All the fundraising deals in the AI infrastructure market (from Q1 2025 to Q1 2026)
Download our beautiful pitch about the AI infrastructure market
In our AI infrastructure market deck, you will find everything you need to understand the market
Between Q1 2025 and Q1 2026, startups in the AI infrastructure market raised over $10.4 billion across 34 verified funding rounds.
The largest deals went to companies building GPU clouds, AI-optimized data centers, and next-generation AI chips, with Crusoe, Cerebras, and Nscale each closing rounds above $1 billion.
Interconnect and networking startups also broke out as a major investment theme, with Ayar Labs and Kandou AI raising $500 million and $225 million respectively in Q1 2026 alone.
And if you want to better understand this new industry, you can download our pitch covering the AI infrastructure market.
Insights
- The top 3 deals in any given quarter accounted for 57% to 94% of all AI infrastructure funding, showing how a handful of mega-rounds shape the entire market.
- Nvidia appeared as an investor in 6 of the 34 AI infrastructure deals tracked, making it the most active strategic backer across compute, cloud, and networking startups.
- Q3 2025 was the peak quarter, with over $4 billion raised in AI infrastructure, largely driven by billion-dollar rounds from Cerebras and Nscale.
- Not a single AI infrastructure deal in this dataset fell below $6.8 million, suggesting that even early-stage companies in this space need significant capital to compete.
- AI cloud and cluster platforms captured $5.3 billion across 13 deals, more than AI compute chips and networking combined, reinforcing that raw capacity is still the scarcest resource.
- Seven distinct networking and interconnect startups raised over $1 billion collectively, a sign that data movement is becoming as important as data processing in AI infrastructure.
- Sovereign and regional AI infrastructure attracted major funding in Europe (Nscale, Ori, NexGen Cloud), Australia (Firmus), and South Korea (Rebellions), proving this is no longer a U.S.-only story.
- Q1 2026 was the most diversified quarter: no single company took more than 20% of total AI infrastructure funding, compared to Q4 2025 where Crusoe alone represented 65%.
- At least five startups (Groq, Cerebras, Rebellions, Etched, Positron) are building custom AI chips to challenge Nvidia, and together they raised over $3 billion in this period.
- Analog and in-memory chip startups like EnCharge AI, d-Matrix, and Mythic collectively raised $500 million, signaling investor appetite for radically different approaches to AI compute.
In our AI infrastructure market deck, we will give you useful market maps and grids
Summary table of the funding deals in the AI infrastructure market (last 5 quarters)
We define the AI infrastructure market as the technologies and services required to run AI training and inference reliably at scale.
We include AI compute (accelerators and servers), AI-optimized cloud/cluster platforms, and the networking and storage required to move and serve model data and outputs.
We exclude end-user AI applications, foundation-model API services as a "model product," and general-purpose data/analytics and MLOps tools that are not necessary to operate the compute-and-cluster layer.
You can also read our detailed analysis to understand how funding activity in the AI infrastructure market has evolved over the last few years.
Also, you should know that we have a dedicated page, updated weekly, with all the latest fundraising deals in the AI infrastructure market.
| Name | What they do | Amount ($M) | Quarter | Source(s) |
|---|---|---|---|---|
| iPronics | Builds programmable optical networking engines for AI data centers | $20.8M | Q1 2025 | iPronics, DCD |
| Lumai | Develops optical processors to cut AI compute cost and energy use | $10M+ | Q1 2025 | Lumai |
| EnCharge AI | Makes analog in-memory AI chips for faster, cheaper inference | $100M+ | Q1 2025 | TechCrunch |
| Lambda | Runs a GPU cloud platform for AI training and inference | $480M | Q1 2025 | Lambda, DCD |
| Together AI | Operates an AI acceleration cloud for training and serving models | $305M | Q1 2025 | Together AI |
| Fluidstack | Provides instant access to large Nvidia GPU clusters for AI workloads | $200M | Q1 2025 | Signalbase |
| Ori | Offers compute-as-a-service and control software for sovereign AI clouds | $175M | Q1 2025 | Sifted |
| Nexthop AI | Builds networking hardware and software for hyperscale AI clusters | $110M | Q1 2025 | Nexthop AI, DCD |
| NexGen Cloud | Runs GPU cloud services for AI training and sovereign infrastructure | $45M | Q2 2025 | DCD |
| nEye Systems | Builds optical circuit switches to speed up AI data center networks | $58M | Q2 2025 | Business Wire |
