AI Chip Startup Funding 2024-2026
In our AI chip market deck, you will find everything you need to understand the market
SUMMARY
This report analyzes every publicly disclosed equity round raised by pure-play AI chip companies between August 2024 and July 2026, a 24-month window covering data-center accelerators for AI training and inference. We only kept rounds of $300K or more, included IPOs and public-market primary offerings as equity deals, and excluded endpoint, edge, CPU, memory, and networking chip companies.
Over this period, fundraising in the AI chip market was large, late-stage, and highly strategic. The dataset includes 29 disclosed deals and $11.25B raised across 22 unique companies.
Capital in the AI chip market is concentrated, but not dependent on one single company. The top deal represents 10.04% of total capital, the top 3 deals reach 28.70%, and the top 10 deals reach 69.10%.
The median round size in the AI chip market is $250M, which is exceptionally high. The average round size is even higher at $388.07M, showing how mega-rounds and public offerings pull the market upward.
Fundraising pace is modest on deal count but heavy on dollars. Deal flow averages 1.21 rounds per month, while capital raised averages $468.91M per month across the 24-month period.
Inference Accelerators lead the AI chip market by activity, with 15 deals and 51.72% of all disclosed rounds. But they account for only 35.85% of capital, meaning the category is crowded but less capital-heavy per deal.
Data Center GPUs and Server AI Processors attract disproportionately large checks. Together, they represent only 27.58% of deals but 46.21% of disclosed capital.
North America dominates the AI chip market by funding volume, with $7.03B raised across 15 deals. Asia-Pacific follows with $3.71B across 10 deals, helped by Chinese GPU companies and public-market financings.
The AI chip market is structurally late-stage. Late-stage, Series C and beyond, and Growth Equity rounds account for 78.94% of disclosed capital, while Seed to Series B rounds account for only 18.70%.
Follow-on funding dominates the visible AI chip market. Only one deal is clearly classified as a first financing, which suggests public funding visibility begins after strong technical or investor validation.
Repeat investors matter because credible AI chip companies are expensive to finance. Disruptive, Atreides, Fidelity, Spark Capital, QIA, M12, 1789 Capital, Samsung-linked investors, Arm-linked investors, and Mirae Asset each appear in more than one disclosed deal.
This market map, featured in our AI chip market deck, highlights top companies and startups in the AI chip market
What are all the funding deals in the AI chip market from August 2024 to July 2026?
The table below lists every disclosed equity round raised by pure-play AI chip companies between August 2024 and July 2026. We count as pure-play AI chip companies those focused on data-center accelerators whose primary purpose is running AI workloads, including training and inference.
Each row shows the company, what it does, its category, the deal date, the funding stage, the round size, the region, the main investors, and the announcement source. For a wider view of how AI accelerators, GPUs, ASICs, and inference chips fit into the compute stack, we cover it in our AI Chip market report.
| Company | What they do | Category | Date | Stage | Deal size | Region | Main investors | Source |
|---|---|---|---|---|---|---|---|---|
| Groq | LPU-based AI inference chips and an inference platform for fast generative AI serving | Inference Accelerators | Aug 2024 | Series D | $640M | North America | BlackRock Private Equity Partners; Cisco Investments; Samsung Catalyst Fund; Neuberger Berman | Groq |
| MatX | Data-center AI chips for large language model training and serving | Training Accelerators | Nov 2024 | Series A | $80M | North America | Spark Capital; former Google technical leaders and other investors | TechCrunch |
| Tenstorrent | AI processors, AI chip IP, and AI compute systems for training and inference workloads | AI ASIC Platforms | Dec 2024 | Series D+ | $693M | North America | Samsung Securities; AFW Partners; Bezos Expeditions; LG Electronics; Fidelity | Tenstorrent |
| HyperAccel | LLM-focused AI semiconductors and LPU server systems for low-latency inference | Inference Accelerators | Dec 2024 | Series A | $41.2M | Asia-Pacific | Not disclosed | WOWTALE |
