By Efosa Udinmwen

Copyright techradar

Skip to main content

Tech Radar Pro

Tech Radar Gaming

Close main menu

the business technology experts

België (Nederlands)

Deutschland

North America

US (English)

Australasia

New Zealand

View Profile

Search TechRadar

Expert Insights

Website builders

Web hosting

Best web hosting
Best office chairs
Best website builder
Best antivirus
Expert Insights

Don’t miss these

Nvidia x Intel press conference: as it happened

After Sandisk, D-Matrix is proposing an intriguing alternative to the big HBM AI puzzle with 10x better performance with 10x better energy efficiency

Nvidia’s biggest customers are lining up to take it down using ASICs and Broadcom could be the winner of that battle

Nvidia gives OpenAI $100 billion to spend mostly on Nvidia hardware and Nvidia share price goes up by $220 billion – why am I not surprised?

Intel will build custom x86 CPUs for Nvidia’s AI infrastructure as world’s largest company invests $5 billion in beleaguered tech firm – and don’t discount a data center x86 APU

Samsung’s biggest rival and Sandisk have come together to standardize High Bandwidth Flash memory – which could mean better AI performance for everyone

Could this be the next big step forward for AI? Huawei’s open source move will make it easier than ever to connect together, well, pretty much everything

This is probably the most important graph of 2025 so far and shows how Nvidia’s fortunes are changing for the worse

The next stage in AI power? XConn set to reveal end-to-end PCIe Gen 6 offering higher bandwidth than ever

Google, AMD and Intel catching up on Nvidia? Survey shows almost a third of AI teams now use non-Nvidia hardware

Nvidia AI sales to reach almost $400 billion by 2028, research claims – but then things will get a bit tricky for the world’s largest company

Mediatek’s crucial role in building the Nvidia GB10 highlighted in new presentation – so could Jensen Huang mull a potential $73 billion acquisition?

Acer joins Nvidia GB10 superchip fan club with $3,999 Veriton AI mini workstation PC – and yes, you can link two to get even more power

AI > Crypto – Bitcoin mining spinoff gets $700 million investment from Nvidia to build hyperscale AI infrastructure using, you’ve guessed it, thousands of Blackwell GPUs

So who are the mystery trio? Over half of Nvidia’s data center revenue comes from three unnamed customers

‘Enfabrica has the coolest technology’: Nvidia spent nearly $1 billion on a chip maker to secure its future on the same day it gave Intel $5 billion – and here’s why this is actually a more important investment

Efosa Udinmwen

26 September 2025

Enfabrica’s ACF-S bridges PCIe and high-speed networking with minimal latency

When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.

(Image credit: ServeTheHome)

Nvidia’s acquisition brings Enfabrica engineers directly into its AI ecosystem
EMFASYS chassis pools up to 18TB of memory for GPU clusters
Elastic memory fabric frees HBM for time-sensitive AI tasks efficiently

Nvidia’s decision to spend more than $900 million on Enfabrica was something of a surprise, especially as it came alongside a separate $5 billion investment in Intel.

According to ServeTheHome, “Enfabrica has the coolest technology,” likely because of its unique approach to solving one of AI’s largest scaling problems: tying tens of thousands of computing chips together so they can operate as a single system without wasting resources.
This deal suggests Nvidia believes solving interconnect bottlenecks is just as critical as securing chip production capacity.

You may like

After Sandisk, D-Matrix is proposing an intriguing alternative to the big HBM AI puzzle with 10x better performance with 10x better energy efficiency

Nvidia’s biggest customers are lining up to take it down using ASICs and Broadcom could be the winner of that battle

Nvidia gives OpenAI $100 billion to spend mostly on Nvidia hardware and Nvidia share price goes up by $220 billion – why am I not surprised?

A unique approach to data fabrics
Enfabrica’s Accelerated Compute Fabric Switch (ACF-S) architecture was built with PCIe lanes on one side and high-speed networking on the other.

Its ACF-S “Millennium” device is a 3.2Tbps network chip with 128 PCIe lanes that can connect GPUs, NICs, and other devices while maintaining flexibility.
The company’s design allows data to move between ports or across the chip with minimal latency, bridging Ethernet and PCIe/CXL technologies.
For AI clusters, this means higher use and fewer idle GPUs waiting for data, which translates into better return on investment for costly hardware.

Are you a pro? Subscribe to our newsletter
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Contact me with news and offers from other Future brandsReceive email from us on behalf of our trusted partners or sponsorsBy submitting your information you agree to the Terms & Conditions and Privacy Policy and are aged 16 or over.
Another piece of Enfabrica’s offering is its EMFASYS chassis, which uses CXL controllers to pool up to 18TB of memory for GPU clusters.
This elastic memory fabric allows GPUs to offload data from their limited HBM memory into shared storage across the network.
By freeing up HBM for time-critical tasks, operators can reduce token processing costs.