| VSORA | Designs ultra-high-performance AI inference chips for data centers | $46M | Q2 2025 | VSORA |
| TensorWave | Runs an AMD-powered cloud for AI training and inference | $100M | Q2 2025 | TechCrunch |
| Baseten | Provides infrastructure for serving AI models in production | $150M | Q3 2025 | Baseten |
| Scintil Photonics | Builds integrated photonic components for AI data center optical links | $58M | Q3 2025 | Scintil |
| Firmus | Develops renewable-powered AI data centers in Australia | $330M | Q3 2025 | Firmus |
| Groq | Makes low-latency AI inference chips and runs its own inference cloud | $750M | Q3 2025 | TechCrunch |
| Upscale AI | Builds high-performance open-standard networking for AI GPU clusters | $100M+ | Q3 2025 | Upscale AI |
| Modular | Builds a unified software layer so AI runs across any chip | $250M | Q3 2025 | Modular |
| Nscale | Runs an AI hyperscaler with GPU cloud and AI data centers in Europe | $1,100M | Q3 2025 | Nscale, DCD |
| Rebellions | Builds AI inference accelerators and rack-scale systems from South Korea | $250M | Q3 2025 | Rebellions |
| Cerebras | Makes wafer-scale AI processors for large-scale training and inference | $1,100M | Q3 2025 | Cerebras, TechCrunch |
| Peak:AIO | Builds AI-first storage infrastructure to keep GPU clusters fed with data | $6.8M | Q4 2025 | DCD |
| Crusoe | Builds vertically integrated AI factories, data centers, and cloud | $1,375M | Q4 2025 | Crusoe |
| d-Matrix | Builds digital in-memory compute chips for data center AI inference | $275M | Q4 2025 | d-Matrix |
| Firmus | Expands renewable-powered AI data center campuses across Australia | $327M | Q4 2025 | Firmus |
| Mythic | Develops analog AI chips with dramatically lower energy consumption | $125M | Q4 2025 | Mythic |
| Fluidstack | Expands large GPU cluster capacity for AI training and inference | $450M | Q1 2026 | Sahm Capital |
| Etched | Designs transformer-specific AI chips to outperform general GPUs | $500M | Q1 2026 | DCD |
| Positron AI | Builds energy-efficient AI inference hardware to lower power costs | $230M | Q1 2026 | TechCrunch |
| Ayar Labs | Builds co-packaged optical interconnects for AI scale-up systems | $500M | Q1 2026 | Ayar Labs |
| Kandou AI | Develops high-speed copper connectivity for AI cluster interconnects | $225M | Q1 2026 | Kandou AI |
| Gimlet Labs | Runs a multi-chip serverless inference cloud across mixed hardware | $80M | Q1 2026 | TechCrunch |
| Rebellions | Scales AI inference chip production globally ahead of IPO | $400M | Q1 2026 | Yahoo Finance |
| Starcloud | Builds orbital AI data centers to run compute in space | $170M | Q1 2026 | Barron's |
In our AI infrastructure market deck, we identify pain points entrepreneurs should prioritize
How has funding activity in the AI infrastructure market changed over time?
Q3 2025 was the most active quarter for AI infrastructure funding, with over $4 billion raised across 9 deals, largely because Cerebras and Nscale each closed $1.1 billion rounds in the same month.
Q2 2025 was the quietest quarter, with just $249 million raised across 4 deals, as no single AI infrastructure startup closed a round above $100 million.
Total AI infrastructure funding in Q1 2026 reached $2.6 billion, which is a 21% increase compared to Q4 2025 ($2.1 billion) and an 82% jump compared to Q1 2025 ($1.4 billion) one year earlier.
If you strip out the top 1-2 mega-deals from each quarter, the underlying activity in the AI infrastructure market looks surprisingly steady: the remaining rounds typically add up to $400 million-$700 million per quarter, suggesting broad-based demand for compute, networking, and cloud startups beyond just the headline names.
| Quarter | Number of deals | Total raised ($) | Comment |
|---|---|---|---|
| Q1 2025 | 8 | $1,401M | Lambda ($480M) and Together AI ($305M) led a balanced quarter across AI cloud and networking startups. |
| Q2 2025 | 4 | $249M | The quietest quarter, with TensorWave ($100M) as the largest round and no mega-deals above $150M. |
| Q3 2025 | 9 | $4,088M | A record quarter for AI infrastructure funding, driven by two $1.1B rounds from Cerebras and Nscale. |
| Q4 2025 | 5 | $2,109M | Crusoe dominated with a $1.375B raise, accounting for about 65% of the quarter's total funding. |
| Q1 2026 | 8 | $2,555M | The most diversified quarter, with funding spread across AI chips (Etched, Rebellions), interconnect (Ayar Labs, Kandou AI), and cloud. |
| All quarters | 34 | $10,402M | Over $10.4 billion raised across 34 verified AI infrastructure deals from Q1 2025 to Q1 2026. |
In our AI infrastructure market deck, we identify repeatable patterns you can use if you’re building in this market
Which startups in the AI infrastructure market raised the largest rounds over the last months?
