Arteris missed the Street on revenue, but the contract base is the print that matters

The thesis from this print is that Arteris is becoming easier to own as a semiconductor IP compounding story, even though the headline revenue line did not cooperate with consensus. What was priced in was $18.6 million of revenue and a slightly wider non-GAAP loss, with the Street expecting EPS of -$0.11. What actually surprised was the split: revenue of $17.4 million missed by -6.2%, while EPS of -$0.09 beat by +18.2%. That combination usually invites skepticism, because software and IP companies can make a quarter look cleaner through expense timing. Here, the more important evidence sits in the contract metrics: management said annual contract value plus royalties reached $74.9 million, while remaining performance obligations reached $104.7 million. Those are not expense-managed numbers; they are backlog and recurring-value indicators. The market may be mispricing the quarter by over-penalizing the revenue miss and underweighting the fact that forward contracted revenue crossed the $100 million milestone for the first time.

The reason the headline miss is not fatal is that revenue is still following the slope management needs for the model to work. On the Street-comparison basis, Q3 revenue was $17.4 million against $18.6 million expected, so the miss was real and should not be waved away. But the company’s own reported basis also matters for internal trajectory, and Nicholas Hawkins framed the quarter as above management’s range: “Total revenue for the third quarter was $17.4 million, up 5% sequentially and 18% year-over-year and above the top end of our guidance range.” The quote matters because it separates two investor questions. Relative to the sell side, the company disappointed. Relative to the plan management had given investors, the company exceeded the revenue bar. That makes the debate less about demand deterioration and more about whether the Street had pulled too much revenue from future periods into Q3.

The financial trajectory supports that interpretation because the model has moved from a low-teens revenue plateau into a higher run-rate without giving up its core margin structure. Revenue has risen from $14.7 million in Q3 FY2024 to $17.4 million in Q3 FY2025, and gross margin stayed essentially in the same high band at 90.1% versus 89.9%. That is the crux for a PM: the company is not buying growth with a material gross-margin concession. The EPS line remains negative, but the loss profile is improving on the metric investors use for this stage of the model, with the quarter’s EPS at -$0.09 versus consensus at -$0.11. A revenue miss paired with stable gross margin and a smaller loss does not prove the model is fixed, but it does argue against the simplest bearish read that demand is weakening and profitability is being deferred.

The contract metrics are the variant perception because they turn the quarter from a miss into a visibility upgrade. Hawkins said annual contract value plus royalties was $74.9 million, up 24% year-over-year, and remaining performance obligations were $104.7 million, up 34% year-over-year. The magnitude matters: ACV plus royalties is now larger than the company’s expected full-year revenue base, while RPO provides a contracted revenue reservoir that is materially above the current annualized run-rate. In semiconductor IP, those numbers are more valuable than a single-quarter revenue print because licenses and royalties convert over design cycles, not clean retail-style reporting periods. The market’s likely mistake is treating the -6.2% revenue miss as a demand signal when the better demand indicators, ACV plus royalties and RPO, both moved to records.

That forward-base argument is stronger because management raised the year rather than simply asking investors to ignore the miss. For the fourth quarter, Hawkins guided revenue to $18.4 million to $18.8 million and non-GAAP operating loss to $2.3 million to $3.3 million. For the full year, he said revenue guidance moved to $68.8 million to $69.2 million, “also an increase of $1 million compared to our prior guidance.” The wording matters because a company that misses the Street and then lifts the full-year target is telling investors the Q3 shortfall was not a broken funnel. It also raises the hurdle for the bear case: to argue that revenue quality is deteriorating, one has to explain why management would simultaneously guide Q4 above Q3 and lift full-year revenue. The cleaner critique is that Arteris remains small and loss-making, but that is different from saying the quarter showed demand slippage.

The cost structure is not yet where public-market semiconductor software investors ultimately want it, but Q3 gave evidence that incremental revenue is beginning to matter. Hawkins cited non-GAAP operating expense of $19.5 million and non-GAAP operating loss of $3.5 million, with free cash flow positive $2.5 million. The important connection is that operating expense still exceeds gross profit, so the company is not past the leverage threshold, but the loss is now small enough that revenue conversion from the RPO base can change the debate. Cash risk also looks contained for now: management ended the quarter with $56.2 million in cash, cash equivalents and investments, and no financial debt. For portfolio construction, that means the investment case is less dependent on near-term financing and more dependent on whether ACV and RPO convert into revenue at the guided pace.

