Science1 distinct publisher3 min readUpdated
Holly Springs is heading to sixteen 20,000 L bioreactors and 1,400 staff by 2028. The constraint on a biologics program is shifting from finding capacity to committing to it early.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
Fujifilm Biotechnologies' Holly Springs, North Carolina, site is on course for sixteen 20,000 L mammalian bioreactors and 1,400 staff by 2028, and so far one customer has been publicly named for the second half of it [3][4][5][6]. That is the current shape of the large-molecule market: the capacity a program will need at the end of the decade is being built now, and the scarce thing is a booked slot rather than a plant.
The arithmetic is worth doing slowly. Fujifilm first committed $2 billion for eight 20,000 L mammalian bioreactors at Holly Springs, then in 2024 added $1.2 billion for eight more by 2028, taking the site to $3.2 billion end to end, with the first phase opened last year [2][3][4]. Sixteen vessels at 20,000 L is 320,000 L of installed mammalian capacity at one address [1]. Add Hillerod, Denmark, which Fujifilm Corp. bought from Biogen in 2019 for $890 million and expanded in 2024 from six to twelve 20,000 L bioreactors, and the two large-scale sites reach 28 vessels and roughly 560,000 L [7][8][9][2].
The named demand behind phase two is argenx, which according to GEN has agreed to manufacture drug substance for efgartigimod alfa-fcab, sold as intravenous Vyvgart and, with hyaluronidase-qvfc, as injectable Vyvgart Hytrulo, in generalized myasthenia gravis and chronic inflammatory demyelinating polyneuropathy [6]. Between the 2024 announcement and the 2028 completion date there are about four years [4]. A sponsor negotiating today is contracting for suites that do not exist yet, and the slots fill during that window.
Headcount is the harder lead time. Holly Springs employs about 800 people and is set to reach 1,400 when phase two is done, which is 600 hires, a 75 percent increase [5][3]. The Hillerod expansion shows where that labor goes: a 2,000 square meter quality control laboratory opened ceremonially in April and began operations in May, staffed by roughly 100 quality personnel doing viral safety testing for drug substance and product release, cell-based potency and ELISA methods, and raw material and total organic carbon cleanability studies, inside a 7,600 square meter building with its own emergency generator [10][11][12]. "Simply, we needed more QC analytical capacities to support all the increased number of batches and the increased number of testing programs," Toshihisa Iida of Fujifilm told GEN [13]. Release testing scales with batches and programs, not with liters.
Iida also drew a line under further construction. His stated focus is getting existing facilities online and running well "because we already have big contracts with big clients," and any next-phase investment comes after more spending on people and quality systems [14]. Read plainly, the 28 vessels now committed are the supply for this cycle. Holly Springs was also one of seven facilities the FDA selected for its PreCheck Pilot Program [1]. Separately, the group's Fujifilm Cellular Dynamics opened a 175,000 square foot iPSC headquarters and manufacturing site in Madison, Wisconsin, in May, including a gene editing center of excellence [15].
What to watch: whether more phase-two customers are named before 2028, since an unfilled 320,000 L is a pricing event as much as a capacity one; whether Iida's operational-excellence condition holds and no third phase is announced [14]; and whether QC and analytical throughput, not bioreactor volume, becomes the queue that actually delays batches [13].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
Fujifilm Biotechnologies' commercial-scale cell culture manufacturing facility in Holly Springs, NC, was selected among seven facilities chosen by the FDA for its PreCheck Pilot Program, designed to strengthen US pharmaceutical manufacturing capabilities.
Last year Fujifilm Biotechnologies opened the first phase of its $3.2 billion end-to-end biomanufacturing facility in Holly Springs, NC.
Fujifilm initially committed $2 billion to a Holly Springs facility consisting of eight 20,000 L mammalian cell culture bioreactors.
In 2024 Fujifilm announced plans to invest an additional $1.2 billion by building eight additional 20,000 L bioreactors at the Holly Springs site by 2028.
About 800 people are based at Holly Springs, a figure set to grow to 1,400 when the second phase is completed.
One biopharma customer has been announced for the second phase: immunology developer argenx, which has agreed to manufacture drug substance for efgartigimod alfa-fcab, marketed as intravenous Vyvgart and, combined with hyaluronidase-qvfc, as injectable Vyvgart Hytrulo, both indicated for adults with generalized myasthenia gravis or chronic inflammatory demyelinating polyneuropathy.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Specific and quantified, but single-source and company-supplied
Every figure - dollar commitments, bioreactor counts, square footage, headcount, the named argenx contract - is precise and internally consistent, and an identified executive is quoted on the record. But the whole cluster rests on one trade publication's roundup drawn from a company interview, with no independent, customer-side or regulatory documentation supplied.
Real plants running, thin disclosed demand
Adoption is visible in physical, already-completed steps: phase one open at Holly Springs, Hillerod expanded to twelve vessels with a QC lab in operation, a Madison iPSC site opened, and FDA PreCheck selection. What is missing is commercial uptake evidence - a single named phase-two customer and no utilization, booked-volume or backlog figures - so committed demand for the 2028 capacity cannot be sized.
Mildly overstated on demand, sober on execution
The capacity and capex facts are concrete, and Iida's own framing is conservative - operational excellence and quality-system investment before any next phase. The overstatement is modest and sits in the demand narrative: 'big contracts with big clients' is unnamed and unquantified while only argenx is disclosed for phase two, and the story's framing that capacity is now abundant enough to shift the constraint to early commitment goes slightly beyond what one publisher's interview evidences.
Company-sourced expansion narrative in trade press
The material originates in a vendor-facing CDMO roundup where a Fujifilm corporate vice president is interviewed about the company's own facilities; Fujifilm has a direct commercial interest in signalling available, regulator-favoured capacity to prospective biologics customers, and the outlet's format aggregates such company updates. No adversarial or customer-side voice is present in the supplied source.
Moderate: checkable physical facts, one publisher
Confidence is limited by sole-publisher coverage and reliance on company statements, but raised by the nature of the claims: bioreactor counts, building sizes, an acquisition price and a named customer program are durable, externally checkable facts, and nothing in the supplied material contradicts them. Forward elements - 2028 completion, 1,400 staff - remain company projections.
science
The self-driving lab is out. Whether AI shows up in your filing is still open.1 distinct publisher
science
A chip with human bone, lung and circulating tumour cells tests the metastasis model gap1 distinct publisher
product
Supreme Court voids "POST MILK GENERATION" for oat drinks, and the workaround with it4 distinct publishers
science
Three federal health actions in one week, and the compliance clock starts before the evidence does1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 15, 2026