed to serve AI models in data centers.
What is the company stage?
Taalas is at the prototype and pre-commercial stage, with demonstration hardware unveiled while production versions remain under development.
How much did they raise?
Taalas raised $169 million in this financing.
What round is it?
Taalas did not publicly disclose a formal Series label for the venture financing.
Why did they raise?
Taalas raised the capital to continue developing its model-specific architecture and move from demonstration hardware toward deployable inference products.
Cerebras Systems raised $1B in February 2026 to expand its wafer-scale AI computing infrastructure.
When was it?
Cerebras Systems announced the financing on February 3, 2026.
Who are they?
Cerebras Systems develops wafer-scale AI processors and complete computing systems for training and extremely fast inference of large neural networks.
Geographical focus?
Cerebras Systems is based in the United States and serves hyperscalers, enterprises, governments and research institutions internationally.
Why do we include them in the AI chip market?
Cerebras Systems belongs in the AI chip market because its Wafer Scale Engine processors are purpose-built accelerators for data-center AI training and inference.
What is the company stage?
Cerebras Systems was at the late-stage commercial and pre-IPO growth stage when the financing was announced.
How much did they raise?
Cerebras Systems raised $1 billion in this financing.
What round is it?
Cerebras Systems raised a Series H round.
Why did they raise?
Cerebras Systems raised the capital to expand its processor roadmap, AI systems and data-center capacity as customer demand increased.
Mastiṣka raised $10M in November 2025 to build sovereign AI accelerator technology in the UAE.
When was it?
Mastiṣka's financing was reported on November 25, 2025.
Who are they?
Mastiṣka is a UAE-based fabless semiconductor company developing data-center AI accelerators and an architecture designed around more efficient workload execution.
Geographical focus?
Mastiṣka has a strong UAE and GCC focus while also targeting sovereign AI customers across South Asia, BRICS markets and the Global South.
Why do we include them in the AI chip market?
Mastiṣka belongs in the AI chip market because its core business is data-center-class AI acceleration rather than general-purpose or endpoint semiconductor products.
What is the company stage?
Mastiṣka was at the seed and pre-product stage, developing accelerator architecture and moving from FPGA validation toward purpose-built silicon.
How much did they raise?
Mastiṣka raised $10 million in this financing.
What round is it?
Mastiṣka raised a Seed round.
Why did they raise?
Mastiṣka raised the capital to develop its accelerator technology, establish a UAE fabless semiconductor operation and move its first products toward market.
Want to know who is really ahead? Our report is here ↓
Sample pages from an earlier edition. The report has been updated with new data since.
d-Matrix raised $275M in November 2025 to increase production of its generative AI inference accelerators.
When was it?
d-Matrix announced the financing on November 12, 2025.
Who are they?
d-Matrix builds PCIe accelerator cards using digital in-memory computing to serve generative AI models efficiently in data centers.
Geographical focus?
d-Matrix is based in the United States and targets hyperscale, enterprise and sovereign AI customers internationally.
Why do we include them in the AI chip market?
d-Matrix belongs in the AI chip market because its Corsair platform is a dedicated data-center accelerator for generative AI inference.
What is the company stage?
d-Matrix is at the early growth and commercial deployment stage, with products shipping and larger installations being prepared.
How much did they raise?
d-Matrix raised $275 million in this financing.
What round is it?
d-Matrix raised a Series C round.
Why did they raise?
d-Matrix raised the capital to expand production and distribution, grow internationally and support larger customer deployments.
FuriosaAI raised $125M in July 2025 to scale production of its RNGD AI inference processor.
When was it?
FuriosaAI announced the financing on July 30, 2025.
Who are they?
FuriosaAI develops RNGD, a Tensor Contraction Processor designed specifically to run large generative AI models efficiently in servers and data centers.
Geographical focus?
FuriosaAI is based in South Korea and is expanding toward global enterprise and data-center customers.
Why do we include them in the AI chip market?
FuriosaAI belongs in the AI chip market because RNGD is a purpose-built data-center inference accelerator designed as an alternative to GPU-based inference.
What is the company stage?
FuriosaAI is at the early growth and production-ramp stage, with customer design wins and larger-volume manufacturing underway.
How much did they raise?
FuriosaAI raised $125 million in this financing.
What round is it?
FuriosaAI raised a Series C bridge round.
Why did they raise?
FuriosaAI raised the capital to increase RNGD production, meet enterprise demand and fund future generations of its AI accelerator roadmap.
If this is useful, the full report goes much further. Have a look ↓
Sample pages from an earlier edition. The report has been updated with new data since.
Written by the New Market Pitch teamWe track new markets so founders and investors can move faster. About us