Chip-to-chip connections: which startup is ahead?
Everything we know about the semiconductor industry, in one place. Get our report whether you build or invest.
Summary
Ayar Labs is ahead in chip-to-chip connections, and its lead over Lightmatter widened in the autumn of 2026. Ayar showed a working rack, while Lightmatter still targets volume production for 2028.
The market is less a single race than a stack of related connection problems. Eliyan works mainly inside packages, Kandou extends copper across boards, and Ayar and Lightmatter push optics closer to processors, switches and rack-scale systems. Copper boards, in-package links and optics solve different problems, so we set out what counts and what does not in our semiconductor market report.
Ayar leads because its work forms a believable route from optical chiplet to finished AI infrastructure. In September 2026 it raised another $150 million, bringing its 2026 funding to $650 million. In October 2026, Wiwynn ran a live AI rack built with Ayar's TeraPHY engines and SuperNova lasers at the OCP Global Summit.
Lightmatter owns the stronger headline performance. Passage reaches higher bandwidth per engine and per fiber. But it has named no hyperscaler running Passage in production, and it points to 2028 and 2029 for large co-packaged optics deployments.
Kandou remains the strongest alternative to an immediate optical transition. Its 260 Gbps board-level result suggests copper can stay competitive across a large middle zone where cost, familiarity and ordinary PCB manufacturing still matter.
The most mature product story and the deepest shipping history belong to different companies. Ayar has the strongest visible pipeline for a new optical platform, while Kandou has already shipped more than 20 million chips using earlier connectivity technology.
Eliyan appears to have produced the most validated engineering output per dollar raised. Its working 3-nanometer silicon and flexible packaging approach make it the clear independent leader in electrical die-to-die links. In July 2026 it raised $145 million at a $1 billion valuation, and in October it began sampling optical chiplets.
Avicena is the most credible architectural wildcard. Its microLED approach trades a few extremely fast lanes for hundreds of slower parallel channels, which could reduce link power and signal-processing complexity if packaging and reliability hold up.
Funding concentration is unusually high. Ayar and Lightmatter account for roughly two-thirds of the seven companies' announced capital, and adding Kandou takes the top three to about 84%. Press reports put Ayar's value above $5 billion after a secondary share sale in September 2026, against Lightmatter's last published $4.4 billion.
No startup has really won yet. The decisive evidence is still missing: named high-volume deployments, quarterly shipment numbers, paid contract values, full system cost and field-reliability data. Both optical leaders now point to 2028 for real volume.
The ranking is Ayar Labs first, Lightmatter second and Kandou AI third. Ayar has the best commercial path and the most visible system proof, Lightmatter the best technical upside, and Kandou the strongest case for delaying the point at which copper must give way to optics.
Written by the New Market Pitch teamWe track new markets so founders and investors can move faster. About us
In our report covering the semiconductor market, we turn what is vague and complex into simple, useful visuals. Market maps like this one show you at a glance where every startup stands.
Which chip-to-chip connection startups should we compare?
The chip-to-chip connection race currently has seven serious independent startups, led by Ayar Labs and Lightmatter, although several companies compete at different points inside the system.
We include companies whose main product moves data between chiplets, processors, memory, packages, boards or nearby AI systems. That gives us Ayar Labs, Lightmatter, Kandou AI, Eliyan, Avicena, Xscape Photonics and Hyperlume.
The scope needs some care because "chip-to-chip connections" covers several distances. Eliyan mainly connects dies inside one package. Kandou stretches electrical connections across packages and circuit boards. Ayar and Lightmatter bring optical links close to processors and switches. Avicena and Hyperlume use microLEDs for short optical links, while Xscape supplies multi-wavelength lasers that can power several optical architectures.
These companies can complement each other as well as compete. An accelerator could eventually use Eliyan inside the package, Kandou across the board and Ayar or Lightmatter between trays or racks.
We exclude ordinary networking startups whose main products are switches, Ethernet systems or network software. We also exclude public semiconductor groups such as Nvidia, Broadcom and Marvell, even though they may ultimately ship far more interconnect hardware than the startups covered here.