| VSORA | High-performance AI inference chips for data-center AI workloads | Inference Accelerators | Apr 2025 | Unknown | $46M | Europe | Not disclosed | VSORA |
| Biren Technology | Data-center AI GPUs and accelerator systems for AI training and inference | Data Center GPUs | Jun 2025 | Growth Equity | $207M | Asia-Pacific | Not disclosed | Yahoo Finance |
| Arago | Photonic AI processors designed to reduce energy consumption in AI compute | Inference Accelerators | Jul 2025 | Seed | $26M | Europe | Earlybird; Protagonist; Visionaries Tomorrow; Generative IQ | Tech.eu |
| Positron AI | Inference-optimized AI accelerator hardware for efficient transformer serving | Inference Accelerators | Jul 2025 | Series A | $51.6M | North America | Atreides Management; other investors | Business Wire |
| FuriosaAI | AI compute chips for large language model and multimodal inference | Inference Accelerators | Jul 2025 | Series C | $125M | Asia-Pacific | Not disclosed | FuriosaAI |
| Groq | LPU-based AI inference chips and infrastructure for fast, low-cost model serving | Inference Accelerators | Sep 2025 | Series D+ | $750M | North America | Disruptive; 1789 Capital; other investors | Groq |
| Rebellions | AI inference processors and infrastructure for data-center AI workloads | Inference Accelerators | Sep 2025 | Series C | $250M | Asia-Pacific | Arm; Samsung; other investors | Rebellions |
| Cerebras Systems | Wafer-scale AI processors and AI infrastructure for training and inference | Server AI Processors | Sep 2025 | Series D+ | $1,100M | North America | Atreides Management; Fidelity; 1789 Capital; other investors | Cerebras |
| Majestic Labs | Memory-forward AI server processors and systems for very large AI inference workloads | Server AI Processors | Nov 2025 | Series A | $100M | North America | Not disclosed | Data Center Dynamics |
| d-Matrix | Digital in-memory compute chips and platforms for generative AI inference in data centers | Inference Accelerators | Nov 2025 | Series C | $275M | North America | Qatar Investment Authority; M12; Mirae Asset; other investors | d-Matrix |
| Moore Threads | Domestic Chinese GPUs for AI training, inference, graphics, and high-performance computing | Data Center GPUs | Dec 2025 | Growth Equity | $1,130M | Asia-Pacific | Public-market investors | Pandaily |
| MetaX | GPUs for AI training, AI inference, and high-performance computing workloads | Data Center GPUs | Dec 2025 | Growth Equity | $596.3M | Asia-Pacific | Public-market investors | MarketWatch |
| Biren Technology | Data-center AI GPUs and accelerator platforms for AI training and inference | Data Center GPUs | Dec 2025 | Growth Equity | $716.85M | Asia-Pacific | Public-market investors | Economic Times |
| Etched | Transformer-focused AI chips for large language model inference and optimization | AI ASIC Platforms | Jan 2026 | Series B | $500M | North America | Not disclosed | Data Center Dynamics |
| Neurophos | Photonic AI chips for exaflop-scale AI inference | Inference Accelerators | Jan 2026 | Series A | $110M | North America | M12; other investors | Neurophos |
| Positron AI | Energy-efficient AI inference hardware and custom silicon for data-center deployment | Inference Accelerators | Feb 2026 | Series B | $230M | North America | Qatar Investment Authority; Arm; Atreides Management; other investors | Business Wire |
| OLIX | Photonic AI inference chips intended to reduce reliance on HBM-heavy GPU architectures | Inference Accelerators | Feb 2026 | Unknown | $220M | Europe | Not disclosed | Financial Times |
| Cerebras Systems | Wafer-scale AI processors and AI compute infrastructure for training and inference | Server AI Processors | Feb 2026 | Series D+ | $1,000M | North America | Atreides Management; Fidelity; 1789 Capital; other investors | Wall Street Journal |
| SambaNova Systems | AI chips, dataflow architecture, and systems for cloud-scale AI inference | Server AI Processors | Feb 2026 | Growth Equity | $350M | North America | Intel; other investors | SambaNova |
| MatX | LLM-focused AI training chips designed to compete with Nvidia GPUs | Training Accelerators | Feb 2026 | Series B | $500M | North America | Spark Capital; other investors | TechCrunch |
| Rebellions | AI inference chips and deployable inference infrastructure for global data-center customers | Inference Accelerators | Mar 2026 | Growth Equity | $400M | Asia-Pacific | Mirae Asset; other investors | Rebellions |
