Fractile, the London-based AI inference chip startup, announced a $720 million Series C on November 11, and its reported valuation rests on one unproven engineering bet: DRAM directly on top of logic, stacked so AI models can generate answers faster. The company has not shipped a production chip, and the UK government's Sovereign AI Fund is among the backers.

Anyone tracking AI hardware gets two things from this round: a concrete picture of how Fractile's memory design differs from GPUs and from SRAM-based rivals, and a clear list of what the company has not yet shown. Reuters, citing a person involved, reported a $7 billion valuation on Wednesday. CEO Walter Goodwin said the funding will let Fractile ship its first chips and systems in 2027.

Four Venture Firms Led Round While UK Sovereign AI Fund Took Part

Thrive Capital, Founders Fund, Lightspeed Venture Partners and Redpoint led the round. New investors include CapitalG, BlackRock, DST Global, Dragoneer, Elad Gil, Balderton, SV Angel, the British Business Bank and the Sovereign AI Fund; returning investors include Accel, Conviction, Factorial and Felicis. Founders Fund also led the $220 million Series B in May, alongside Accel and Factorial Funds.

Fractile lists offices in London, Bristol, Palo Alto and Taipei and is hiring across its stack.

Government, Financial Times and Reuters Report Different Valuations for Same Round

The UK government says the round values Fractile at over £5 billion (approximately $6.6 billion), according to TechCrunch's report on the round. The Financial Times and a person cited by Reuters put it at around $7 billion, according to the same report, which hedged that figure. Some outlets still list $6.5 billion, which Bloomberg reported in August as a pre-money figure, meaning the value before new capital is added. Adding $720 million of new capital to $6.5 billion puts it near $7.2 billion, close to Reuters' figure.

The step up is steep. Fractile raised a $220 million Series B in May, valuing it near $1 billion, led by Accel, Founders Fund and Factorial Funds, after a seed round of about $15 million in 2024. That makes the new valuation more than six times the May figure.

The size of the round also moved as it came together. Bloomberg reported in August that Fractile was in talks to raise about $600 million at a $6.5 billion pre-money valuation, with some of the money invested at lower prices. Companies House filings show about £577 million (approximately $765 million) of Series C shares allotted between September 9 and October 23, Tech Funding News reported, in three share classes priced at £801.54, £707.25 and £1,452.35 per share. Sifted reported £540 million (approximately $716 million).

The coverage reviewed does not explain the gap, but the arithmetic is suggestive. Dealroom reports that the £37.4 million C-2 class came from earlier investments converting into shares, and subtracting it from £577 million leaves about £540 million. The announced $720 million equals roughly £543 million at today's rate. Because shares sold at prices that differ by about a factor of two, a single headline valuation depends on which class sets it.

How Does Stacking DRAM on Logic Differ From HBM and SRAM?

AI inference, the stage where a trained model produces answers, is limited mostly by memory bandwidth rather than arithmetic. Each generated token requires streaming the model's weights from memory to the processor, so the chip that moves weights fastest produces tokens fastest. Dealroom's summary of an October No Priors interview says Goodwin argued each new model demands far more memory bandwidth and that data-center inference economics come down to cost per gigabyte.

Chip designers have three memory options, and each has a cost. SRAM sits on the die and is fast but small: Nvidia's Groq 3 LPU holds 500 MB at 150 TB/s, while a Rubin GPU holds 288 GB of HBM4 at 22 TB/s, so a 256-chip LPU rack holds only 128 GB in total. An SRAM cell needs four to six transistors per bit while DRAM needs one transistor and one capacitor, which is why DRAM is denser and cheaper per bit. HBM stacks up to 32 DRAM dies linked by through-silicon vias beside the processor, and HBM4 reaches up to 2 TB/s and 64 GB per stack.

Fractile's earlier design, as VKTR described it in August, put compute and memory on one die using SRAM. Goodwin said the company concluded by late 2023 that SRAM could not scale, and it pivoted to pulling high bandwidth out of low-cost DRAM in work with memory vendors. The company now places DRAM directly on top of logic, aiming for more bandwidth than HBM and more capacity than SRAM. Coverage as recent as August still describes the chip as SRAM-based, a description that conflicts with the DRAM design Fractile now describes.