Celestial AI is the most important exclusion. Marvell completed the acquisition on February 2, 2026 after announcing approximately $3.25 billion in upfront consideration. Marvell's later acquisition accounting recorded $3.5 billion in total purchase consideration. Celestial AI would otherwise belong near the top of this comparison, but it no longer operates as an independent startup.
The seven companies below have raised about $2.84 billion altogether. Ayar Labs and Lightmatter alone account for roughly two-thirds of that capital. Adding Kandou takes the top three to about 84%.
| Startup | Main focus | Announced cumulative funding |
|---|---|---|
| Ayar Labs | Optical I/O chiplets and external laser systems | About $1.02 billion |
| Lightmatter | Photonic interconnects, optical engines and photonic interposers | $850 million |
| Kandou AI | High-speed electrical links across chips, boards and memory systems | Approximately $505 million |
| Eliyan | Electrical die-to-die and chip-to-chip interface technology | About $245 million |
| Avicena | Parallel microLED optical links for processors and memory | $120 million |
| Xscape Photonics | Multi-wavelength laser sources for optical interconnects | $81 million |
| Hyperlume | Low-power microLED optical connections | $12.5 million |
Is Ayar Labs now pulling away from Lightmatter?
Yes, a little. Ayar Labs leads Lightmatter and the gap widened in 2026, although one named production contract could still reverse the order.
The market has a top two, followed by Kandou as the main electrical challenger. Both optical leaders built broad manufacturing ecosystems and moved beyond lab demos. Ayar now has more capital: about $1.02 billion raised against Lightmatter's $850 million. Ayar holding more capital than Lightmatter is one view of where the money goes, which we follow in our report on semiconductors.
Ayar's pieces fit together. TeraPHY offers up to 8 terabits per second per optical engine. Designs with Alchip and GUC place several engines around accelerators, taking package connectivity beyond 100 terabits per second. Wiwynn's investment was disclosed in September 2026, and in October Wiwynn ran a live co-packaged optics rack with GUC and Ayar at the OCP Global Summit.
Ayar has also kept its product focused. TeraPHY handles optical data movement beside an accelerator or switch. SuperNova provides the light source. UCIe connects the optical engine to the surrounding silicon. Customers can add these pieces without redesigning the whole processor around one proprietary platform.
Lightmatter has the stronger headline specifications. Its Passage portfolio includes 32-terabit and 64-terabit optical engines, 1.6 terabits per second through a single fiber and photonic interposers that can sit beneath large compute packages.
That broader architecture could ultimately support more bandwidth, but it also creates more integration work. Customers must qualify the photonic silicon, electronic interfaces, packaging, fibers, lasers, cooling and repair process together.
Lightmatter answered at the same event with a 204.8-terabit switch tray built with Flex and a 102.4-terabit compute tray. Those are validation platforms, not customer deployments. Ayar offers the shorter path from optical engine to finished AI system. Lightmatter could still overtake it if a named hyperscaler puts Passage into a production cluster.
On an informal leadership index, we would place Ayar at 100 and Lightmatter around 90. Kandou sits closer to 60 because its newest AI product has only recently reached working silicon.
Our research about semiconductors gets updated several times a month, so you always get the latest view of where things stand.
Which chip-to-chip startup has the most mature product and strongest customer traction?
Ayar Labs has the most mature new optical product and the strongest visible customer pipeline, while Kandou AI has the deepest history of shipping connectivity silicon.
Ayar describes TeraPHY as production-ready and has tested it under rapid temperature changes while maintaining its target error performance. The engine uses established CMOS and silicon-photonics processes, connects through UCIe and fits packaging methods already used for advanced processors and switches.
Its partnerships cover most of the route from chiplet to rack. Alchip and GUC create accelerator and custom ASIC designs. Wiwynn contributes board, server and rack engineering. Jabil has worked on external laser infrastructure. Nvidia, AMD, Intel, MediaTek and Alchip have invested in Ayar or worked closely with it.
The sequence matters more than any single announcement. Ayar demonstrated optical links, aligned with UCIe, passed reliability testing, created accelerator reference designs and showed a live rack. Ayar is reported to aim for qualification in the second half of 2027 and mass-market products in early 2028.