You may like

After Sandisk, D-Matrix is proposing an intriguing alternative to the big HBM AI puzzle with 10x better performance with 10x better energy efficiency

Nvidia’s biggest customers are lining up to take it down using ASICs and Broadcom could be the winner of that battle

Nvidia gives OpenAI $100 billion to spend mostly on Nvidia hardware and Nvidia share price goes up by $220 billion – why am I not surprised?

Enfabrica said reductions could reach up to 50% and allow inference workloads to scale without overbuilding local memory capacity.
For large language models and other AI workloads, such capabilities could become essential.
The ACF-S chip also offers high-radix multipath redundancy. Instead of a few massive 800Gbps links, operators can use 32 smaller 100Gbps connections.
If a switch fails, only about 3% of bandwidth is lost, rather than a large portion of the network going offline.
This approach could improve cluster reliability at scale, but it also increases complexity in network design.
The deal brings Enfabrica’s engineering team, including CEO Rochan Sankar, directly into Nvidia, rather than leaving such innovation to rivals like AMD or Broadcom.
While Nvidia’s Intel stake ensures manufacturing capacity, this acquisition directly addresses scaling limits in AI data centers.
You might also like

Take a look at the best VPN with antivirus available right now
These are the best password managers we’ve found so far
‘NVLink is the key’: Analysts ponder on Intel + Nvidia and what it means for TSMC, AMD and others

Efosa Udinmwen

Freelance Journalist

Efosa has been writing about technology for over 7 years, initially driven by curiosity but now fueled by a strong passion for the field. He holds both a Master’s and a PhD in sciences, which provided him with a solid foundation in analytical thinking. Efosa developed a keen interest in technology policy, specifically exploring the intersection of privacy, security, and politics. His research delves into how technological advancements influence regulatory frameworks and societal norms, particularly concerning data protection and cybersecurity. Upon joining TechRadar Pro, in addition to privacy and technology policy, he is also focused on B2B security products. Efosa can be contacted at this email: udinmwenefosa@gmail.com

You must confirm your public display name before commenting

Please logout and then login again, you will then be prompted to enter your display name.

After Sandisk, D-Matrix is proposing an intriguing alternative to the big HBM AI puzzle with 10x better performance with 10x better energy efficiency

Nvidia’s biggest customers are lining up to take it down using ASICs and Broadcom could be the winner of that battle

Nvidia gives OpenAI $100 billion to spend mostly on Nvidia hardware and Nvidia share price goes up by $220 billion – why am I not surprised?

Intel will build custom x86 CPUs for Nvidia’s AI infrastructure as world’s largest company invests $5 billion in beleaguered tech firm – and don’t discount a data center x86 APU

Samsung’s biggest rival and Sandisk have come together to standardize High Bandwidth Flash memory – which could mean better AI performance for everyone

Could this be the next big step forward for AI? Huawei’s open source move will make it easier than ever to connect together, well, pretty much everything

Latest in Pro

Watch out – hackers are using AI to make phishing emails even more convincing

Nearly 150,000 patient records exposed in major healthcare data breach – here’s what we know

This devious malware has jumped from Meta over to Google Ads and YouTube to spread – here’s how to stay safe

Microsoft flags dangerous XCSSET macOS malware targeting developers – so be on your guard

I helped create Amazon’s Alexa – here’s why I think businesses should embrace humanity’s analogous imperfection to succeed

US Government tells agencies to patch Cisco firewalls immediately, or face attack

Latest in News

Sticking with Windows 10 next month? Here’s what you need to know about your choices and risks you might be taking

Quordle hints and answers for Saturday, September 27 (game #1342)

NYT Strands hints and answers for Saturday, September 27 (game #573)

The UK wants mandatory digital ID – but over one million Brits are demanding to scrap the plan over privacy concerns

NYT Connections hints and answers for Saturday, September 27 (game #839)

How to watch Ballerina online from anywhere

LATEST ARTICLES

Sticking with Windows 10 next month? Here’s what you need to know about your choices and risks you might be taking

Grumpy about Windows 10 support ending? Try this Microsoft-centered crossword to soothe and distract you

Apple Maps on iOS 26 could finally turn my head away from Google Maps

Snap up one of Ninja’s most unique and highly-rated air fryers for a record-low price

This 1TB external SSD has a 1.8-inch color touchscreen, encryption, supports Apple Find My tracking and… two fidget toys

TechRadar is part of Future US Inc, an international media group and leading digital publisher. Visit our corporate site.

Contact Future’s experts

Terms and conditions

Privacy policy

Cookies policy

Advertise with us

Web notifications

Accessibility Statement

Future US, Inc. Full 7th Floor, 130 West 42nd Street,

Please login or signup to comment

Please wait…