Bottleneck Solvers

BN · rank #12 · 2026-09-04

RMBS

NASDAQ · $9.12B (USD)

SignalBuy3.1strength 3.1 / 10
verdict · PROCEED

The exact numbers the algorithm saw.

Bottleneck themes
The structural shortage this name supplies into.
Advanced AI chip packaging and memory
Factor scores and the inputs behind this pick.
Composite score
Z-score blend of the factors below; drives the rank.
0.51
Forward revenue growth
Consensus forward revenue growth.
+19.7%
Forward net margin
Consensus forward net margin.
+32.3%
Net margin TTM
Trailing twelve month net margin.
+31.7%
Margin expansion
Forward minus trailing net margin (percentage points).
+0.6%
Forward PEG
Forward P/E to growth. Below 1 is cheap for the growth.
3.80
Debt / FCF
Net debt relative to free cash flow. Lower is safer.
0.07×
Analyst upside
Spread between the consensus 12m target and the current price.
+74.9%
Last EPS surprise
Most recent reported EPS versus consensus.
+10%
Market cap (USD)$9.12B

The AI research card

Independent qualitative review of each pick before the order is placed. On the rare day the research service is unavailable, the paper book trades on the quant ranking alone and no card appears here.

Summary

Rambus is a fabless semiconductor and IP licensing company supplying DDR5 memory interface chips, HBM/GDDR/PCIe/CXL controller IP, and security IP to data center and AI infrastructure customers.

Rationale

Record Q2 2026 revenue (+20% YoY), ~45% non-GAAP operating margins, a 10% EPS beat, and an expanding PCIe 7.0/CXL 3.1 portfolio directly confirm the bottleneck-solver thesis—Rambus sits at the memory and interconnect chokepoints that AI scale-out cannot bypass.

Material risks

  • 1Royalty revenue (~$84M/quarter) depends on licensing agreements with a concentrated set of DRAM manufacturers (Samsung, SK Hynix, Micron); a non-renewal or renegotiation at lower rates would structurally impair ~40% of revenue with little lead time.
  • 2Forward PEG of 3.8x signals the market has already priced significant execution; any deceleration in AI infrastructure capex or a product cycle delay in HBM4/PCIe 7.0 adoption compresses the multiple sharply given limited margin for error.

AI verdict council

Each pick is reviewed independently by 3 models before any order. 2 of 3 voted to proceed.

OpenAI
Caution
Claude
Proceed
Gemini
Proceed