Academic work on DRAM-on-logic stacks shows what is at stake. A 2025 arXiv paper on Stratum puts HBM's through-silicon via pitch at about 10 µm (0.0004 inches) against about 1 µm (0.00004 inches) for copper-to-copper hybrid bonding, which allows far more connections between memory and logic. A 2026 evaluation framework paper cites access energy of 0.66 to 0.88 picojoules per bit for 3D-stacked DRAM, 77% to 83% below HBM, a figure it takes from another source. These are estimates for the technology class, not Fractile measurements. Fractile has not said which bonding process it uses.

Thermal Limits, Yield and Packaging Remain Open Engineering Questions for Stacked DRAM

Stacking brings its own limits. In 3D chips, heat builds up inside the stack and hotspots need careful management, and in wafer-to-wafer bonding a single defective die can ruin an entire stack. Samsung's zHBM concept, which stacks DRAM directly on a processor, targets roughly four layers rather than the 12 or 16 of conventional HBM because of heat. A 2026 DeepStack modeling study found that, under 7 nm, effective bandwidth peaks below about 10 stacked DRAM layers and then declines. The coverage reviewed for this article includes no Fractile figures on layer count, bandwidth, power, thermal behavior or yield.

Manufacturing runs through Taiwan. A Fractile job listing for a Taiwan-based packaging engineer describes owning packaging yield and ramp while working with outsourced assembly-and-test providers and substrate suppliers.

Tight DRAM Supply Could Test Premise That Commodity Memory Is Cheap Capacity

Fractile's pivot depends on DRAM being the cheap, abundant substrate. The 2026 memory market does not look like that. Wikipedia's DRAM entry reports early-2026 price increases exceeding 200% for some DDR and flash memory since early 2025, and its HBM entry reports HBM crowding out commodity DRAM, with Micron citing a 3-to-1 wafer conversion ratio between HBM and DDR5. Tom's Hardware reported in August that memory makers shifted capacity toward HBM and server DRAM, starving commodity markets. Goodwin has said Fractile works with memory vendors, but whether that insulates it from market swings is not established. The company has not said who supplies its DRAM dies or on what terms, so the exposure is an open question rather than a documented problem.

Performance Claims Range From 25 Times to 100 Times Faster With No Public Benchmarks

Fractile has described its speed advantage differently over time. At its 2024 debut it said its design could run large models up to 100 times faster and 10 times cheaper than existing hardware. Computer Weekly reported the company's claim as up to 50 times faster at about 10% of the cost. In more recent investor materials, 25 times faster at one-tenth the cost is reportedly the framing. The company's May announcement targeted throughput rising from 40 tokens per second to more than 1,000, W.Media reported. The coverage reviewed includes no public benchmark data.

Goodwin has argued that fast inference matters most for AI agents, with a goal of multi-trillion-parameter models at many thousands of tokens per second, according to Dealroom's summary of the October interview.

UK Sovereign AI Fund Joined Round but Has Not Disclosed Its Stake

The UK Sovereign AI Fund was announced in April with £500 million (approximately $663 million) of public money and invests directly in British AI companies. The Telegraph reported that its checks typically run from £1 million (approximately $1.3 million) to £10 million (approximately $13.3 million) and that it entered at a slightly lower valuation than the final price, according to Dealroom's summary; the size of its stake has not been disclosed, and Fractile and the government declined to comment. The British Business Bank also took part in the round.

The government said it wants Britain to hold a larger stake in the hardware underpinning its future economy, TechCrunch's report on the round noted. Fractile had already announced a £100 million UK expansion (approximately $133 million) in February, and Computer Weekly reported the plan covers its London and Bristol sites.

The public money goes into a hardware company with no shipped production chip. Dealroom's analysis flags that risk, along with the roughly sixfold valuation rise in months and signs that some investors are pulling back from the largest AI rounds.