Lightmatter comes next. Passage products are sampling, evaluation systems are available, and GUC plans to bring co-packaged optical products based on the platform to hyperscale customers. GlobalFoundries, ASE, Amkor, Synopsys and Cadence support its manufacturing and design ecosystem.
Lightmatter has not named a cloud company running a production cluster built around Passage. It says two major players are designing with its M1000 platform, but that is company-stated and unnamed. Evaluation systems show serious interest, not paid adoption at scale.
Kandou says more than 20 million chips using its connectivity technology have shipped. Most came from earlier USB, PCIe, retimer and consumer-connectivity products rather than Tigerwing, so the figure proves manufacturing experience rather than current AI traction.
Eliyan has working advanced-node silicon. Avicena began shipping 1-terabit evaluation kits in August 2026. Xscape sells a commercial laser module. No independent startup has disclosed enough shipment, revenue or field-reliability data to claim mass adoption.
Which chip-to-chip startup has the best performance?
Lightmatter leads optical bandwidth, Kandou AI has the strongest long-distance copper result, and Avicena has reported the lowest transmitter-energy figure.
Lightmatter has sampled a Passage chiplet carrying 1.6 terabits per second through one fiber by placing 16 wavelengths on that fiber. Its L200 family reaches 32 or 64 terabits per second per engine, while larger configurations can provide more than 200 terabits per second around one processor package.
Ayar's individual TeraPHY engine reaches 8 terabits per second. Ayar scales by placing several engines beside one accelerator, creating package designs above 100 terabits per second.
In June 2026, Kandou demonstrated a 260-gigabit-per-second electrical connection over more than 200 millimeters of ordinary printed circuit board. Reported energy stayed below one picojoule per bit, with a bit error rate below 10-19.
Eliyan has delivered 64 gigabits per second per bump on 3-nanometer silicon. Its technology targets shorter connections inside packages, where hundreds of closely spaced bumps create several terabits per second of total bandwidth. Its NuLight optical chiplets, sampling since October 2026, span 1.6 to 12.8 terabits per second.
Avicena uses hundreds of relatively slow microLED channels rather than a few extremely fast serial lanes. It has reported transmitter energy around 80 femtojoules per bit under specified test conditions. Its newest kit carries 1 terabit per second over 335 channels running at up to 3 gigabits per second each.
The figures measure different parts of the connection, so direct comparison has limits. A transmitter-only energy number will naturally look better than a figure covering drivers, receivers, lasers, fibers and control circuitry.
| Startup | Strongest disclosed result | Where it looks strongest |
|---|---|---|
| Lightmatter | 1.6 Tbps per fiber and up to 64 Tbps per optical engine | Optical bandwidth and fiber density |
| Ayar Labs | 8 Tbps per engine and more than 100 Tbps per accelerator design | Modular package-level scaling |
| Kandou AI | 260 Gbps over more than 200 mm at below 1 pJ/bit | High-speed copper across circuit boards |
| Eliyan | 64 Gbps per bump on 3 nm silicon | Dense electrical links inside packages |
| Avicena | 1 Tbps kit over 335 lanes and about 80 fJ/bit at the transmitter | Low-power parallel optical links |
| Xscape Photonics | Eight wavelengths and more than 1 watt of optical output | Shared laser infrastructure |
In our market report about semiconductors, we look at what the winners do differently, so you can learn from the best instead of guessing.
Is Lightmatter's 2028 volume date too late to catch Ayar Labs?
Not yet too late, because Ayar also targets 2028 for volume. But Lightmatter now needs a named production customer more than another bandwidth record.
Passage attacks a real physical constraint. Processors traditionally send data through connections placed around their edges. As chips become larger and need more bandwidth, that limited perimeter becomes crowded. Lightmatter can place a photonic layer beneath the processor or stack an electronic interface directly on photonic silicon. Crowded chip edges separate technology that works from technology still on paper, a split we sort in our market report about semiconductors.
Lightmatter has turned to the practical problems around that architecture. It introduced detachable fiber arrays, a liquid-cooled laser system for dense racks and smaller Passage engines for near-package use. In August 2026 it launched an open silicon-photonics initiative at OCP with 19 companies. In October it showed switch and compute trays with Flex.
