Atum Works

Cover Block

PUBLIC

Attribute	Value
Company Name	Atum Works
Tagline	Gigascale 3D nanomanufacturing for semiconductors
Headquarters	Mountain View, CA, USA
Founded	2024
Stage	Seed
Business Model	B2B
Industry	Deeptech
Technology	Hardware
Geography	North America
Growth Profile	Venture Scale
Founding Team	Co-Founders (3+)
Funding Label	Undisclosed

Links

PUBLIC

• Website: https://atum.works

Data Accuracy: GREEN -- Confirmed by Y Combinator company profile.

Company Overview

PUBLIC

Atum Works incorporated in 2024, a deeptech startup based in Mountain View, California, founded by three engineers with backgrounds at Caltech and NASA [Y Combinator, Spring 2025]. The company's formation was driven by a shared vision to apply advanced 3D printing principles to semiconductor manufacturing, an ambition recognized with NASA awards for Most Visionary Concept and Best Product Development prior to the startup's launch [Y Combinator, Spring 2025].

Key milestones have unfolded rapidly. The team built its first 3D nano-printer prototype and established a manufacturing facility in Mountain View shortly after founding [Y Combinator, Spring 2025]. In Spring 2025, Atum Works was accepted into the Y Combinator accelerator program, a move that provided initial capital and structured mentorship [Y Combinator, Spring 2025]. The company also reported securing letters of intent for co-development, including one with NVIDIA, and plans to sell its first products within the 2025 calendar year [Y Combinator, Spring 2025].

Data Accuracy: YELLOW -- Key operational claims (prototype, LOIs, sales timeline) are sourced solely from the company's Y Combinator profile and lack independent verification. Foundational details (incorporation date, location, team origin) are consistent across multiple public profiles [Crunchbase].

Product and Technology

MIXED

The company's core proposition is a hardware system for semiconductor manufacturing, described as a "3D nano-printer" intended to fabricate multi-material, three-dimensional structures directly onto wafers [Y Combinator, Spring 2025]. This positions it as a potential alternative or complement to the extreme ultraviolet (EUV) lithography systems that define the current state of the art, with a claimed cost advantage of one-tenth that of traditional 2D lithography processes [Y Combinator, Spring 2025]. The target resolution is sub-100 nanometers, a scale necessary for advanced chipmaking, though the specific materials and throughput capabilities are not detailed in public sources.

Public information suggests the technology is at a functional prototype stage. The company reports having built its first 3D nano-printer and established a manufacturing facility or lab in Mountain View, California [Y Combinator, Spring 2025]. A key early validation signal is a letter of intent for co-development with NVIDIA, indicating initial engagement from a major player in the semiconductor ecosystem, though the scope and commercial terms of this arrangement are [PRIVATE] [Y Combinator, Spring 2025]. The stated goal is to begin selling first products within the 2025 calendar year.

• Technical differentiation. The system's proposed ability to create "arbitrary 3D metal structures" suggests a departure from the planar, layer-by-layer patterning of lithography, potentially enabling novel chip architectures like monolithic 3D integration [Y Combinator, Spring 2025].
• Development wedge. The initial application appears focused on printing "future devices for advanced chips," which may include specialized interconnects, sensors, or other components not easily made with existing methods [Y Combinator, Spring 2025].

Data Accuracy: YELLOW -- Core claims sourced from the company's Y Combinator profile; the NVIDIA LOI and prototype status are unverified by independent reporting.

Market Research

PUBLIC

Atum Works is entering a semiconductor manufacturing market where the physical limits of traditional 2D lithography are increasingly seen as a primary bottleneck for continued performance gains, particularly for AI hardware. The company's thesis, as stated in its launch materials, is that a new scaling law will be required, shifting from planar transistor density to three-dimensional integration [Y Combinator, Spring 2025]. This creates a potential opening for a novel manufacturing approach, though the commercial landscape for such a fundamental process shift is nascent and difficult to size with precision.

No third-party market research reports specifically sizing the addressable market for gigascale 3D nanomanufacturing equipment were identified. Analysts can anchor against the broader semiconductor fabrication equipment (fab equipment) market, which is a multi-hundred-billion-dollar industry. For context, the global market for semiconductor manufacturing equipment was valued at approximately $109 billion in 2024, according to industry association SEMI [SEMI, 2024]. Within this, the lithography segment, dominated by companies like ASML, represents a critical and high-value portion. Atum's claim of enabling 3D structures at one-tenth the cost of 2D lithography, if validated, would target a portion of this existing capital expenditure.

Demand drivers are clear and well-documented across the sector. The insatiable compute requirements for training and inference of large AI models are pushing chip architectures toward greater complexity and specialization, which in turn strains existing manufacturing paradigms. Industry roadmaps, including those from the IEEE International Roadmap for Devices and Systems (IRDS), explicitly call for advancements in 3D integration, heterogeneous packaging, and new interconnect technologies as essential paths forward [IEEE IRDS, 2023]. This provides a technical tailwind for any technology promising a more economical path to 3D structures at the nanoscale.