Former Co-Founder Yuhang Song Left in 2024 Amid Reported Concerns About China Links

Fractile was founded in 2022 by Goodwin and co-founder Yuhang Song, who has since left. City AM reported on February 5 that Song departed in May 2024 amid concerns about his links to China, and said he studied at Beihang University in Beijing, which is on the U.S. Commerce Department's entity list. The report states there is no suggestion Song engaged in wrongdoing or that any intellectual property was transferred, and a Fractile spokesperson said he left to pursue other technical and business interests. Sifted also reported the departure in February.

Anthropic Chip Deal Rests on Reports Citing Unnamed Sources

Bloomberg reported in August that Anthropic had agreed an initial deal to buy about $250 million of Fractile's chips, with both sides intending to expand it. Deliveries are not expected until 2027, and the contract would equal roughly a third of the $720 million round. VKTR said The Information first reported the talks in May, when they were described as early-stage. TechCrunch's report on the round does not mention the deal. Tech Funding News described the order as only an initial agreement, according to unnamed sources, and found no disclosed revenue.

Disclosure: Anthropic makes Claude, the AI model that helped research this article.

Etched Shipped a Rack to Jane Street While Fractile Has Shipped No Production Chip

Fractile joins a well-funded field of inference-chip challengers, and several rivals are further along. Etched raised $700 million at a $21 billion valuation on August 18 and shipped its first rack to Jane Street in July. Cerebras went public on May 14, raising $5.5 billion at $185 per share, a fully diluted valuation of $56.4 billion at the IPO price. Nvidia licensed Groq's technology in December 2025, and Groq's founder and other engineers joined Nvidia. London-based Olix, valued at $3.3 billion, plans to deliver its first chip to customers in the second half of 2027, and TechCrunch also names Axelera among the well-funded challengers.

Goodwin Has Argued Britain Needs Domestic AI Hardware as Part of Its Compute Plan

Goodwin started Fractile in 2022 after working with early foundation models during his PhD. In an opinion piece for The Engineer, he called the government's compute aims the right level of ambition, backed a significant share of that compute being publicly held, and argued the UK needs its own hardware as an indispensable link in the chain, pointing to chip-design hubs in Cambridge, Bristol and Oxford. The government's £1.1 billion AI hardware plan (approximately $1.5 billion), announced in June, names Fractile and Olix among British hardware companies.

What Is Proven, Contested and Still Unshown After Fractile's $720M Round

Verified: Fractile announced a $720 million Series C led by four venture firms, with the UK Sovereign AI Fund and British Business Bank participating, and it describes a design that stacks DRAM directly on logic. Contested: the valuation, which the government puts above £5 billion and Reuters puts near $7 billion, and the round size, which ranges from about £540 million to £577 million depending on how shares are counted. Unshown: public benchmarks, thermal and yield data, DRAM supply terms, the size of the public stake and company confirmation of the Anthropic deal. Fractile's first chips and systems next year are due in 2027, and until then the $7 billion figure prices a design, a team and a customer report rather than shipped silicon.

Currency conversions are approximate and based on exchange rates as of November 11, 2026.

Frequently Asked Questions

What is an AI inference chip?

An AI inference chip runs a trained model to produce answers, as opposed to training the model. Because generating each token requires reading the model's weights from memory, inference speed depends heavily on memory bandwidth, which is why Fractile, Groq, Cerebras and Etched compete on memory design.

How is Fractile's stacked DRAM different from HBM?

HBM stacks DRAM dies beside the processor and connects them through an interposer, while Fractile says it places DRAM directly on top of logic to get more bandwidth than HBM and more capacity than SRAM. The open questions are heat, manufacturing yield and DRAM supply, and the coverage reviewed includes no Fractile figures for any of them.

Who invested in Fractile's Series C?

Thrive Capital, Founders Fund, Lightspeed Venture Partners and Redpoint led the round. Other investors include CapitalG, BlackRock, DST Global, Dragoneer, Balderton, the British Business Bank and the UK Sovereign AI Fund, plus returning backers such as Accel.

How much public money is in Fractile?

The Sovereign AI Fund's stake has not been disclosed. The Telegraph reported that the fund typically writes checks of £1 million (approximately $1.3 million) to £10 million (approximately $13.3 million), according to Dealroom's summary of the Telegraph report, and the British Business Bank also took part.