Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

How would Meta deploy its networking clusters? How would that be different at Oracle, Nebius or Crusoe? Clara Ee on the SemiAnalysis AI Networking Model: a bottoms-up build of every buyer's clusters that gives the BOM of each one and the networking equipment TAM.

SemiAnalysis

179,465 subscribers

23,279 views • 13 days ago •via X (Twitter)

Science & Technology Finance

Anya Rossi• Live Now

Private livecam show

5 Comments

SemiAnalysis13 days ago

Learn more about the AI Networking Model and talk to our team:

Luminara Stocks13 days ago

Networking is the line item nobody models until a cluster is already late. Would love to see how different Oracle's looks from Meta's

The SI Maximalist13 days ago

Even the cables have buyer personas.

Leonid Invests13 days ago

Bottoms-up networking BOM per buyer sounds fun. Which cluster surprised you most? 🤔

港美10倍股玩家13 days ago

比起只看网络设备市场有多大，更想看不同买家的集群拆开后，设备清单究竟差在哪里。

Related Videos

Cerebras represents a whole NVL72 rack on a single wafer. By routing around defects and staying on-die, they bypass the networking power bottleneck that traditional GPU clusters face.

Cerebras represents a whole NVL72 rack on a single wafer. By routing around defects and staying on-die, they bypass the networking power bottleneck that traditional GPU clusters face.

SemiAnalysis

82,410 views • 5 months ago

$AVGO helps hyperscalers build the custom chips and networking fabric needed to run massive AI systems. If $GOOGL, $META, OpenAI or Anthropic want to build their own inference chips then Broadcom is one of the first calls because it has the IP, packaging expertise and networking portfolio required to scale custom AI systems. Thats why Broadcom remains one of the strongest infrastructure positions in the AI economy.

$AVGO helps hyperscalers build the custom chips and networking fabric needed to run massive AI systems. If $GOOGL, $META, OpenAI or Anthropic want to build their own inference chips then Broadcom is one of the first calls because it has the IP, packaging expertise and networking portfolio required to scale custom AI systems. Thats why Broadcom remains one of the strongest infrastructure positions in the AI economy.

Shay Boloor

77,765 views • 5 months ago

The future of AI depends on how compute, storage, networking, and software work together. Junchen Jiang, Co-founder of Tensormesh, on what it takes to build the next generation of AI infrastructure.

The future of AI depends on how compute, storage, networking, and software work together. Junchen Jiang, Co-founder of Tensormesh, on what it takes to build the next generation of AI infrastructure.

AI at AMD

116,017 views • 4 months ago

$NVDA networking is now running at a $60B+ annualized pace, growing ~200% YoY and nearly the size of $AVGO total 2025 revenue. As AI clusters scale, the bottleneck becomes how fast GPUs, CPUs, memory and storage move data together because even the best chips underperform if the network cannot keep up. This is why Nvidia is selling the full AI factory stack.

$NVDA networking is now running at a $60B+ annualized pace, growing ~200% YoY and nearly the size of $AVGO total 2025 revenue. As AI clusters scale, the bottleneck becomes how fast GPUs, CPUs, memory and storage move data together because even the best chips underperform if the network cannot keep up. This is why Nvidia is selling the full AI factory stack.

Shay Boloor

61,104 views • 5 months ago

How To Introduce Yourself At A Networking Event. 🤝 Follow @typecapital for more of what actually works when you're in the room! #networking #businesstips #firstimpressions #startuplife #elevatorpitch

How To Introduce Yourself At A Networking Event. 🤝 Follow @typecapital for more of what actually works when you're in the room! #networking #businesstips #firstimpressions #startuplife #elevatorpitch

tcburning

73,600 views • 2 months ago

WE’RE MOVING INTO THE NEXT PHASE OF AI Networking is the key -- widening the highways so AI traffic can flow efficiently: • $AVGO custom silicon for accelerators • $ALAB bandwidth for scalable inference • $MRVL optical backbone between clusters

WE’RE MOVING INTO THE NEXT PHASE OF AI Networking is the key -- widening the highways so AI traffic can flow efficiently: • $AVGO custom silicon for accelerators • $ALAB bandwidth for scalable inference • $MRVL optical backbone between clusters

Shay Boloor

103,926 views • 1 year ago

Kimi K3 is too large to run efficiently on a single accelerator so the workload must be distributed across many chips that exchange data at extremely high speeds. $CRDO and $ALAB benefit from the high speed connectivity inside AI clusters while $MRVL and $AVGO provide the networking, optical and custom silicon infrastructure required to scale them. Thats why inference constraints are bullish for the entire surrounding system including memory, networking, optics, switches, power, cooling and orchestration software.

Kimi K3 is too large to run efficiently on a single accelerator so the workload must be distributed across many chips that exchange data at extremely high speeds. $CRDO and $ALAB benefit from the high speed connectivity inside AI clusters while $MRVL and $AVGO provide the networking, optical and custom silicon infrastructure required to scale them. Thats why inference constraints are bullish for the entire surrounding system including memory, networking, optics, switches, power, cooling and orchestration software.

