# MIT Technology Review reports Luminous Computing’s plan for optical AI chips

[Canonical reading](https://escapemeridian.com/readings/mit-technology-review-reports-luminous-computing-s-plan-for-d5b78f41)

MIT Technology Review reported on June 13, 2019, that startup Luminous Computing was developing an optical chip to accelerate AI computing. Unlike conventional electrical chips, the proposed design uses lasers to send light through tiny structures called waveguides; different colors can carry multiple pieces of data at once. The article says this could increase data movement while using less power, but describes the chip as a prototype, not a demonstrated product in widespread use.

Luminous cofounder and chief technology officer Mitchell Nahmias said the prototype was more energy efficient than other state-of-the-art AI chips. That is the company’s reported comparison, not evidence in the article of performance in deployed systems. The article also describes practical hurdles: optical chips require devices such as lasers and electro-optic modulators, and an MIT professor and Lightmatter advisor said managing those components could be challenging as production scales. Other startups and established chipmakers were also pursuing optical chips.

The article presents faster, lower-power AI computing as a possible benefit for uses such as autonomous vehicles and smartphone assistants. Whether Luminous could overcome production challenges, and whether its prototype’s reported efficiency would translate into useful performance in deployed systems, remained open questions.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

## Original source

MIT Technology Review · Article

Original source date: Jun 13, 2019 (2019-06-13T09:05:36Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-10-29T00:37:14.553Z. First observed: 2026-10-28T14:09:54.870Z. Retrieved: 2026-10-28T14:11:52.093Z.

- [Bill Gates just backed a chip startup that uses light to turbocharge AI](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Limits

Text as retrieved on October 28, 2026; originally published June 13, 2019. The page was edited after publication; the retrieved text may include later additions, which are not treated here as part of the original report. The efficiency comparison is attributed to Luminous’s chief technology officer; the article does not provide independent test details or deployed-system results. The article reports potential applications and obstacles, not outcomes from production use.


## Open question

How would Luminous’s prototype perform in deployed systems, and could its optical components be managed at production scale?


## Source-stated

Luminous was developing a chip that uses light rather than electrons to carry out computations for AI.

> While conventional semiconductors use electrons to help carry out the demanding mathematical calculations that power AI models, Luminous is using light instead.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

Luminous said its prototype was more energy efficient than other state-of-the-art AI chips; the article does not report a deployed-system comparison.

> Mitchell Nahmias, another cofounder of Luminous and its chief technology officer, says its current prototype is three orders of magnitude more energy efficient than other state-of-the-art AI chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

The article describes optical chips as requiring components including lasers and electro-optic modulators, and reports a concern about managing them as production scales.

> Optical chips need everything from lasers to electro-optic modulators for controlling light to make them work, which is a big reason they haven’t yet caught on widely.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

An MIT professor and Lightmatter technical advisor said Luminous might find it challenging to manage the devices needed to manipulate light as production increases.

> Dirk Englund, an MIT professor who’s also a technical advisor to Lightmatter, thinks Luminous may find it challenging to manage the various devices required to manipulate light when it starts to ramp up production of its chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Escape Meridian interpretation

Escape Meridian reads the proposed design’s data-carrying approach as a possible route to faster, lower-power AI computing, but the article does not establish results in deployed systems.

> By using different colors of light to move multiple pieces of data through waveguides at the same time, it can outstrip the data-carrying capabilities of conventional electrical chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.


## Source media

Media retrieval status: unchecked.


## Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.


## Editorial provenance

Version 1. Publication ID: 939e122a-7f74-293b-5f4c-683dd33cfde3.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T00:36:46.581Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge\_admitted\_by\_model.

Observation: ca2d92ee04e8758f38cc18a382869e60d2e8f51c348fe529a97da0d417ba0403. Source text hash: 94c4b510a0aece826b79b0c8df1df88f8e7677a6b4a71048f09f878b86c98648.


## Revision and source-check history

### Version 1 — 2026-10-29T00:37:14.553Z

MIT Technology Review reported on June 13, 2019, that startup Luminous Computing was developing an optical chip to accelerate AI computing. Unlike conventional electrical chips, the proposed design uses lasers to send light through tiny structures called waveguides; different colors can carry multiple pieces of data at once. The article says this could increase data movement while using less power, but describes the chip as a prototype, not a demonstrated product in widespread use.

Luminous cofounder and chief technology officer Mitchell Nahmias said the prototype was more energy efficient than other state-of-the-art AI chips. That is the company’s reported comparison, not evidence in the article of performance in deployed systems. The article also describes practical hurdles: optical chips require devices such as lasers and electro-optic modulators, and an MIT professor and Lightmatter advisor said managing those components could be challenging as production scales. Other startups and established chipmakers were also pursuing optical chips.

The article presents faster, lower-power AI computing as a possible benefit for uses such as autonomous vehicles and smartphone assistants. Whether Luminous could overcome production challenges, and whether its prototype’s reported efficiency would translate into useful performance in deployed systems, remained open questions.

## Original source

MIT Technology Review · Article

Original source date: Jun 13, 2019 (2019-06-13T09:05:36Z).