Its partnerships also look increasingly industrial. GUC supports custom silicon, GlobalFoundries provides silicon-photonics manufacturing, and ASE and Amkor provide advanced packaging. Cadence and Synopsys help chip designers integrate the required interfaces.
The timeline is the weak point. Lightmatter's CEO has described near-package optical shipments in 2027, large co-packaged deployments in 2028 and 2029, and a production-ready Passage M1000 by late 2028. A named hyperscaler running Passage in production would count for more than any peak-capacity gain.
Can Ayar Labs reach volume production by 2028?
Probably, as Ayar Labs has the strongest scaling plan among optical chip-to-chip startups. But it has not published the yields or shipment volumes that would prove mass manufacturing.
TeraPHY uses standard semiconductor manufacturing and packaging flows. The optical engine sits beside a processor or switch and communicates through an electrical UCIe interface. SuperNova keeps the light source outside the expensive compute package.
Separating the laser helps with heat, replacement and product upgrades. Ayar can improve the light source without forcing customers to redesign every processor package.
The supply chain now covers more than fabrication. Alchip and GUC handle accelerator integration. Jabil contributes optical and laser expertise. Detachable fiber connections simplify assembly and servicing. Wiwynn brings the system engineering needed to move from chip-level designs to rack equipment.
Ayar raised $500 million in March 2026 and $150 million more in September, aimed at high-volume TeraPHY manufacturing. It also opened a design center in Bengaluru. That is how a company moves from engineering validation toward production.
The hardest numbers remain private. We do not know Ayar's monthly output, packaging yield, unit price or field-failure rate. The Wiwynn rack proves the system works, not that it can be built by the thousand. For now, Ayar looks production-ready rather than proven at hyperscale.
In our document detailing everything you need to know about semiconductor, you will always find fresh funding data, taken straight from our own verified funding database.
Could Kandou AI keep copper competitive for chip-to-chip connections?
Kandou AI could delay the move to optical chip-to-chip connections because many customers will keep choosing copper whenever it remains fast enough.
Copper is familiar, inexpensive and easy to manufacture. It uses ordinary circuit boards and avoids lasers, fibers and optical alignment. Its weakness is that electrical losses rise quickly as speed and distance increase.
Kandou uses coordinated signaling across several wires to push that limit further. In June 2026, Tigerwing silicon on a 4-nanometer process transmitted 260 gigabits per second across more than 200 millimeters of circuit board at below one picojoule per bit.
A 200-millimeter connection can link separate processor, memory or switch packages on the same board. Customers could keep components physically separate and easier to manufacture while gaining some of the benefits of a large multi-chip package.
Kandou also has more shipping experience than most optical startups, and parts of its signaling work have influenced industry standards.
Tigerwing still needs a qualified customer deployment. Copper will eventually lose at very long distances and extreme bandwidth, but Kandou can win the large middle zone across boards and nearby packages.
Is Eliyan leading electrical chiplet connections?
Eliyan leads the independent startups building electrical die-to-die connections inside advanced chip packages.
NuLink connects chiplets across silicon bridges, interposers or ordinary organic packaging. Its main appeal is the ability to support different bump sizes and substrates without forcing every customer into the most expensive packaging. Costly packaging is a pain point customers feel, and we study what they would pay to solve in our report covering the semiconductor market.
Eliyan says NuLink can deliver high bandwidth and low latency through organic substrates in suitable designs. That could preserve more of the cost advantage that chiplets are supposed to create.
The company has delivered working 3-nanometer silicon at 64 gigabits per second per bump. It has also developed versions for Samsung's 4-nanometer process and longer chip-to-chip distances.
In July 2026, Eliyan closed a $145 million Series C at a $1 billion valuation, led by Seligman Ventures. Cisco and Lumentum joined earlier backers such as AMD, Arm, Meta, Coherent, Samsung, Intel Capital and SK hynix. That group covers processors, networking, cloud infrastructure, optics and memory.
In October 2026 Eliyan also began sampling NuLight optical chiplets, which moves it toward optical territory. The missing evidence is still a production processor shipping with NuLink. Until that appears, Eliyan leads the electrical subcategory and presses Kandou for third place without taking it.