These startups raised the most recently in the AI infrastructure market:
- Crusoe raised $1.375 billion in Q4 2025 to build vertically integrated AI factories and gigawatt-scale data centers, making it the largest single round in this dataset.
- Cerebras raised $1.1 billion in Q3 2025 to scale its wafer-scale AI processors, one of the most ambitious alternatives to Nvidia-based training hardware.
- Nscale raised $1.1 billion in Q3 2025 to build out AI-focused data centers and GPU cloud capacity, starting from Europe and expanding globally.
- Groq raised $750 million in Q3 2025 as demand for its low-latency inference chips surged, reaching a $6.9 billion valuation.
- Etched raised $500 million in Q1 2026 to develop transformer-specific AI chips designed to outperform general-purpose GPUs on inference.
- Ayar Labs raised $500 million in Q1 2026 to ramp volume production of co-packaged optical interconnects, backed by Nvidia and AMD.
- Lambda raised $480 million in Q1 2025 to expand its GPU cloud platform with backing from Nvidia, ARK Invest, and In-Q-Tel.
- Fluidstack raised $450 million in Q1 2026 to finance continued GPU cluster expansion after an earlier $200M round in 2025.
- Rebellions raised $400 million in Q1 2026 as a pre-IPO round to fund mass production and global rollout of its AI inference chips from South Korea.
- Firmus raised $330 million in Q3 2025 to build a sovereign green AI factory in Tasmania, Australia, with Nvidia as an investor.
And, yes, we do cover most of them in our our beautiful pitch about the AI infrastructure market.
You may also want to check our ranking of the most funded startups in the AI infrastructure market as well as our list of the most valued startups.
In our AI infrastructure market deck, we answer all the common questions from investors and entrepreneurs
Is the AI infrastructure market shifting toward smaller or bigger deals?
Across all five quarters, the average deal size in the AI infrastructure market was about $306 million, reflecting the capital-intensive nature of building compute, cloud, and networking systems at scale.
Quarter by quarter, the average AI infrastructure deal ranged from $62 million in Q2 2025 to $454 million in Q3 2025. The jump happened because Q3 2025 included three rounds above $750 million, while Q2 2025 had no single deal above $100 million.
Even if you exclude the top two outlier deals each quarter, the average AI infrastructure round still lands around $100 million to $150 million, which suggests this is a market where building meaningful compute capacity simply requires large checks.
| Quarter | Number of deals | Average deal size ($) | Deals below $2M | Deals above $50M |
|---|---|---|---|---|
| Q1 2025 | 8 | $175M | 0 | 6 |
| Q2 2025 | 4 | $62M | 0 | 2 |
| Q3 2025 | 9 | $454M | 0 | 9 |
| Q4 2025 | 5 | $422M | 0 | 4 |
| Q1 2026 | 8 | $319M | 0 | 8 |
| All quarters | 34 | $306M | 0 | 29 |
In our AI infrastructure market deck, we help you understand how the market is structured
How concentrated was funding activity in the AI infrastructure market?
AI infrastructure funding was highly concentrated throughout this period: in every single quarter, the top 3 deals accounted for at least 57% of total capital raised. In Q4 2025, Crusoe alone captured 65% of all AI infrastructure funding that quarter, making it the most top-heavy period in the dataset.
By contrast, Q1 2026 was the most balanced quarter for AI infrastructure investment. The top deal (Etched at $500 million) represented only about 20% of the quarter's total, suggesting that investors spread their bets across compute, interconnect, and cloud startups rather than concentrating on one company.
| Quarter | Number of deals | % by Top 1 | % by Top 3 | % by Top 10 |
|---|---|---|---|---|
| Q1 2025 | 8 | 34.3% | 70.3% | 100% |
| Q2 2025 | 4 | 40.2% | 81.9% | 100% |
| Q3 2025 | 9 | 26.9% | 72.2% | 100% |
| Q4 2025 | 5 | 65.2% | 93.8% | 100% |
| Q1 2026 | 8 | 19.6% | 56.8% | 100% |
| All quarters | 34 | 13.2% | 34.4% | 82.0% |
In our AI infrastructure market deck, we have designed useful charts to give you full market clarity
Which categories in the AI infrastructure market received the most funding?
AI cloud and cluster platforms captured $5.3 billion across 13 deals, representing about 51% of all AI infrastructure funding in this period. This makes sense: building and operating GPU clouds, AI data centers, and compute-as-a-service platforms requires massive upfront capital for hardware and facilities.
AI compute (chips and accelerators) came second with $3.8 billion across 11 deals. This category reflects the intense race to build alternatives to Nvidia, with startups like Groq, Cerebras, Rebellions, and Etched each raising hundreds of millions to bring custom silicon to market.