| EVAS Intelligence / Yixing Intelligence | RISC-V AI chips and computing platforms for large-model training and inference | Training Accelerators | Apr 2026 | Series B | $211M | Asia-Pacific | Not disclosed | Pandaily |
| Fractile | Specialized inference chips and server-rack architectures to accelerate AI query serving | Inference Accelerators | May 2026 | Series B | $220M | Europe | Not disclosed | Wall Street Journal |
| Groq | AI inference cloud infrastructure based on inference-optimized chip technology | Inference Accelerators | Jun 2026 | Growth Equity | $650M | North America | Disruptive; other investors | Groq |
| OXMIQ Labs | Licensable AI GPU architecture for custom AI silicon and sovereign AI workloads | AI ASIC Platforms | Jul 2026 | Series A | $35M | Asia-Pacific | Not disclosed | Economic Times |
In our AI chip market deck, we identify pain points entrepreneurs should prioritize
OUR METHODOLOGY TO BUILD THIS TRACKER
We built this AI chip funding tracker by reviewing every publicly disclosed equity round raised by pure-play AI chip companies between August 2024 and July 2026. A company counts as pure-play when more than 80% of its activity is dedicated to data-center AI accelerators, GPUs, TPUs, AI ASICs, or related server systems for AI training and inference.
We applied four filters to build the dataset. First, we only included equity rounds, including IPOs and public-market primary offerings. Second, we only counted rounds of $300K or more. Third, we only kept pure-play AI chip companies. And fourth, every entry had to be confirmed by a direct company announcement, a press release, or a tier-1 media report, with the source URL preserved for every row.
We excluded general-purpose CPUs, networking chips, memory components, endpoint AI chips, phone chips, PC chips, automotive chips, and IoT edge chips. The final dataset contains 29 disclosed deals across 22 unique companies, and every average, median, share, and concentration ratio is computed on that disclosed sample. Privately raised rounds that were never publicly announced are necessarily missing, which is a known limitation of any public-only AI chip funding tracker.
How active has fundraising been in the AI chip market?
As of July 2026, fundraising in the AI chip market has been active in dollars but relatively selective in deal count. Over the past 24 months, companies raised 29 disclosed equity rounds and $11.25B combined, which works out to 1.21 deals per month.
The AI chip market is not producing a constant stream of small rounds. Monthly deal count has a median of only 1.00, meaning many months had one disclosed round or none at all.
Dollar flow is much larger than the deal count suggests. Average capital raised per month reached $468.91M, but the median month was only $204.80M, confirming that large financings drove the overall picture.
The clearest activity spike came between September 2025 and February 2026. That six-month period included Cerebras, Groq, Rebellions, d-Matrix, Moore Threads, MetaX, Biren, Etched, Positron AI, SambaNova, MatX, and OLIX.
If you want to go deeper on the companies behind this activity, we cover them in our AI Chip market report.
How concentrated has fundraising been in the AI chip market?
As of July 2026, fundraising in the AI chip market is concentrated, but not single-company dependent. Over the past 24 months, the top deal represents 10.04% of total capital, the top 3 deals represent 28.70%, and the top 5 deals represent 41.74%.
The top 10 deals account for 69.10% of disclosed capital in the AI chip market. That means the market total is shaped by a small group of large rounds, but spread across several companies.
This is different from a market where one runaway company defines the category. Cerebras, Moore Threads, Groq, Biren, Tenstorrent, MetaX, MatX, Etched, Rebellions, and SambaNova all contribute to the large-round stack.
The right reading is that AI chip fundraising is a portfolio of scale-up bets. The market is concentrated because credible chip companies are expensive, not because only one company attracted all investor demand.
How much of the AI chip funding signal is driven by outliers?
As of July 2026, a large share of the AI chip funding signal is driven by outliers. Over the past 24 months, 25 of 29 disclosed deals were above $50M, and 22 deals were above $100M.
