"The Fastest AI Infrastructure"

Cerebras Systems is an AI infrastructure company known for pioneering wafer-scale computing technology. Founded in 2016 and headquartered in Sunnyvale, California, it has emerged as a leader in delivering high-performance AI inference and training. In September 2025, Cerebras completed a $1.1 billion Series G funding round led by Fidelity Management & Research Company and Atreides Management, bringing its post-money valuation to $8.1 billion. The company initially filed to go public in September 2024 but withdrew its IPO plans following this major private funding round, choosing to further expand its customer base and data center infrastructure ahead of a future public listing.

‍

Business Model

Cerebras Systems is an AI infrastructure company known for pioneering wafer-scale computing technology. Founded in 2016 and headquartered in Sunnyvale, California, it has emerged as a leader in delivering high-performance AI inference and training. In September 2025, Cerebras completed a $1.1 billion Series G funding round led by Fidelity Management & Research Company and Atreides Management, bringing its post-money valuation to $8.1 billion. The company initially filed to go public in September 2024 but withdrew its IPO plans following this major private funding round, choosing to further expand its customer base and data center infrastructure ahead of a future public listing.

‍

Scale and Growth

Cerebras generates revenue mainly through sales of its CS-3 wafer-scale systems, its CSoft software platform, and professional services such as data preparation, model design, training support, and optimization. It also operates a growing cloud business offering on-demand access to its systems, including its Inference Cloud launched in August 2024. Professional services typically account for 25–33% of new customer engagements. The company claims major performance advantages, including more than 10x faster training times and over 10x faster output versus GPU-based competitors, which supports its premium pricing strategy.

A notable milestone in 2025 was the launch of Sonar, a Perplexity AI search product powered by Cerebras infrastructure, delivering over 1,200 tokens per second and providing near-instant, high-quality answers for Perplexity Pro users. ‍

‍

Competitive Edge

<row>

<col>

The AI accelerator market reveals a strategic spectrum from general-purpose platforms to specialized solutions. NVIDIA maintains dominance with versatile GPUs serving diverse AI workloads, while Groq disrupts with ultra-specialized hardware achieving unprecedented LLM inference speeds. AMD and AWS challenge NVIDIA with broad capabilities, as Google TPU and Intel Gaudi balance workload flexibility with optimization. Cerebras and SambaNova target specific high-value segments with specialized architectures. This landscape highlights the tension between versatility and performance optimization in AI infrastructure.

</col>

<col>

</col>

</row>

‍

<row>

<col>

Targeted Impact

Cerebras Systems accelerates AI breakthroughs by enabling faster, more efficient deep learning to solve complex global challenges.

</col>

<col>

</col>

<col>

What Comes Next?

Cerebras Systems is rapidly expanding its AI inference cloud infrastructure across multiple continents while scaling manufacturing capacity to meet surging demand for real-time AI applications.

</col>

</row>

<row>

<col>

</col>

<col>

Who will benefit?

Cerebras Systems' innovations benefit AI researchers, supercomputing centers, pharmaceutical firms, and cloud service providers.

</col>

<col>

</col>

</row>

‍

Cerebras Systems' Timeline

‍

‍

Bull and Bear Case

<row>

<col>

</col>

<col>

</col>

</row>

<row>

<col>

Market Opportunity & Expansion

• AI infrastructure Total Addressable Market (TAM) projected to reach around $400B by 2030, with rapid adoption in inference workloads and expansion of cloud AI services
• Cerebras projects TAM growth from $131B (2024) to $453B (2027), a 51% CAGR, driven by GenAI adoption and enterprise AI integration
• Cerebras' expansion into five new US and international data centers (as of September 2025) enhances AI infrastructure capacity and broadens global market reach

</col>

<col>

Market Risks & Competitive Threats

• Direct competition from Nvidia (dominant in AI hardware), and emerging startups (Lightmatter, Tenstorrent) leveraging multi-chip packaging, threatens Cerebras’ niche defensibility
• Downturn in GenAI investment or capital market volatility could compress margins and stall infrastructure growth, as investors may reconsider aggressive AI bets
• Customer concentration risk: G42 accounted for ~83% of 2023 revenue and ~87% in H1 2024, limiting diversification and market reach

</col>

</row>

<row>

<col>

Competitive Advantages & Defensibility‍

• Wafer-Scale Engine (WSE) is the largest AI processor in production: 56x size of leading GPU, delivers compute and memory bandwidth advantages for large-scale AI model training and inference
• Protected by extensive patent portfolio on wafer-scale integration and packaging
• Strategic customer focus builds loyalty; partnerships with Meta and Perplexity AI for ultra-fast inference highlight technology lead

</col>

<col>

Operational Challenges & Execution Risks

• Cerebras’ shift toward a vertically integrated hardware-plus-cloud model increases its own capex and opex burden, pressuring margins as it scales
• Potential scale bottlenecks as company ramps manufacturing and data center operations, execution missteps could delay deployments or escalate costs
• Cerebras expanded its executive team to manage growth in AI Cloud, inference, and global operations, underscoring the execution challenges of rapid scale-up

</col>

</row>

<row>

<col>

Financial Performance & Unit Economics

• 2024 revenue: $500M, +535% YoY increase over 2023
• 2023 performance: Revenue reached $78.7M, up from $24.6M in 2022, while net loss narrowed to $127.2 M from $177.7 M
• Token-based pricing ($0.10 per M tokens for 8B models) adds scalable usage-based revenue