Learning from others' mistakes is the cheapest lesson there is. In our report covering the semiconductor market, we look at what broke startups before you, so you don't repeat it.
Could Avicena's microLED links beat conventional silicon photonics?
Avicena has the most credible alternative to laser-based optical chip-to-chip connections, but its microLED platform remains several steps behind Ayar and Lightmatter commercially.
Conventional silicon-photonics systems concentrate data into a small number of extremely fast lanes. Avicena instead uses hundreds of microLEDs and photodetectors connected through multicore fiber.
Each lane runs at a more modest speed, but the large number of lanes creates high total bandwidth. Its newest evaluation kit carries 1 terabit per second, and Avicena has reported more than two terabits per second per millimeter of bandwidth density.
Slower parallel lanes may need less demanding signal processing than a few ultra-fast serial channels. MicroLED manufacturing could also benefit from equipment and techniques developed for displays.
Avicena has acquired Nanosys assets and is working with TSMC on photodetectors. Its investors include SK hynix, Micron Ventures, Lam Research and Hitachi Ventures.
Customers are only beginning to test the complete system. Avicena still needs to prove long-term LED reliability, fiber alignment, packaging yield and total link power outside the laboratory.
Which chip-to-chip startup uses its funding best?
Eliyan has produced the most validated chip-to-chip technology per dollar raised, while Ayar Labs has built the strongest case for spending at a much larger scale.
Eliyan has raised about $245 million across three announced rounds, most of it in its July 2026 Series C. It has delivered working 3-nanometer silicon, several packaging options and versions covering both die-to-die and longer chip-to-chip connections. That is still less than one-third of the capital raised by either Ayar or Lightmatter.
Avicena has also used its $120 million efficiently. It developed a new microLED architecture, acquired manufacturing assets, worked with TSMC and shipped 1-terabit evaluation kits.
Kandou has raised roughly $505 million. That produced several generations of connectivity silicon, more than 20 million shipped units and the new Tigerwing architecture.
Ayar's roughly $1.02 billion supports photonic dies, electronic interfaces, laser systems, packaging, test equipment and manufacturing capacity. Lightmatter has raised $850 million for an even broader product range covering several distances and laser infrastructure.
We cannot calculate true capital efficiency without revenue, gross margin and operating-expense data. Based on public engineering output, Eliyan has achieved the most for its money.
| Startup | Funding | What the capital has produced |
|---|---|---|
| Eliyan | About $245 million | Working advanced-node silicon and several packaging options |
| Avicena | $120 million | A new microLED architecture and 1 Tbps evaluation kits |
| Kandou AI | Approximately $505 million | More than 20 million historical shipments and working Tigerwing silicon |
| Ayar Labs | About $1.02 billion | Production-ready optical chiplets and a live rack-scale system with Wiwynn |
| Lightmatter | $850 million | The widest photonic product portfolio and advanced evaluation platforms |
| Xscape Photonics | $81 million | A commercial eight-wavelength laser module |
| Hyperlume | $12.5 million | Early microLED interconnect development |
Your time is limited. In our research about semiconductors, we put numbers on each part of the market, so you can see where it makes sense to focus.
Which chip-to-chip startup is moving fastest now?
Ayar Labs has the strongest momentum, while Lightmatter keeps releasing products, Kandou AI is making the sharpest comeback and Eliyan is accelerating.
In 2026 Ayar combined $650 million of fresh capital, expansion in Taiwan and India, Nvidia ecosystem access and a live rack with Wiwynn. A secondary sale in September 2026 priced it above $5 billion, according to press reports; Ayar's September announcement gave no valuation.
Lightmatter has been more active on the product side. It sampled its high-density Passage chiplet, introduced a smaller optical engine, added a liquid-cooled laser system and showed rack trays with Flex in October 2026. Product launches and rack demos month after month are signals we track every quarter in our report on semiconductors.
Lightmatter has also broadened its strategy. Customers can now enter through near-package and co-packaged products instead of starting with the largest photonic interposer.
Kandou has changed direction more dramatically. New leadership repositioned the company around AI infrastructure, the business raised $225 million, and Tigerwing reached working silicon soon afterward. Eliyan raised $145 million in July 2026, launched a 224-gigabit NuLink-XD SerDes the same month and shipped optical chiplet samples in October.