Megarounds above $50M represent 86.21% of all disclosed deals in the AI chip market. That is unusually high and shows that public funding visibility starts late in the technical validation cycle.
The total raised outside rounds above $50M is only $148.20M. That is less than a single mid-sized Series B in this dataset, so small rounds barely affect the dollar picture.
The median round size of $250M is a better indicator than the lowest visible rounds. In the AI chip market, a visible financing often means enough capital to support tape-outs, software stacks, packaging, and customer deployment.
This chart, featured in our AI chip market deck, shows how Nvidia is leading in AI chips
Is the AI chip market broad with many targets, or narrow with few fundable companies?
As of July 2026, the AI chip market is narrow in terms of fundable companies, even though the technical approaches are diverse. Over the past 24 months, only 22 unique companies produced the 29 disclosed equity rounds in the dataset.
The same names appear repeatedly. Groq raised three times, while MatX, Positron AI, Rebellions, Biren Technology, and Cerebras each raised more than once during the 24-month window.
This repeat financing pattern matters because AI chip development is capital-intensive and execution-heavy. Investors appear to keep financing teams that have already passed technical, customer, or strategic validation.
The market is not broad in the way a software market might be broad. There is no long public tail of $5M to $20M rounds, and the dataset includes zero deals below $20M.
Is the AI chip market mostly an early-stage formation market or a late-stage scaling market?
As of July 2026, the AI chip market behaves much more like a late-stage scaling market than an early-stage formation market. Over the past 24 months, late-stage, Series C and beyond, and Growth Equity rounds captured 78.94% of disclosed capital.
Early-stage rounds from Seed to Series B captured only 18.70% of capital. That is not because no new ideas exist, but because visible AI chip funding tends to appear only after expensive technical validation.
Seed and Series A together are especially small in dollar terms. Seed accounts for just one deal and 0.23% of capital, while Series A accounts for 6 deals and 3.71% of capital.
Series B is where the early-stage label starts to look misleading. Series B rounds total $1.66B, helped by Etched, MatX, Positron AI, Fractile, and EVAS Intelligence.
For more context on how capital is moving from concept to deployment in this market, see our deeper analysis of the AI chip market.
Which categories attract the most investor attention in the AI chip market?
As of July 2026, Inference Accelerators attract the most investor attention in the AI chip market by deal count. Over the past 24 months, the category produced 15 disclosed deals, equal to 51.72% of the full dataset.
Inference Accelerators also lead by total capital, with $4.03B raised and 35.85% of disclosed dollars. Groq, Positron AI, d-Matrix, Rebellions, Fractile, Neurophos, OLIX, FuriosaAI, Arago, VSORA, and HyperAccel all sit in this group.
The category’s dominance reflects a clear customer pain point. As AI usage shifts toward serving models at scale, investors are funding chips that promise lower latency, lower power use, and better economics for inference.
Still, inference is less capital-heavy per deal than GPU and server-processor categories. It has over half of deals but only about one-third of capital, which suggests many differentiated approaches are being tested.
This chart, featured in our AI chip market deck, shows annual funding in AI chip startups
Which categories attract disproportionately large checks in the AI chip market?
As of July 2026, Data Center GPUs and Server AI Processors attract the most disproportionately large checks in the AI chip market. Over the past 24 months, both categories captured much more capital share than deal share.
Data Center GPUs had only 4 deals, or 13.79% of activity, but raised $2.65B, or 23.55% of total capital. Its capital-share-to-deal-share ratio is 1.71, the highest in the dataset.
Server AI Processors show a similar pattern. The category also had 4 deals, but raised $2.55B, giving it a 1.64 capital-share-to-deal-share ratio.
These categories are funded like infrastructure platforms, not component experiments. Investors and public markets appear to assign premium capital intensity to Nvidia-style substitution, wafer-scale systems, and full-stack AI compute platforms.
We break down these category differences further in our market report covering AI chip funding categories.
Which geographies matter most for fundraising in the AI chip market?
