
Our CEO Chris Walker explores the photonic milestones that have transformed the compute conversation


Accelerating the future: Photonic compute in transit

Our CEO Chris Walker explores the photonic milestones that have transformed the compute conversation


“We’re effectively adding processing to where there is none right now” - our CPO @NNOPT recently spoke to @IEEESpectrum about the advances of photonic compute for AI ⬇️ eu1.hubs.ly/H0zZr5m0

AI workloads create and transform large volumes of intermediate data — data that must move between memory and processing elements and, increasingly, across distributed computing resources. The cost of moving data is becoming as much of an architectural concern as the cost of computing on it. What we need are smarter systems. Most computing infrastructure follows a familiar pattern: move data to where computation takes place, perform the operation, then move the result onwards. Faster memory, specialised accelerators and higher-bandwidth interconnects have made that process considerably more efficient, but the underlying relationship remains largely unchanged. Computation and data movement are still treated as distinct, separate activities. This raises a different architectural question: what if some computation could happen while the data was already moving? Imon Kundu & Flavio Bergamaschi for @PIC_magazine ⬇️ eu1.hubs.ly/H0zW_LN0

"Having investors on our cap table who truly understand the strength and complexities of the Northern innovation ecosystem, while also recognising the global commercial potential of the technologies being developed here, makes a significant difference." “That said, the challenges facing deeptech businesses in the North are largely the same as those experienced across the UK. Companies developing advanced hardware require access to patient, long-term capital. Unlike software businesses, which can often bring products to market relatively quickly, deeptech companies typically face substantial upfront development costs, complex engineering challenges and longer commercialisation timelines. Investors need to understand those realities and be prepared to support businesses throughout that journey." Our CPO @NNOPT spoke to @ProlificNorth about the investment landscape in the North of England ⬇️ eu1.hubs.ly/H0zTk9j0

📣 We're thrilled to announce the appointment of Robert Coenen as our new Senior Director of Business Development. Rob brings more than 20 years of experience in photonics, optical communications, and quantum computing, including senior sales and BD roles at Jabil, IQM, Q-CTRL, Quantum Machines and Intel Silicon Photonics. He holds a Ph.D. in Electrical Engineering from the University of British Columbia, where he specialized in fiber-optic communications. Welcome aboard Rob!

FEC is one of the hidden heroes of AI infrastructure. It keeps high-speed optical links reliable, but decoding adds latency, power and silicon overhead. Our latest report explores whether photonic Compute-in-Transit could change that equation in scale-out networks. Download the report ⬇️ eu1.hubs.ly/H0zMmh70

Adding compute into the optical path delivers more useful work for every transmitted bit for ultimate system efficiency in a heterogeneous AI architecture. This approach is substantiated in the Digital Infrastructure 2035 report, with @techUK predicting that the data centres of 2035 will combine classical silicon, neuromorphic chips, quantum processors and optical interconnects and processors while an intelligent orchestration layer directs each workload to the compute best suited to it. Read more ➡️ eu1.hubs.ly/H0zFTD_0

We're thrilled to be finalists in this year's Yorkshire Tech Awards 🎉 We've been shortlisted in the Tech Innovation of the Year category for our work developing photonic Compute-in-Transit for efficient AI infrastructure. Congratulations to all other finalists, we'll see you at the ceremony on 16th December 🍾 #YTAwards

GPUs have transformed artificial intelligence because they are exceptionally good at highly parallel computation, but the performance of an AI system depends on far more than the arithmetic capability of the processor: information must also travel between memory, accelerators, servers and racks. This means that as clusters expand, what happens between the processors is becoming as critical to performance as what happens on them. This is one reason optical technologies are being rapidly adopted throughout AI networking infrastructure - photonics provides the bandwidth necessary to move enormous quantities of information efficiently. At Optalysys we have asked ourselves whether optical infrastructure can contribute more than bandwidth alone. When data is already travelling as light, why wait for an inefficient conversion back to digital and pass through a separate processor before any useful operation can take place? eu1.hubs.ly/H0zxLQt0

