Cornelis raises $205 million for a scale-up fabric that ships inside other vendors' racks
The Intel spinout's Active Compute Fabric targets Nvidia's NVLink but reaches customers only through platform builders such as Qualcomm, on terms neither company has described.
The Investor · Invest desk
What happened
- Cornelis Networks, the Intel spin-off based in Wayne, Pa., said Monday it had raised $205 million and that its entry into scale-up networking has the potential backing of Qualcomm.
- Cryptobriefing.com reported that IAG Capital Partners led the round, which will fund production of the CN5000 and CN6000 switches and take the company deeper into scale-up networking.
- The CN5000 switch is shipping at 400 Gbps, and the multi-protocol CN6000 at 800 Gbps is expected to become more broadly available in the fourth quarter of 2026.
Why it matters
- What it limits constraint Cornelis's scale-up income now depends on someone else shipping a rack, so the timing of that revenue belongs to platform builders such as Qualcomm.
- Who has to decide decision Enterprise buyers get a second name to price scale-up against, and the argument they will make internally is cost, availability and vendor dependency.
- What it sets up precedent An open rack gets a resale channel once Qualcomm opens a data center partner program; Spelman expects channel firms to take on rack delivery and servicing when it does.
Next story · Product · yesterday
Reviewers find Nvidia's RTX Spark well ahead on local AI and behind the top chips on CPUConfirmed4 publishers
The scale-up business Cornelis has just entered may never produce a Cornelis box. CRN reported that the move may not result in channel-ready products directly from the startup, and that the company will instead put its Active Compute Fabric architecture into the hands of firms building rack-scale platforms, Qualcomm among them [10]. Scale-out, the existing business, links servers across a large cluster; scale-up links processors inside a single rack-scale system [17].
The fabric combines programmable compute with lossless data transport and in-fabric acceleration, moving some processing off the accelerators and into the network itself [5]. It supports Ultra Ethernet and UALink, both consortium efforts aimed at Nvidia's proprietary interconnect stack [6]. The switch shipping now is the CN5000 at 400 Gbps. The CN6000 handles multiple protocols at 800 Gbps and is expected to be broadly available in Q4 2026 [11], double the throughput [21], six years after Intel spun Omni-Path out into Cornelis [7][18].
Two accounts of the Qualcomm relationship are on the record, and one claims more than the other. CRN wrote that Cornelis did not share details of what it called a "collaboration", and described the scale-up move as having Qualcomm's potential backing [22][1]. Cryptobriefing.com called it a strategic partnership that the two companies plan to showcase at the AI Infra Summit [23]. Both put Tony Pialis, general manager of Qualcomm's data center business, alongside Cornelis CEO Lisa Spelman this week [4]. Pialis said in a statement that "Giving customers more choice and flexibility across the infrastructure stack will be critical as AI moves toward rack-scale architecture" [9]. Qualcomm plans to debut its first rack-scale platform this year with its AI200 accelerator, with successor products in the years after [8].
The demand argument is a procurement one. An executive at a large U.S. systems integrator, who asked CRN not to be identified, said "If you're not married to one architecture, there are cost implications as well. There are availability implications as well, and the dependencies implication as well. So from a purely business point of view, this is great" [14]. Resellers cannot sell it yet; Qualcomm has not announced a partner program for its data center business [13]. Spelman said "I think there's a world in which some of these channel players become mechanisms for rack delivery and servicing of that enterprise" [16].
I would treat the $205 million as development money for an architecture with one publicly named prospective host. Cornelis did not say how the round splits between scale-up development, CN6000 manufacturing and customer deployments [20]. The open-alliance reading fails if Qualcomm's rack arrives with NVLink inside: Nvidia has spent more than a year promoting NVLink to rival accelerator makers as the de facto scale-up method, and that campaign has been winning their support [12][19]. What Cornelis has in production is supercomputing work, including the Lynx machine at Lawrence Livermore National Laboratory, which runs on its networking [15].
What to watch
- Whether Qualcomm's AI200 successor products, due in the years after this one, are built on Active Compute Fabric.
- Whether any rack-scale platform builder other than Qualcomm names Cornelis as its scale-up fabric.
- The first disclosed commercial terms of the Cornelis-Qualcomm collaboration.
Information, not investment advice. This story is general information about companies and markets, drawn from cited reporting. It is not a recommendation to buy, sell or hold any security; check the cited sources before acting.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence55
- Adoption32
- Hype gap+35
- Incentives72
- Confidence60
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Cornelis Networks, an Intel spin-off based in Wayne, Pa., announced on Monday that its scale-up networking move has the potential backing of Qualcomm and that it has raised $205 million in new funding from investors.