Date basis: Date from the publisher’s own records, through its API.

Added to Escape Meridian: 2026-10-29T00:37:14.553Z. First observed: 2026-10-28T14:09:54.870Z. Retrieved: 2026-10-28T14:11:52.093Z.

- [Bill Gates just backed a chip startup that uses light to turbocharge AI](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Limits

Text as retrieved on October 28, 2026; originally published June 13, 2019. The page was edited after publication; the retrieved text may include later additions, which are not treated here as part of the original report. The efficiency comparison is attributed to Luminous’s chief technology officer; the article does not provide independent test details or deployed-system results. The article reports potential applications and obstacles, not outcomes from production use.


## Open question

How would Luminous’s prototype perform in deployed systems, and could its optical components be managed at production scale?


## Source-stated

Luminous was developing a chip that uses light rather than electrons to carry out computations for AI.

> While conventional semiconductors use electrons to help carry out the demanding mathematical calculations that power AI models, Luminous is using light instead.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

Luminous said its prototype was more energy efficient than other state-of-the-art AI chips; the article does not report a deployed-system comparison.

> Mitchell Nahmias, another cofounder of Luminous and its chief technology officer, says its current prototype is three orders of magnitude more energy efficient than other state-of-the-art AI chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

The article describes optical chips as requiring components including lasers and electro-optic modulators, and reports a concern about managing them as production scales.

> Optical chips need everything from lasers to electro-optic modulators for controlling light to make them work, which is a big reason they haven’t yet caught on widely.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source-stated

An MIT professor and Lightmatter technical advisor said Luminous might find it challenging to manage the devices needed to manipulate light as production increases.

> Dirk Englund, an MIT professor who’s also a technical advisor to Lightmatter, thinks Luminous may find it challenging to manage the various devices required to manipulate light when it starts to ramp up production of its chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Escape Meridian interpretation

Escape Meridian reads the proposed design’s data-carrying approach as a possible route to faster, lower-power AI computing, but the article does not establish results in deployed systems.

> By using different colors of light to move multiple pieces of data through waveguides at the same time, it can outstrip the data-carrying capabilities of conventional electrical chips.

- [Supporting source](https://www.technologyreview.com/2019/06/13/867/ai-chips-uses-optical-semiconductor-machine-learning/)

## Source and contributor context

MIT Technology Review · Publisher

The publisher is identified here. No individual author has been inferred.


## Source media

Media retrieval status: unchecked.


## Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.


## Editorial provenance

Version 1. Publication ID: 939e122a-7f74-293b-5f4c-683dd33cfde3.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T00:36:46.581Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge\_admitted\_by\_model.

Observation: ca2d92ee04e8758f38cc18a382869e60d2e8f51c348fe529a97da0d417ba0403. Source text hash: 94c4b510a0aece826b79b0c8df1df88f8e7677a6b4a71048f09f878b86c98648.


### passed — 2026-10-29T00:36:46.581Z — anthropic/claude-opus-5.5

Publication: 939e122a-7f74-293b-5f4c-683dd33cfde3

attribution — pass: The reading attributes the report to MIT Technology Review. It attributes the efficiency comparison to Luminous CTO Mitchell Nahmias and the production-scaling concern to an MIT professor who advises Lightmatter (Dirk Englund), matching the source.

> Dirk Englund, an MIT professor who’s also a technical advisor to Lightmatter, thinks Luminous may find it challenging to manage the various devices required to manipulate light

summary — pass: The reading identifies the actor (startup Luminous Computing) and the development (optical AI chips). It explains waveguides and the use of multiple colors of light in plain language, and notes hurdles and competition. It omits the $9 million seed round and the 'three orders of magnitude' figure, but this does not distort the meaning. The title and question follow from the evidence.

> It uses lasers to beam light through tiny structures on its chip, known as waveguides.

date — pass: June 13, 2019 matches the UTC day of originalPublished. The evidence is a publisher page, not an archive copy, so no archive disclosure is needed. The page was edited in 2020, and its Most Popular sidebar describes later items, but the reading neither uses nor relies on them. There is no hindsight framing, and plans and claims are not presented as outcomes.

support — pass: Each claim's excerpt is an exact quote from the core article body, which is plainly part of the original piece rather than a later edit. The efficiency figure is kept attributed and the lack of deployed-system evidence is noted. The applications mentioned (autonomous vehicles, smartphone assistants) are in the source. The interpretation is labeled and hedged.

> Mitchell Nahmias, another cofounder of Luminous and its chief technology officer, says its current prototype is three orders of magnitude more energy efficient than other state-of-the-art AI chips.

media — pass: No media is selected, and no verified media candidates were supplied, so nothing could be misidentified.

## Recorded references

Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter.

No public references recorded here yet.

[Reference data (JSON)](/readings/d5b78f41ffc0df1d9c526ea60d5f7fcf67d3594cf7f4f09f4e163f64c1c25e61/references.json)

## Corrections and challenges

[Send a correction or an opposing source](https://escapemeridian.com/corrections/)

[JSON record](https://escapemeridian.com/readings/mit-technology-review-reports-luminous-computing-s-plan-for-d5b78f41.json)