Shay Boloor

84,649 views • 3 months ago

.Bridgit Mendler on the founding thesis of Northwood. "We really were recognizing the same problem, that networking for the space economy is really what drives its value. And there's a significant bottleneck, and that bottleneck actually is not in space. It's actually on Earth. The way that you actually solve that networking problem for space is through introducing a model that gets you to the scale that you need, not just for one individual company or a mission, but for the industry overall." Full "Valley of Depth" episode in comments 👇

.Bridgit Mendler on the founding thesis of Northwood. "We really were recognizing the same problem, that networking for the space economy is really what drives its value. And there's a significant bottleneck, and that bottleneck actually is not in space. It's actually on Earth. The way that you actually solve that networking problem for space is through introducing a model that gets you to the scale that you need, not just for one individual company or a mission, but for the industry overall." Full "Valley of Depth" episode in comments 👇

Mo Islam

148,426 views • 4 months ago

Jensen Huang says that Nvidia’s full AI infrastructure package (chip, networking, data centre) is so efficient that competitors can price their chips at $0 and Nvidia would still be the better option.

Jensen Huang says that Nvidia’s full AI infrastructure package (chip, networking, data centre) is so efficient that competitors can price their chips at $0 and Nvidia would still be the better option.

Bearly AI

1,613,332 views • 1 year ago

What if every attendee at a conference had an AI counterpart networking on their behalf? For this year's AI on the Lot 2026 Amazon MGM Studios -- we built exactly that. Here's how it worked 🧵

What if every attendee at a conference had an AI counterpart networking on their behalf? For this year's AI on the Lot 2026 Amazon MGM Studios -- we built exactly that. Here's how it worked 🧵

Parag K. Mital

45,252 views • 5 months ago

$NVDA is evolving into the platform behind the entire AI factory spanning compute, CPUs, networking, interconnect, software, rack-scale systems and now the financing that helps customers build them. Owning more of that stack lets Nvidia capture more of every dollar spent on AI infrastructure while sitting across the bottlenecks that determine how quickly the entire market can scale.

$NVDA is evolving into the platform behind the entire AI factory spanning compute, CPUs, networking, interconnect, software, rack-scale systems and now the financing that helps customers build them. Owning more of that stack lets Nvidia capture more of every dollar spent on AI infrastructure while sitting across the bottlenecks that determine how quickly the entire market can scale.

Shay Boloor

153,566 views • 2 months ago

How to learn networking the fun way.

How to learn networking the fun way.

𝐇𝐚𝐦𝐳𝐚 | Networking Guy

44,095 views • 2 months ago

Jensen Huang says $NVDA invested $2B in $MRVL because the TAM for networking and custom silicon is entering an inference boom. Nvidia is becoming the platform layer for a more distributed AI factory architecture increasingly built around its networking fabric.

Jensen Huang says $NVDA invested $2B in $MRVL because the TAM for networking and custom silicon is entering an inference boom. Nvidia is becoming the platform layer for a more distributed AI factory architecture increasingly built around its networking fabric.

Shay Boloor

111,005 views • 7 months ago

What's the secret to a 36-hour pre-seed round? Always take that networking opportunity. Listen to the full episode of Build Mode to hear how founder and CEO of @geckomaterials Capella Kerst, convinced investors on the spot with a mobile demo:

What's the secret to a 36-hour pre-seed round? Always take that networking opportunity. Listen to the full episode of Build Mode to hear how founder and CEO of @geckomaterials Capella Kerst, convinced investors on the spot with a mobile demo:

TechCrunch

16,721 views • 6 months ago

speaking of networking I love how talkative & confident Taehyun is in every interview and how he always tries to keep the conversation going.

speaking of networking I love how talkative & confident Taehyun is in every interview and how he always tries to keep the conversation going.

nat☘︎ ́

43,769 views • 8 months ago

$AVGO has locked in some of the most important AI infrastructure customers on Earth across the next several years: • $GOOGL long-term deal to design and supply future TPU generations plus AI-rack networking through 2031 • $META initial MTIA commitment above 1GW with a multi-gigawatt roadmap through 2029 • Anthropic ~3.5GW of next-gen TPU-based compute starting 2027 (on top of ~1GW live in 2026) Clear validation that custom silicon and networking will be massive layers of the AI economy.

$AVGO has locked in some of the most important AI infrastructure customers on Earth across the next several years: • $GOOGL long-term deal to design and supply future TPU generations plus AI-rack networking through 2031 • $META initial MTIA commitment above 1GW with a multi-gigawatt roadmap through 2029 • Anthropic ~3.5GW of next-gen TPU-based compute starting 2027 (on top of ~1GW live in 2026) Clear validation that custom silicon and networking will be massive layers of the AI economy.

Shay Boloor

208,399 views • 5 months ago

"AI is really changing the way we need to think about networking." 💡 David Hughes, chief product officer, HPE Aruba Networking, discusses networking for AI and AI for networks. Listen to the interview from #HPEDiscover Barcelona. 👉

"AI is really changing the way we need to think about networking." 💡 David Hughes, chief product officer, HPE Aruba Networking, discusses networking for AI and AI for networks. Listen to the interview from #HPEDiscover Barcelona. 👉

HPE

4,924,765 views • 1 year ago

Education, community activities and a lot of networking! Take a look at what we got up to on the second day of #BinanceCampus ⤵️

Education, community activities and a lot of networking! Take a look at what we got up to on the second day of #BinanceCampus ⤵️

Binance

175,389 views • 3 years ago

Computer networking | TCP/IP Explained | How does TCP work? #networking #cybersecurity #tcp

Computer networking | TCP/IP Explained | How does TCP work? #networking #cybersecurity #tcp

DannyLegend.Cyber

24,000 views • 22 days ago