- [2]
The fresh capital will support the development of future scale-up and scale-out products as well as manufacturing and customer deployments for its new CN6000 family of scale-out products.
- [3]
The $205 million round was led by IAG Capital Partners and will fund production of the company's CN5000 and CN6000 switches and push the company deeper into the scale-up networking market.
- [4]
Pialis is expected to join Cornelis CEO Lisa Spelman for her keynote at the AI Infra Summit this week.
- [5]
The Active Compute Fabric architecture combines programmable compute capabilities with lossless data transport and in-fabric acceleration, offloading some processing tasks to the network itself rather than making the AI chips handle everything.
- [6]
Cornelis supports technologies including Ultra Ethernet and UALink, both backed by industry consortiums seeking alternatives to Nvidia's proprietary interconnect stack.
- [7]
Cornelis traces its lineage to Intel's Omni-Path Architecture, which Intel developed for supercomputing workloads before spinning it out in 2020, and the company has since pivoted from that HPC heritage toward AI networking.
- [8]
Qualcomm plans to debut its first rack-scale platform this year with its AI200 accelerator chip, with successor products to come out in the following years.
- [9]
Tony Pialis, general manager of Qualcomm's data center business, said in a statement: "Giving customers more choice and flexibility across the infrastructure stack will be critical as AI moves toward rack-scale architecture."
- [10]
Cornelis's foray into scale-up networking may not result in channel-ready products directly from the startup; instead Cornelis will use its newly revealed Active Compute Fabric architecture to give firms building rack-scale computing platforms like Qualcomm an alternative to Nvidia's proprietary NVLink interconnect.
- [11]
The CN5000 switch is already shipping with 400 Gbps throughput; the CN6000, which supports multiple protocols at 800 Gbps, is expected to become more broadly available in Q4 2026.
- [12]
Nvidia has spent more than a year promoting its proprietary NVLink interconnect technology as the de facto method for scale-up networking to competitors building rack-scale solutions with their own accelerator chips.
- [13]
Qualcomm has not yet announced a partner program for its data center business.
- [14]
An executive at a large U.S. systems integrator, who asked not to be identified, said: "If you're not married to one architecture, there are cost implications as well. There are availability implications as well, and the dependencies implication as well. So from a purely business point of view, this is great."
- [15]
The Lynx supercomputer at Lawrence Livermore National Laboratory runs on Cornelis networking, and the company has additional implementations across government and academic institutions.
- [16]
Spelman said: "I think there's a world in which some of these channel players become mechanisms for rack delivery and servicing of that enterprise."
- [17]
Cornelis's traditional scale-out networking business is focused on high-speed interconnects for servers across large clusters, while scale-up networking enables high-speed connections between processors inside rack-scale systems.
- [18]
The CN6000's expected broad availability in Q4 2026 falls six years after Intel spun the Omni-Path business out in 2020.
- [19]
Scale-up networking is critical to rack-scale AI computing and is helping Nvidia gain support from AI chip rivals.
- [20]
Neither published account states how the $205 million is allocated among the listed uses: future scale-up and scale-out product development, CN5000 and CN6000 production, and customer deployments.
- [21]
The CN6000's 800 Gbps is twice the throughput of the shipping CN5000's 400 Gbps.
- [22]
Cornelis didn't share any details about what it called a "collaboration" with Qualcomm, though leaders from both companies signaled the potential to integrate their technologies together.
- [23]
Cryptobriefing.com reported a strategic partnership with Qualcomm Technologies, and said the two companies plan to showcase the collaboration at the AI Infra Summit.
Sources
2 independent publishers whose own reporting we read for this story.
- crn.comCornelis Mounts Scale-Up Challenge To Nvidia, Teases Potential Qualcomm Tie-Up
1 article · September 14, 2026
- cryptobriefing.comCornelis raises $205M to challenge Nvidia’s dominance in AI networking
1 article · September 14, 2026
Topics and entities
Topics
- Scale-up networking
- Rack-Scale AI Systems
- Open interconnect standards
- AI Infrastructure Funding
Entities
- Lawrence Livermore National Laboratory
- Nvidia
- Cornelis Networks
- Qualcomm
- IAG Capital Partners
- AI Infra Summit
- Active Compute Fabric
- NVIDIA NVLink
- UALink
- Intel