Networking and interconnect startups raised $1.1 billion across 7 deals, which may sound smaller but represents a breakthrough. Until recently, AI networking was treated as plumbing. Now investors see data movement inside AI clusters as a bottleneck worth funding at scale.
| Category | Number of deals | Total raised ($) | Startups and amount |
|---|---|---|---|
| AI cloud / cluster | 13 | $5,287M | Lambda ($480M), Together AI ($305M), Fluidstack ($200M + $450M), Ori ($175M), NexGen Cloud ($45M), TensorWave ($100M), Baseten ($150M), Nscale ($1,100M), Firmus ($330M + $327M), Crusoe ($1,375M), Gimlet Labs ($80M), Starcloud ($170M) |
| AI compute | 11 | $3,786M | Lumai ($10M+), EnCharge AI ($100M+), VSORA ($46M), Groq ($750M), Cerebras ($1,100M), Rebellions ($250M + $400M), d-Matrix ($275M), Mythic ($125M), Etched ($500M), Positron AI ($230M) |
| Networking / interconnect | 7 | $1,072M | iPronics ($20.8M), Nexthop AI ($110M), nEye Systems ($58M), Upscale AI ($100M+), Scintil Photonics ($58M), Ayar Labs ($500M), Kandou AI ($225M) |
In our AI infrastructure market deck, we cover the latest tech updates shaping the market
Who are the biggest investors in the AI infrastructure market?
Nvidia is the most active investor in AI infrastructure startups, appearing in 6 of the 34 deals tracked. Nvidia backed companies across the entire stack, from GPU cloud providers like Lambda and Nscale to interconnect startups like Ayar Labs and Scintil Photonics. This breadth shows Nvidia is not just selling chips but actively shaping which AI infrastructure companies scale up.
Maverick Silicon participated in 3 AI infrastructure rounds (TensorWave, Upscale AI, Kandou AI), all focused on compute and networking. Maverick Silicon is building a portfolio across the hardware layers that connect and power AI clusters.
Prosperity7, the venture arm of Saudi Aramco, also backed 3 AI infrastructure deals (Together AI, TensorWave, Gimlet Labs). Prosperity7's involvement signals strong sovereign and strategic interest from the Middle East in securing access to AI compute capacity.
CapitalG, Alphabet's independent growth fund, invested in 2 AI infrastructure deals (nEye Systems and Baseten), focusing on the inference and networking layers of the AI infrastructure stack.
SoftBank Group backed 2 deals in the AI infrastructure market (Mythic and Kandou AI), placing bets on both alternative AI chips and high-speed connectivity for AI clusters.
Neuberger Berman participated in 2 major AI infrastructure rounds (Groq and Ayar Labs), backing both a leading inference chip maker and a top optical interconnect company.
Arm invested in 2 AI infrastructure deals (Rebellions and Positron AI), supporting chip startups that are building inference hardware on architectures that could complement Arm's own ecosystem.
Fidelity backed 2 of the largest AI infrastructure rounds in the dataset (Cerebras at $1.1B and Nscale at $1.1B), showing that major asset managers see AI compute and cloud as a generational investment opportunity.
General Catalyst participated in 2 AI infrastructure deals (Together AI and Modular), investing in both the AI cloud layer and the compute software abstraction layer.
Qualcomm Ventures backed 2 AI infrastructure startups (Upscale AI and TensorWave), diversifying into AI cluster networking and AMD-based cloud compute.
Disclaimer: this investor list may be incomplete; we focus on publicly disclosed lead and prominent recurring investors, so some frequent minority participants may be underrepresented. "Total funded" does not represent the amount personally invested by an individual investor. Instead, it refers to the aggregate amount raised across all fundraising rounds in which the investor participated.
| Investor | Number of deals | Total funded ($) | Startups |
|---|---|---|---|
| Nvidia | 6 | $2,618M | Lambda, nEye Systems, Scintil Photonics, Nscale, Firmus (Q3 2025), Ayar Labs |
| Maverick Silicon | 3 | $425M | TensorWave, Upscale AI, Kandou AI |
| Prosperity7 | 3 | $485M | Together AI, TensorWave, Gimlet Labs |
| CapitalG | 2 | $208M | nEye Systems, Baseten |
| SoftBank Group | 2 | $350M | Mythic, Kandou AI |
| Neuberger Berman | 2 | $1,250M | Groq, Ayar Labs |
| Arm | 2 | $480M | Rebellions (Q3 2025), Positron AI |
| Fidelity | 2 | $2,200M | Cerebras, Nscale |
| General Catalyst | 2 | $555M | Together AI, Modular |
| Qualcomm Ventures | 2 | $200M | Upscale AI, TensorWave |
In our AI infrastructure market deck, we track adoption trends and shifts in consumer behavior
Related blog posts
- All funding deals in AI infrastructure
- Which startups have raised the most funding in AI infrastructure?