As of July 2026, North America matters most for AI chip fundraising by dollars, while Asia-Pacific is the second major geography. Over the past 24 months, North America captured $7.03B and Asia-Pacific captured $3.71B.
North America represents 62.46% of disclosed capital and 51.72% of disclosed deals in the AI chip market. Its average round size is $468.64M, and its median round size is $500M.
Asia-Pacific represents 32.99% of capital and 34.48% of deals. The region is close to proportional on dollars and deal count, but its funding profile is shaped by Chinese GPU and public-market financings.
Europe is present but much smaller. It produced 4 deals and $512M, equal to 13.79% of deals but only 4.55% of capital.
Is the AI chip opportunity set broad or concentrated in one hub?
As of July 2026, the AI chip opportunity set is concentrated in two major hubs rather than one. Over the past 24 months, North America and Asia-Pacific together captured 95.45% of disclosed capital and 86.20% of disclosed deals.
North America is the private mega-round hub. Groq, Cerebras, Tenstorrent, MatX, Etched, Positron AI, d-Matrix, Neurophos, SambaNova, and Majestic Labs show the region’s depth across architectures.
Asia-Pacific is more mixed. It includes Korean inference players, Indian and Chinese AI accelerator companies, and several Chinese GPU companies using public markets or growth-equity style financings.
Europe has credible technical formation but weaker scale-financing capacity. Arago, VSORA, OLIX, and Fractile show differentiated work in photonics and inference, but no European round reached the billion-dollar level.
This chart, featured in our AI chip market deck, compares the main business model options for AI accelerator chip companies
Is the AI chip market a market of small experiments or scaled financings?
As of July 2026, the AI chip market is clearly a market of scaled financings, not small experiments. Over the past 24 months, 25 of 29 disclosed rounds were $50M or larger.
The dataset contains no disclosed rounds below $20M. It includes 4 rounds between $20M and $50M, and every other round sits above $50M.
The median deal size is $250M, while the average is $388.07M. That gap shows that the AI chip market is pulled upward by billion-dollar and near-billion-dollar financings.
This size distribution makes sense for the category. A credible AI chip company needs capital for design teams, tape-outs, packaging, software, developer tools, go-to-market, and customer deployment.
If you want to track which scaled financings matter most, explore our full market deck on AI chip startups.
Who are the investors that appear the most in AI chip fundraising?
As of July 2026, repeat investors in AI chip fundraising are mostly large financial investors, strategic semiconductor investors, and infrastructure-oriented backers. Over the past 24 months, only a limited group appears in more than one disclosed deal.
Disruptive appears across Groq’s 2025 and 2026 rounds. Atreides Management and Fidelity appear across Cerebras-related rounds, while Spark Capital appears in both MatX financings.
Qatar Investment Authority appears in d-Matrix and Positron AI. M12 appears in d-Matrix and Neurophos, while Mirae Asset appears in d-Matrix and Rebellions.
Strategic participation is also important. Samsung-linked investors appear in Tenstorrent and Rebellions, while Arm-linked strategic participation appears in Rebellions and Positron AI.
One caveat matters: round announcements usually disclose total round size, not each investor’s personal check. So repeat-investor lists show participation breadth, not exact capital committed by investor.
This chart, featured in our AI chip market deck, shows how revenue is split across customer segments in the AI chip market
INSIGHTS
The insights below come from reviewing every disclosed equity round in the AI chip market between August 2024 and July 2026. They are not row-by-row summaries. They are the reusable patterns that kept showing up across the 29-deal dataset, and they are meant to stay useful when reading any future AI chip funding announcement.
- The AI chip market is structurally late-stage, even when some companies are still young. Nearly 79% of disclosed capital went to Series C or later, Series D+, or Growth Equity rounds. This means funding totals mostly describe scale-up races, not company formation.
- The absence of sub-$20M disclosed rounds is a strong signal. Public visibility in the AI chip market begins after expensive technical validation, not at the earliest company-building stage.
- Inference is the broadest funding theme, but not the most capital-intensive per deal. It holds 51.72% of deals but only 35.85% of capital. That suggests many investors see inference as the most accessible wedge against incumbent GPU economics.