"Across the ecosystem, architectures that don’t just deliver more compute, but rethink where that compute happens are gaining traction. Innovators and designers at the forefront of AI are now exploring whether we can turn the network itself into active compute infrastructure. From the vantage point of memory, processing-in-memory is targeting decode heavy premium workloads to reduce data movement in pursuit of speed, energy and cost gains. While memory pooling, sharing and dynamic allocation make capacity more available. That capacity can follow workloads and be more effective in serving larger models and supporting longer contexts. Disaggregated memory can improve utilization and enable larger models or contexts, but physically separating memory from the accelerator introduces a transport path with different latency and bandwidth characteristics from local HBM. If memory can become part of compute, why not the data path?" - our CEO Chris Walker in his latest post over on our blog: eu1.hubs.ly/H0ztLQZ0

Conventionally, transport and compute are largely treated as separate functions: data is moved, received and reconstructed before the useful operation happens. With photonic Compute-in-Transit, selected operations can become part of the movement event itself. This doesn’t mean replacing conventional XPUs with photonics systems. While photonics is naturally suited to particular classes of linear mathematical operations, electronics remains ideal for general-purpose control, nonlinear processing and many other workloads GPUs perform exceptionally well. What heterogenous architectures offer us is the opportunity to explore better division of labor. Read out latest post from @NNOPT eu1.hubs.ly/H0zqKkB0

The future of AI infrastructure won’t be defined by a single processor, memory or computing technology. As power, memory and data movement constraints grow we’re seeing advances in combining XPUs, specialist accelerators, advanced memory, optical interconnects and emerging compute technologies into heterogenous architectures. Heterogeneity allows each component in the system to be used where it delivers the greatest system-level advantage. In this article for @techUK, Optalysys CPO and Co-Founder @NNOPT argues that this shift creates a significant opportunity for the UK to turn its strengths in photonics, semiconductor design and advanced computing into strategically vital positions within the global AI supply chain – but doing so will require collaboration across hyperscalers, hardware developers, networking companies, deep-tech innovators and government alongside faster commercialisation of the UK’s research strengths. Read the full article: eu1.hubs.ly/H0zgRcK0

We've touched down at the Optica + Advanced Photonics Coalition PECC Summit in Sunnyvale! We're looking forward to connecting with leading technical minds from across the photonics ecosystem 💡 If you'd like to experience a brand new live demo of our photonic Compute-in-Transit technology that delivers greater energy efficiency for AI, swing by our stand (T9). You can also pre-book your meeting here bookings.cloud.microsoft/boo…

Across the ecosystem, architectures that don’t just deliver more compute, but rethink where that compute happens are gaining traction. Innovators and designers at the forefront of AI are now exploring whether we can turn the network itself into active compute infrastructure. In his latest article, our CEO Chris Walker asks "if memory can become part of compute, why not the data path?" eu1.hubs.ly/H0z8rck0 Heading to the PECC Summit and interested in finding out more? Book some time in with the team to be taken through a live demo of our photonic Compute-in-Transit technology ➡️ bookings.cloud.microsoft/boo…

New exciting opportunities 🇮🇳 🏢@HexawareGlobal - 240 new jobs in Leeds. ⚡ @AssetCool taking its innovative tech global. 🤖 @Optalysys growing its AI business. 🔌 @wilsonpoweruk Solutions expanding in 🇮🇳. These partnerships mean more jobs & opportunities for local people.

Next week we'll be exhibiting at the @OpticaWorldwide & @AdvancePhotonic PECC Summit 🚀 Visit the team on Stand T9 to see and interact with a live demo of our photonic Compute-in-Transit technology and talk to us about: - Photonic compute within the data path for AI inference - Distributed AI and collective operations - How compute can be integrated into optical links - Opportunities to collaborate across chips, networking and optical infrastructure Book some time in with the team ➡️ bookings.cloud.microsoft/boo…