Key adjacent and substitute markets include advanced packaging and through-silicon via (TSV) technologies, which are currently the industry's primary methods for achieving 3D-like integration. Companies like Taiwan Semiconductor Manufacturing Company (TSMC) with its 3D Fabric technology and Intel with its Foveros Direct are making significant investments in this area. Atum's proposed technology appears to aim earlier in the manufacturing flow, at the transistor and interconnect level within a single die, rather than at the package level. Regulatory and macro forces are dominated by global supply chain security concerns and national industrial policies, such as the U.S. CHIPS and Science Act, which are funneling tens of billions of dollars into domestic semiconductor manufacturing and R&D. This policy environment could lower barriers to adoption for a novel, U.S.-based equipment vendor, though it does not reduce the immense technical risk.

Given the absence of a directly applicable TAM, the following table presents analogous market sizing from established industry sources to provide a frame of reference for the segment Atum aims to disrupt.

Market Segment	2024 Size (Estimated)	Source	Notes
Semiconductor Manufacturing Equipment	$109 billion	SEMI	Total global sales for wafer fab equipment, packaging, and testing.
Wafer Fab Equipment (WFE)	~$90 billion	Gartner, 2024	Includes lithography, etch, deposition, and inspection tools.
Advanced Packaging Equipment	$7.8 billion	Yole Group, 2024	Forecast for equipment used in 2.5D/3D packaging and fan-out.

The analyst takeaway is that while the total addressable market for semiconductor equipment is vast, Atum Works is targeting a hypothetical sub-segment that does not yet exist in a commercialized form. Its success is contingent not on capturing share from an existing market, but on creating a new one by proving its technology offers a superior price-performance paradigm for 3D integration. The demand environment is favorable, but the go-to-market challenge is of the highest order.

Data Accuracy: YELLOW -- Market sizing is based on analogous, well-established industry reports. The specific application for Atum's technology is not sized by third parties.

Competitive Landscape

MIXED Atum Works enters a competitive arena not by challenging the current leaders in 2D lithography, but by proposing a new, adjacent manufacturing paradigm for which there are few direct peers [Y Combinator, Spring 2025].

Given the absence of named, direct competitors in the structured sources, the analysis proceeds without a formal comparison table. The competitive map must be drawn from first principles of the semiconductor equipment and advanced packaging sectors.

The competitive landscape for 3D nanomanufacturing is currently fragmented across several distinct segments. Atum Works does not compete with ASML or Nikon in extreme ultraviolet (EUV) lithography for leading-edge 2D transistor patterning, a multi-billion dollar duopoly. Instead, its initial wedge appears aimed at the 3D integration and advanced packaging space, where incumbents like Applied Materials and Lam Research provide tools for deposition and etching to build stacked structures [PUBLIC]. A more direct, albeit still nascent, competitive set includes startups exploring alternative 3D fabrication methods, such as those using directed self-assembly or novel imprint lithography, though none were surfaced by name in this research. Adjacent substitutes also exist in the form of outsourced semiconductor assembly and test (OSAT) providers, like ASE Group and Amkor Technology, which offer 3D packaging services using established, if less revolutionary, techniques.

The company's claimed defensible edge today rests almost entirely on its founding team's technical credentials and early validation. The team's background from Caltech and NASA, coupled with awards from those institutions, provides a talent moat for deep physics and systems engineering [Y Combinator, Spring 2025]. The reported letter of intent for co-development with NVIDIA, if executed, would be a significant channel and validation advantage, potentially locking in a flagship partner early [Y Combinator, Spring 2025]. However, this edge is perishable. It depends on translating academic insight into a reliable, production-worthy tool, a process where well-capitalized incumbents could rapidly replicate concepts once proven. The lack of disclosed patents or proprietary process details in public materials leaves the durability of this technical lead an open question.

Atum Works is most exposed to the immense scale and customer relationships of established semiconductor capital equipment vendors. A company like Applied Materials, with its decades of process integration knowledge and global service footprint, could decide to internalize a similar 3D printing approach and use its existing sales channel to dominate the segment. Furthermore, the startup is vulnerable to alternative 3D integration technologies that may achieve scale and yield faster, such as hybrid bonding, which is already in high-volume manufacturing for some applications. Atum's technology must not only work but must demonstrate a clear cost-per-structure and throughput advantage over these evolving, incumbent-driven methods.

The most plausible 18-month scenario hinges on the execution of its first product sales and the NVIDIA partnership. The winner in this near-term frame is likely to be the entity that successfully places a beta tool in a customer's fab for process development. If Atum Works can deliver its first 3D nano-printer to a partner like NVIDIA and generate validated electrical data from 3D test structures, it will have secured a beachhead. The loser in this scenario would be any other academic spin-out or small startup pursuing a similar 3D printing thesis but failing to secure a top-tier design partner, thereby stalling in the "lab curiosity" phase without a path to semiconductor qualification.