- Data Center GPUs remain premium-funded despite fewer deals. They represent only 13.79% of rounds but 23.55% of capital. Direct Nvidia-style substitution still attracts large checks when companies can credibly target training and inference infrastructure.
- Server AI Processors are financed like infrastructure platforms. Their 1.64 capital-share-to-deal-share ratio shows that wafer-scale and full-system approaches are treated differently from component startups.
- Training-focused startups are scarce in the dataset. They represent only 3 deals and 7.03% of capital. Investors appear to view inference cost as the more immediate commercial wedge.
- The top 10 deals account for 69.10% of capital. Any headline about total AI chip funding should be read as a statement about a small number of mega-funded companies.
- The market is concentrated but not single-company dependent. The largest deal is only 10.04% of total capital. That means the market is a portfolio of several large platform bets rather than one runaway financing.
- The median round size of $250M is unusually high. Credible AI chip challengers must finance silicon cycles, software stacks, packaging, and deployment at the same time. Small venture checks rarely support that full path.
- North America is the center of private scale financing. It captures 62.46% of disclosed capital and has a $500M median round size. The region is funding multiple architectures at platform scale.
- Asia-Pacific is the only geography close to proportional on capital and deal count. It holds 34.48% of deals and 32.99% of capital. That makes it a true second hub, not a peripheral market.
- Europe has technical formation but weaker scale financing. It produced 13.79% of deals but only 4.55% of capital. Without billion-dollar rounds, Europe remains visible but underweighted.
- The center of gravity shifted sharply after mid-2025. September 2025 through February 2026 produced the largest financing cluster. The surge was driven by late-stage and public-equity rounds, not broad early-stage formation.
- Repeat financing is one of the strongest signals in the AI chip market. Groq, MatX, Positron AI, Rebellions, Biren, and Cerebras all raised multiple times. Recapitalization is a better signal than one-time fundraising headlines.
- China’s AI chip funding pattern is different from the US pattern. Chinese GPU companies use public markets and growth equity more visibly, while US companies raise private mega-rounds across inference, wafer-scale, dataflow, and transformer-specific architectures.
- The market is increasingly about systems, not just chips. Groq, SambaNova, Cerebras, and Rebellions sell infrastructure, racks, cloud capacity, or full platforms. Their financing logic is closer to infrastructure scale-up than pure semiconductor design.
- Memory movement is a recurring bottleneck theme. d-Matrix, Positron AI, Majestic Labs, Fractile, OLIX, and Neurophos all attack memory, bandwidth, or energy constraints. Investors are funding architectures that reduce the cost of moving data, not just raw compute claims.
- Photonic approaches are meaningful by company count but not yet dominant by repeat large follow-ons. Arago, OLIX, and Neurophos show investor interest, but the biggest repeated financings still favor electronic, wafer-scale, or deployable system platforms.
- Strategic investors matter because AI chips need deployment channels. Arm, Samsung, Microsoft’s M12, Intel-linked participation, AMD-related talent, and sovereign capital all point to the same need: credibility beyond benchmarks.
- Large rounds should not be read as proof of Nvidia displacement. They are better understood as runway, strategic optionality, and the cost of staying in the race long enough to prove deployment.
- A useful diligence rule is to weight customer deployments, repeat financing, and strategic investor participation more heavily than benchmark claims. The AI chip market often funds companies before broad third-party validation is available.
Groq (2024 Series D), TechCrunch (MatX Series A), Tenstorrent (Series D), VSORA (inference chip round), Tech.eu (Arago), Business Wire (Positron AI Series A), FuriosaAI (Series C bridge), Groq (2025 round), Rebellions (Series C), Cerebras (Series G), d-Matrix (Series C), Pandaily (Moore Threads IPO), MarketWatch (MetaX IPO), Economic Times (Biren IPO), Neurophos (Series A), Business Wire (Positron AI Series B), SambaNova (2026 financing), TechCrunch (MatX Series B), Rebellions (pre-IPO round), Groq (2026 growth round)
Related blog posts
- All funding deals in the AI chip market
- Which companies have raised the most funding in the AI chip market?
- Which companies are the most valued in the AI chip market?