</col>

<col>

Financial Sustainability Concerns

• $127M net loss on $78.7M revenue (2023) shows high fixed costs and R&D load
• Margin pressure: Gross margins fell to 41.1% in H1 2024 from 50.5% in H1 2023, driven by a heavier hardware mix and G42 pricing discounts
• Financial model relies on continually improving unit economics; potential for negative surprises if cloud services scale slowly or pricing falls

</col>

</row>

<row>

<col>

Management Team & Execution Capability

• CEO Andrew Feldman & CTO Sean Lie previously founded SeaMicro, acquired by AMD in 2012
• Recent leadership hires add expertise in AI cloud (Meta), enterprise sales (MongoDB), and cybersecurity (FICO), supporting scale-up
• Strong institutional investor base signals confidence in execution and strategy

</col>

<col>

Regulatory or Structural Headwinds

• Persistent regulatory scrutiny: 2024–2025 IPO delays caused by national security review and CFIUS (foreign investment) bottlenecks
• Ongoing regulatory risk may affect future M&A, customer deals, and public listing timing, especially for hardware exports and U.S. operations
• Policy shifts could impact cross-border manufacturing, supply chain, and talent mobility

</col>

</row>

<row>

<col>

Strategic Positioning & Optionality

• Lays out path for IPO in 2026 following capital raise, with flexibility to remain private if market conditions require
• Diversifies business model: transitioning from hardware sales to cloud-based AI compute services broadens revenue capture
• Expanding U.S. manufacturing and data center footprint buffers against supply chain and geopolitical risks

</col>

<col>

Valuation & Timing Risks

• $8.1B post-money valuation (Sept 2025) reflects peak AI cycle; risks of repricing if AI investment cools or public markets correct.
• Cerebras withdrew IPO plans in October 2025, delaying liquidity and posing risk of market sentiment shifting unfavorably for next offering.

</col>

</row>

Prominent News

Cerebras CEO explains IPO withdrawal, says AI chipmaker still intends to go public, October 6, 2025 - The CEO confirmed his company pulled its S-1 filing but emphasized that the intention to list remains, citing a need to refresh financials and strategic disclosures. Click here to read more.‍‍‍

‍Cerebras Systems raises $1.1B at $8.1B valuation to accelerate AI chip race, Oct 8, 2025 - The AI chipmaker closed a Series G round, bringing in $1.1 billion and valuing the company at $8.1 billion, with the funds earmarked to fast-track development and competition in the AI hardware space. Click here to read more.‍

Cerebras Sets New AI Speed Record on MBZUAI and G42’s K2 Think at 2,000 Tokens/Second Inference Performance,  September 10, 2025 - Cerebras, in collaboration with MBZUAI and G42, announced that their K2 Think model now runs at a record-breaking 2,000 tokens per second inference speed on Cerebras’ hardware, outperforming much larger models in speed and efficiency. Click here to read more.‍

‍

Revenue

The graph below displays the revenue and revenue growth of Cerebras Systems from 2019 to 2024.

• Revenue was up 534.97% in 2024 from $78.74 million in 2023

‍

‍

Return on Investment (ROI)

The graph below displays Cerebras Systems' ROI over the past 180, 90, and 30 days, computed as of October 06, 2025.

• 180D: ↑ +5.77% since April 10, 2025‍
• 90D: ↑ +6.89% since July 09, 2025‍
• 30D: ↑ +10.75% since September 07, 2025

‍

‍

‍

Funding Round in USD

• Primary Funding Round: Cerebras Systems raised $1.1 billion in Series G on September 19, 2025, at a $8.10 billion valuation and a share price of $36.23
• Investors:  Series G saw participation from 1789 Capital, Alpha Wave Global, Altimeter, Atreides Management, Lp, Benchmark, Fidelity, Tiger Global Management, and Valor‍
• Current Implied Valuation: $7.73 billion on October 06, 2025, according to PM Insights‍
• Change Since Last Primary Round: Down 4.57% from its Series G valuation

‍

<row>

<card>

Primary Round Valuation

<span class="date">2025-09-19</span>
<span class="value">$8.10B</span>

</card>

<card>

Implied Valuation

<span class="date">2025-10-06</span>
<span class="value">$7.73B</span>

</card>

<card>

Change Since Last Round

<span class="date">2025-09-19 → 2025-10-06</span>
<span class="value red">-4.57%<span>

</card>

</row>

‍

‍

‍

Quarterly Bid:Ask Volume Ratio

PM Insights highlights the quarterly bid:ask volume ratio of Cerebras Systems where there were at least three institutionally-sized broker contributions within that period. The stacked bar chart below shows the total quarterly bid and ask volumes alongside their relative ratios.

• Highest volume: $243.25 million in Q1 2025‍
• Lowest volume: $7.50 million in‍ Q4 2024‍
• Largest bid-side activity: 40% of $175.58 million in Q2 2024
• Largest ask-side activity: 100% of $7.50 million in Q4 2024 and 100% of $30.12 million in Q3 2025

‍

‍

‍

Volatility of Composite Price

The volatility of the composite price is calculated based on price movements within a 90-day rolling window ending on each respective date. The graph below illustrates the trend in volatility of composite price for Cerebras Systems for a one year period from October 07, 2025 until October 06, 2025.

• Volatility started at 13.55% at the beginning of the observed period
• It rose sharply through early 2025, peaking at 25.45% in March
• After April, it trended downward, reaching a low of 6.01% in September 2025
• It recovered slightly toward the end of the period, ending at 11.47% in October 2025

‍

‍