報告_OCP2026_Credo_寬慢光學勝銅_20261015

PDF 原檔：報告_OCP2026_Credo_寬慢光學勝銅_20261015_original.pdf

講者公開簡報（OCP 官方 Recorded Content 頁提供下載）。文字以 PyMuPDF 逐頁抽取，### p<n> 為頁碼；圖表內容見原檔。

圖片清單（已驗證 2026-10-19）

原始內容

p1

（本頁無可抽取文字，內容在圖中）

p2

Bill Brennan President and CEO, Credo OCTOBER 2026 Optics that Beat Copper on Reliability, Power, AND Cost

p3

We have made statements in this presentation that are forward-looking statements. In some cases, you can identify these statements by forward-looking words such as “may,” “might,” “will,” “should,” “expects,” “plans,” “anticipates,” “believes,” “estimates,” “predicts,” “potential” or “continue,” the negative of these terms and other comparable terminology. These forward-looking statements, which are subject to risks, uncertainties and assumptions about us, may include projections of our future financial performance, our anticipated growth strategies and anticipated trends in our business and in the industry in which we operate. These statements are only predictions based on our current expectations and projections about future events. There are important factors that could cause our actual results, level of activity, performance or achievements to differ materially from the results, level of activity, performance or achievements expressed or implied by the forward-looking statements, including those factors discussed in our registration statement, including (but not limited to) those discussed under the caption entitled “Risk Factors” in our registration statement. You should not rely upon forward-looking statements as predictions of future events. Although we believe that the expectations reflected in the forward-looking statements are reasonable, we cannot guarantee that the future results, levels of activity, performance or events and circumstances reflected in the forward-looking statements will be achieved or occur. These forward-looking statements speak only as of the date of this prospectus. We undertake no obligation to update publicly any forward-looking statements for any reason after the date of this prospectus to conform these statements to actual results or to changes in our expectations, except as required by applicable law. You should read this prospectus and the documents that we reference in this prospectus and have filed with the SEC as exhibits to the registration statement of which this prospectus is a part with the understanding that our actual future results, levels of activity, performance and events and circumstances may be materially different from what we expect. Safe Harbor Disclosure

p4

Why do we Use Copper Where we Can? 100M hrs MTBF and MTBI Copper is Passive Low Power <$0.10 /Gb/s Low Cost Reliability

p5

• 2D Array of Emitters & Receivers – of many types • Little to no channel compensation Metric Red µLED Blue µLED µVCSEL Link Power pJ/bit < 1 < 2 < 2 Fiber Count 10,000s 1,000s 100s Usable reach ≤ 2 m ≤ 10 m ≤ 50 m Applications Scale-in Scale-in Scale-up Scale-in Scale-up Scale-out Wider. Slower. Better.

p6

Copper-Level Reliability — In Optics. Reliability Low Power Data Spare Mgmt. Gearbox Host interface + Mgmt. Crossbar MCU • Layer Zero Redundancy - Copper Level link Reliability • ZeroFlap Performance - MTBF / MTBI of 100M hrs Low Cost ….

p7

µEmitter Optics — 60% Less Power than Copper Reliability Low Power Low Cost Copper is passive, but SerDes aren’t… 400Gb bidi ~ 5pJ/bit per Tx/Rx pair VR200 NVL72 Copper Backplane 260TB/s = 10kW ~5% of VR200 Rack Power µEmitters @ <2pJ/bit save 60% Saves 6kW/Rack or 80W/GPU

p8

At $100/W, Low Power = Low Cost Reliability Low Power Low Cost How much is a Watt worth? OpEx $5/Watt CapEx $20/Watt Opportunity Cost  $100/W Nscale $45B 460MW Lambda $35B 350MW SpaceX $45B 300MW $100 DAC = $1,600 Opportunity Cost VR200 Backplane = $600k Lost Opportunity Cost

p9

µEmitter – Optics that Beat Copper Reliability Low Power Low Cost 100M hrs MTBF and MTBI Copper is passive Serdes are 5pJ/bit Cheap to buy $100/W opportunity cost to use Up to 60% less power Near copper cost to buy Up to 60% lower cost to use 100M hrs MTBF and MTBI Zero clean / zero maintenance Copper µEmitter

p10

Credo’s 1.6T ZeroFlap ALC 50m Reach 12W/end 10x longer, 10x lighter, and lower power Volume in 2027 100M hours MTBF & MTBI PILOT Integrated

p11

2027 One Architecture. A Decade of Roadmap. Scale-up Scale-in 2028 / 29 2030+ Scale-out 6.4T NPO Petabit Scale-In 1.6T ALC 3.2T ALC

5Tb/mm CPO Faster emitters + tighter pitch = n² scaling

p12

Introducing the ZeroFlap Blindmate Backplane 30Pb/s / 1M fiber µEmitter Backplane • 100M hrs MTBF / MTBI – zero clean / zero maintenance • 15% over-provisioning – hitless fiber failover • Uses Senko blind-mate connectors and Credo ferrule • Compared to today’s copper backplane: • 30x as dense – 1.8Pb/tray • 30x copper reach - 50m • 90% less weight

p13

Bringing µEmitters to OCP

p14

Emitters Fiber & Connectors Integrators End Users Introducing Short Reach µEmitter Optics Hyperscaler 2xGPU (SRMO OCP Workstream) 2nd Emitter

p15

Same Reliability. Lower Power and Cost. Reliability Low Power Low Cost 100M hrs MTBF and MTBI Copper is passive Serdes are 5pJ/bit Cheap to buy $100/W opp. cost to use Up to 60% less power Near copper cost to buy Up to 60% lower cost to use 100M hrs MTBF and MTBI Zero clean / zero maintenance Copper µEmitter

p16

• Visit the Credo booth and see our ALCs and µEmitter Optical Backplane. • Bring your next interconnect challenge. Let’s evaluate where µEmitters fit. Copper-level reliability. Up to 60% lower power. Lower ownership cost. Wider. Slower. Better.

p17

Thank you.