Ayar has the most complete momentum because its activity stretches from funding to manufacturing and a working system. Lightmatter is advancing at least as quickly technically, but its customer story has not moved as far.
Could Nvidia, Broadcom or Marvell crush the chip-to-chip startups?
Large semiconductor companies could erase the lead of weaker chip-to-chip startups, but Ayar Labs and Lightmatter have become strategically useful enough to survive or attract a major buyer.
Nvidia ships Quantum-X Photonics switches in 2026. Broadcom sells Tomahawk 6 Davisson, a 102.4-terabit co-packaged optics switch. Marvell owns Celestial AI and showed 2-nanometer optical demos at ECOC in September 2026.
These companies already sell processors or networking chips in huge volumes. Their customers trust their manufacturing, software and support, and they can bundle connectivity with products buyers already need.
Marvell's purchase of Celestial AI shows one likely outcome. A strong interconnect startup may create more value inside a large semiconductor platform than as an independent company.
Ayar and Lightmatter still have room to build independent businesses because hyperscalers designing custom accelerators often want components that work across several processor and switch vendors. Open interfaces such as UCIe also make neutral suppliers easier to adopt.
Reliable optical chiplets require mixed electronic-photonic design, advanced packaging, fiber attachment, testing and years of manufacturing work. Large incumbents can copy the concept faster than they can reproduce the complete production knowledge.
In our market report about semiconductors, we always try to gather several sources and make estimates that make sense, so you can work with solid numbers.
What would prove that a chip-to-chip startup has really won?
A named, high-volume production deployment would settle the chip-to-chip startup race more convincingly than any funding round, partnership or benchmark.
The first missing number is shipment volume. We need to know how many optical engines, electrical interfaces or microLED links leave factories each quarter and how many operate in real AI systems.
The second is paid contract value. A multi-year supply agreement worth several hundred million dollars would outweigh a long list of development partnerships.
Manufacturing economics would also change the ranking. Customers need the full cost of photonic dies, electrical interfaces, lasers, fibers, connectors, packaging, testing, cooling and repair.
Reliability data will become more important as products reach the field. Buyers need predictable laser lifetimes, low failure rates and a practical way to replace damaged components.
Real workload results would be especially persuasive. A production cluster running large models would show whether projected gains in training speed and throughput survive software overhead, congestion and system failures.
Ayar has the strongest indirect evidence, now including a live rack. Lightmatter has the best technical case. Kandou has the longest shipping history. The first to turn that into a named, paid and repeatable deployment will pull clearly ahead.
Which chip-to-chip connection startups are actually ahead?
Ayar Labs is the strongest chip-to-chip connection startup overall, followed by Lightmatter and then Kandou AI. The order has not changed, but Ayar's lead is wider.
Ayar wins because it combines a mature optical engine, a standard UCIe interface, credible chip-design partners, about $1 billion of funding and a rack that already runs live with Wiwynn.
Lightmatter ranks second. Passage offers higher bandwidth density and a broader range of optical architectures, but volume is planned for 2028 and no hyperscaler is named. A large named customer could still move it into first place.
Kandou ranks third because copper remains valuable wherever it can still meet the required speed and distance. The company has shipped millions of earlier connectivity chips and proved Tigerwing in silicon in June 2026.
Eliyan holds fourth place and leads electrical links inside advanced packages, but its $1 billion valuation round and first optical samples bring it close to Kandou. Avicena ranks fifth because microLEDs could eventually provide a simpler optical architecture. Xscape ranks sixth as an important laser supplier without control of the complete link. Hyperlume remains seventh because it is earlier and less publicly validated.
We place Ayar at 100 and Lightmatter around 90. Kandou sits further back near 60 because its latest AI technology has not yet shown the same system-level adoption path.
A hyperscaler production deployment would probably put Lightmatter first. A large Tigerwing contract could pull Kandou closer to the top two. A production processor shipping with NuLink would likely lift Eliyan to third. A successful processor or memory integration could move Avicena above Eliyan.
