Wall Street Initiates Coverage on Cerebras with Price Targets Up to $300
Alina Collins
Morgan Stanley, UBS, Needham, and Wedbush all initiated coverage on Cerebras (CBRS) simultaneously, setting price targets at $250–$300; the consensus is clear — the wafer-scale chip holds a differentiated edge in fast inference that Nvidia cannot easily replicate.
Four firms moved at once — what did they say?
Morgan Stanley rated Cerebras "overweight" with a $250 target. UBS and Needham both rated it "buy" at $300. Wedbush rated it "outperform" at $270.
This means → the median target across all four is roughly $275, every one above the current price, and not a single neutral or sell rating among them.
In plain terms = Wall Street is unusually unanimous on this newly listed company. The only debate is *how bullish*, not *whether* to be bullish.
What makes Cerebras's chip different?
Cerebras's core product is the WSE — a wafer-scale engine that turns an entire silicon wafer into a single chip, tens of times larger than a typical GPU.
This means → the massive chip area lets more computation happen on one die, cutting the time data spends moving between chips — critical for fast-inference workloads.
Wedbush analyst Matt Bryson put it most directly: "In the inference market, speed — not raw compute — determines the value of output."
How solid are the big-customer contracts?
OpenAI is the anchor client. In January 2026 it signed a compute agreement worth over $20 billion, committing to deploy roughly 750 MW of Cerebras capacity by 2028, with an option for an additional 1.25 GW.
Amazon AWS announced a distributed-inference partnership with Cerebras in March; the purchase scale has not been disclosed.
This means → Cerebras's revenue visibility is heavily concentrated in one OpenAI contract. If the AWS partnership converts to material orders, it becomes a second growth curve.
How did UBS arrive at the $300 target?
UBS projects Cerebras reaching $11 billion in sales by fiscal C29 (roughly 2029), applies a 10× EV/Sales multiple, and discounts back 18 months to reach $300.
In plain terms = UBS is betting that revenue can ramp from near-zero to the tens-of-billions level within three years — whether that assumption holds is the make-or-break for the target price.
Needham analyst Bolton added: if OpenAI exercises the additional option or Amazon purchases materially, current models have significant upside.
What does the IPO performance tell us?
Cerebras priced its IPO at $185 per share, raising $5.55 billion. Shares opened as high as $350 before pulling back.
This reflects intense market enthusiasm for the "AI inference" thesis, but also that short-term expectations had already been priced in.
This means → the four firms' targets ($250–$300) sit roughly between the IPO price and the opening-day high — Wall Street sees fair value in that band, neither the $185 bargain nor the $350 frenzy.
Content is for reference only, not financial advice.