Data Accuracy: YELLOW -- Landscape analysis is inferred from industry structure; specific competitor claims are not publicly corroborated.

Opportunity

PUBLIC

If Atum Works can deliver on its core technical promise, the prize is a foundational role in the next scaling law for semiconductors, a multi-trillion-dollar market where manufacturing breakthroughs command premium valuations.

The headline opportunity is to become the de facto standard for 3D nanoscale additive manufacturing, a category-defining hardware platform analogous to ASML's position in lithography. The company's stated goal of building a "3D ASML" [Y Combinator, Spring 2025] frames this ambition directly. Evidence that this outcome is reachable, not merely aspirational, includes the team's NASA awards for visionary concept and product development [Y Combinator, Spring 2025] and their reported acquisition of a co-development LOI with NVIDIA [Y Combinator, Spring 2025]. An early engagement with a tier-one chip designer suggests the core value proposition,10x cost reduction for submicron 3D structures,is being taken seriously by potential anchor customers, providing a critical wedge into the ecosystem.

Multiple, distinct paths to scale exist from this initial wedge. The following scenarios outline plausible routes to massive adoption.

Scenario	What happens	Catalyst	Why it's plausible
Foundry Partnership	Atum's printer is adopted as a specialty tool within a major semiconductor foundry's advanced packaging or interposer line.	A successful pilot project with a partner like NVIDIA leads to a multi-unit purchase order.	The company has already built its first printer and set up a manufacturing facility [Y Combinator, Spring 2025], demonstrating a path to producing sellable tools. A co-development LOI provides a direct conduit to a potential first major customer.
New Device Enabler	The technology unlocks a novel chip architecture (e.g., monolithic 3D integration) that becomes essential for next-generation AI hardware, creating a captive market.	A leading AI chip designer publicly prototypes a revolutionary 3D chip using Atum's process.	The founders' recognition in the Forbes 30 Under 30 Science list for "nanoscale 3D-printer for microchips" [Forbes, 2025] signals expert validation of the technical premise. The team's Caltech and NASA engineering background aligns with the deep research required for such a breakthrough.

Compounding for Atum Works would manifest as a classic hardware ecosystem flywheel. An initial design win with a flagship customer generates proprietary process knowledge and device libraries. This intellectual property makes the printer more valuable for subsequent customers designing for that specific process, creating a data and design moat. Each new device fabricated on the platform adds to a library of proven, high-performance structures, reducing the integration risk for future adopters. The cited LOI with NVIDIA [Y Combinator, Spring 2025] represents the first potential turn of this flywheel, where co-development could yield process IP that becomes a barrier to entry for later competitors.

The size of the win, should a Foundry Partnership scenario fully play out, can be contextualized by looking at the valuation of companies that own critical semiconductor manufacturing equipment. ASML, the monopoly provider of extreme ultraviolet (EUV) lithography systems, carries a market capitalization exceeding $300 billion. While Atum Works is at the earliest possible stage, a successful capture of a new, essential manufacturing niche could support a valuation on the order of dedicated equipment firms like Lam Research or Applied Materials, which trade in the $50-$100 billion range. This is a scenario-based outcome, not a forecast, but it illustrates the premium the market places on companies that define a critical layer in the semiconductor stack.

Data Accuracy: YELLOW -- Core opportunity claims (LOI, cost reduction, product plans) are sourced solely from the company's Y Combinator profile. Team accolades are corroborated by Forbes [2025]. The size-of-win comparables are based on public market data for established peers, not on Atum-specific projections.

Sources

PUBLIC

1. [Y Combinator, Spring 2025] Atum Works: Gigascale Nanomanufacturing | https://www.ycombinator.com/companies/atum-works

2. [Crunchbase] Atum Works - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/atum-works

3. [SEMI, 2024] SEMI Reports 2024 Global Semiconductor Equipment Sales of $109 Billion | https://www.semi.org/en/news-media-press/semi-press-releases/2025-global-semiconductor-equipment-sales-forecast

4. [IEEE IRDS, 2023] IEEE International Roadmap for Devices and Systems | https://irds.ieee.org/editions/2023

5. [Gartner, 2024] Gartner Forecasts Worldwide Semiconductor Manufacturing Equipment Revenue to Decline 2.8% in 2024 | https://www.gartner.com/en/newsroom/press-releases/2024-07-10-gartner-forecasts-worldwide-semiconductor-manufacturing-equipment-revenue-to-decline-2-percent-in-2024

6. [Yole Group, 2024] Advanced Packaging Equipment Market 2024 | https://www.yolegroup.com/product/report/advanced-packaging-equipment-market-2024

7. [Forbes, 2025] 30 Under 30 Science 2026: New Discoveries From The Cosmos To The Nanoscale | https://www.forbes.com/sites/alexknapp/2025/12/02/30-under-30-science-2026-new-discoveries-from-the-stars-to-the-nanoscale