EP:32: CBRS: AI’s New Leader? With the Pre-IPO Champ
In this episode of Pitch The PM, Doug Garber speaks with Chris Murphy about his pre-IPO investment strategy and why $CBRS is his largest investment. Chris previously spent 17 years scaling a multi billion fund and is now focused on his pre-IPO PA (personal account) investment strategy.
Chris has dug-in deep, hitting the road and reading heaps of expert call transcripts from Alpha Sense and Third Bridge. He believes inference speed will become one of the most valuable advantages in the race to AGI (artificial general intelligence). The conversation examines the advantages of Cerebras’ wafer-scale chip design, the growing importance of fast inference for agentic AI workflows and the price/cost trade-off. OpenAI was the first mover, securing initial capacity and options. AWS has a term sheet for capacity. The chip market is evolving and Doug & Chris debate the implications for NVIDIA, Google and Amazon.
Tickers: $CBRS, $NVDA, $GOOG, $AMZN, $META
[00:00:00] Chris Murphy explains why AI inference speed will become a key competitive advantage.
[00:00:20] Doug Garber introduces Chris Murphy and discusses Cerebras’ IPO and price discovery.
[00:00:57] Chris Murphy shares his background in public and private market investing.
[00:03:44] Chris outlines his framework for evaluating pre-IPO investments.
[00:04:31] Why Chris believed inference would become the AI “money maker.”
[00:05:08] Chris explains how Cerebras removes bottlenecks in agentic AI workflows.
[00:06:07] Discussion on token economics, inference speed, and use case advantages.
[00:08:28] Cerebras’ position in fast inference versus Groq and NVIDIA.
[00:09:30] Breakdown of Cerebras’ wafer-scale chip design and routing fabric.
[00:11:48] Chris discusses the research behind his Cerebras investment thesis.
[00:12:35] Expert network transcripts, proprietary calls, and diligence methods.
[00:13:14] Chris walks through his revenue framework for OpenAI, AWS, and others.
[00:15:12] Discussion on customer concentration risk and OpenAI exposure.
[00:15:53] Why Amazon could benefit from combining Trainium with Cerebras chips.
[00:17:15] Competitive positioning versus Google TPUs and NVIDIA GPUs.
[00:18:39] Why inference is expected to outgrow training workloads.
[00:20:33] Chris explains his OpenAI revenue assumptions through 2033.
[00:23:48] How AWS and Anthropic fit into the broader Cerebras opportunity.
[00:26:24] Chris discusses Cerebras as an “arms dealer” in the AGI race.
[00:29:49] Feedback from engineers and users testing Cerebras chips.
[00:32:59] How AI systems route workloads across chips and inference models.
[00:34:54] Chris explains assumptions around Cerebras’ AWS inference share.
[00:36:35] Discussion on pricing, leasing models, and AWS exclusivity.
[00:39:49] Breakdown of G42, MBZUAI, and other customer relationships.
[00:42:55] Chris explains why the long-term revenue opportunity makes investors uncomfortable.
[00:45:38] Discussion on customer warrants and partnerships with OpenAI and AWS.
[00:49:25] Speculation around potential relationships with Meta and Google.
[00:50:16] Gross margin expectations, scalability, and long-term profitability.
[00:52:43] Doug Garber summarizes execution risks and the AI infrastructure landscape.
[00:55:19] Chris shares his long-term outlook on Cerebras and position sizing.
[00:59:16] Final thoughts on AI enthusiasm, IPO risks, and investing discipline.
Links:
Doug Garber on LinkedIn: https://www.linkedin.com/in/doug-garber-42aa508 Chris Murphy on LinkedIn: https://www.linkedin.com/in/chrismurphyhq
Powered by AlphaSense. Complimentary access — https://www.alpha-sense.com/Pitch/
Complimentary pre-built CBRS excel model — https://istari.insyncanalytics.com/pitchthepm/?model=IPO&template=CBRS
Subscribe to our newsletter: https://pitchthepm.beehiiv.com/subscribe
This episode is for informational purposes only and does not constitute investment advice. Full disclosure: https://www.pitchthepm.com/disclaimer