Ayar Labs is ahead. Lightmatter remains the most dangerous challenger, while Kandou offers the strongest alternative route by extending copper rather than replacing it right away.
| Rank | Startup | Why it ranks here |
|---|---|---|
| 1 | Ayar Labs | Strongest mix of product maturity, accelerator integration, manufacturing funding and a live rack-scale system |
| 2 | Lightmatter | Best optical performance and broadest platform, with volume planned for 2028 and no named production customer |
| 3 | Kandou AI | Proven shipping experience and a compelling copper interface that reached working silicon in 2026 |
| 4 | Eliyan | Leading electrical die-to-die specialist, valued at $1 billion in July 2026 and now sampling optical chiplets |
| 5 | Avicena | Most promising alternative optical architecture, but still early in customer qualification |
| 6 | Xscape Photonics | Valuable multi-wavelength laser supplier without control of the complete interconnect platform |
| 7 | Hyperlume | Relevant microLED technology at a much earlier commercial stage |
Our methodology
This analysis compares the independent startups building chip-to-chip connections for AI systems, from electrical die-to-die links inside advanced packages to board-level copper, optical I/O chiplets, photonic interposers, microLED links and shared laser infrastructure.
We assessed the companies across six dimensions: technical performance, product maturity, customer traction, manufacturing readiness, capital efficiency and recent momentum. The final ranking is a structured judgment across those dimensions rather than a mechanical average.
We prioritized working silicon over roadmaps, validated products over laboratory concepts, system integration over general partnerships, and production activity over evaluation-stage interest. Funding was treated as an enabler, not proof of leadership by itself.
Performance figures were interpreted within the architecture and distance they measure. Die-to-die electrical links, board-level copper, optical chiplets, photonic interposers and microLED systems solve related but different problems, so their bandwidth and energy figures are not directly interchangeable.
Commercial traction was judged by how far each announcement moved a product toward deployment. A named high-volume production customer would carry the most weight, followed by qualified system integration, working evaluation hardware and development agreements.
Manufacturing partnerships were treated as stronger evidence when they covered a concrete step such as fabrication, packaging, ASIC design, fiber attachment, testing, laser supply or rack integration. Investor participation was used only as supporting context.
Funding totals are approximate because private companies do not always separate equity, debt, grants, strategic investments and round extensions in the same way. We used announced cumulative figures where available and avoided treating small differences as precise.
Celestial AI was excluded because Marvell completed its acquisition on February 2, 2026. Marvell initially announced approximately $3.25 billion in upfront consideration, while its later Form 10-Q recorded $3.5 billion in total purchase consideration under acquisition accounting.
Key Ayar Labs sources include the TeraPHY UCIe optical-chiplet announcement, the TeraPHY product page, Ayar's reliability and thermal-testing update, its rack-level work with Wiwynn, and the company's optical I/O product overview.
Key Lightmatter sources include the Passage L200 announcement, the L200 product page, the 1.6 Tbps-per-fiber demonstration, and the wider Passage portfolio.
For Kandou AI, we used the Tigerwing 260 Gbps result, the $225 million strategic financing announcement, and the Tigerwing tape-out and product-positioning update.
For Eliyan, we used the 64 Gbps-per-bump 3-nanometer result, the NuLink packaging and standards overview, and the Series B financing announcement.
Key sources include Ayar Labs on its $650 million 2026 funding, AI Weekly on the Series E extension and reported valuation, Tech Times on the Wiwynn live rack at OCP, Futurum on Lightmatter and Flex at OCP 2026, Taipei Times on Lightmatter's production timeline, Avicena on its 1 Tbps evaluation kits, Avicena's press-release archive, Eliyan on its $145 million Series C, Eliyan on its NuLight optical chiplets, Xscape Photonics' eight-wavelength laser and $37 million financing announcement, Xscape's earlier $44 million Series A announcement, Marvell's original Celestial AI acquisition announcement, Marvell's completion notice, and Marvell's Form 10-Q acquisition disclosure.
If you build or invest in the semiconductor market, you need to know which business models pay off now and which ones are fading. That's why our report covering it has a whole section on it.
Written by the New Market Pitch teamWe track new markets so founders and investors can move faster. About us