The royalty detail is especially relevant because it links Arteris to silicon complexity rather than just seat expansion or license renewals. Hawkins said variable royalties for the trailing period were up 36%, and he tied that to royalties growing at roughly twice the rate of licenses. That is the most important qualitative claim in the call because it quantifies the upside mechanism: if customers ship more devices containing Arteris IP, the company participates after the design win rather than only at contract signing. Karel Janac also pointed to advanced foundry nodes, “particularly 5-nanometer, 3-nanometer, 2-nanometer and as we head into the Angstrom era of silicon.” The node language matters because network-on-chip complexity increases with heterogeneous compute and chiplet-like architectures. The investable question is whether this turns Arteris into a small royalty lever on more complex semiconductor systems, and the 36% royalty growth is the first number in the pack that supports that view.

That same complexity lens also explains why the peer comparison is more favorable than the company’s size would suggest. Arteris is nowhere near the scale of SNPS at $2,276.0 million of quarterly revenue or CDNS at $1,474.2 million, but its gross margin of 89.9% sits much closer to high-quality IP and EDA economics than to lower-margin semiconductor product models. ARM posted 93.1% gross margin with +20.1% revenue growth, while CDNS posted 95.8% gross margin with +18.7% revenue growth. Arteris is smaller and still loss-making, so it does not deserve those business-model multiples mechanically. But the comparable point is that its Q3 revenue growth of +18.3% and near-90% gross margin place it in the same economic family as the IP-rich peers, not the commodity end of the subsector. The mispricing risk is that investors screen out Arteris because absolute revenue is only $17.4 million, while the unit economics resemble the businesses that scale best once sales coverage and R&D intensity normalize.

The supply-chain read-through is narrower than usual because the data pack lists no named customers and no named suppliers for Arteris, so there is no defensible customer-specific call to make from this print. That absence is itself relevant: the quarter’s second-order implications should be routed through peers and end-market design activity, not through a named account thesis. For ARM, the read-through is constructive but not company-specific: Arteris’ royalty growth of 36% and ARM’s +20.1% revenue growth both point to sustained monetization of semiconductor IP as designs move into more complex nodes. For SNPS and CDNS, the read-through is that design activity remains active enough to support EDA and IP spend, but Arteris’ $104.7 million RPO milestone suggests smaller specialized IP vendors can still take budget inside accounts dominated by larger toolchain suppliers. There is no data here to claim share loss at SNPS or CDNS, only that Arteris is growing its contracted base alongside them.

The call delivery reinforced the thesis, but with one caution that deserves attention. The tone history shows Q3 sentiment at 0.34 and guidance tone at 0.35, while uncertainty fell to 63.9 from the prior call’s 77.2. That is a better setup than the raw revenue miss would imply, because management sounded more constructive on guidance while using less uncertain language. The caution is Q&A evasiveness, which moved to 31.1 in Q3. That conflicts with the improved guidance tone and lower uncertainty, so the right interpretation is not unqualified confidence. Management gave firmer forward numbers, but the Q&A pattern suggests investors were probing areas where answers may not have been as direct as the prepared script. In a small IP company, that matters because conversion timing, royalty sensitivity and large-deal concentration can all make quarterly revenue harder to forecast.

The tone data also helps explain why this print should be treated as a thesis checkpoint rather than a clean victory lap. Prepared sentiment was 0.50, while Q&A sentiment was 0.12, which means the optimism was concentrated in management’s scripted framing rather than the analyst exchange. That does not invalidate the numbers, but it tells PMs where to press. The prepared remarks gave investors record ACV plus royalties and record RPO; the Q&A apparently did not eliminate all concerns around the path from contract value to recognized revenue. This is where discipline matters: do not pay for the full RPO number as if it were same-quarter revenue, but also do not ignore a 34% year-over-year increase in contracted future revenue because the quarter missed the Street by -6.2%.

The near-term setup is therefore asymmetric but not riskless: Q4 needs to prove that the Q3 miss was timing, not optimism. The first watch item is revenue against the $18.4 million to $18.8 million guide; a print below that range would break the variant perception because it would turn a one-quarter Street miss into a conversion problem. The second is ACV plus royalties exiting 2025 inside the $74 million to $78 million range; falling below that would challenge the view that the record $74.9 million base is durable. The third is cash conversion, with Q4 non-GAAP free cash flow guided to $0.2 million to $3.2 million and full-year non-GAAP free cash flow guided to $2.5 million to $5.5 million. Confirmation is simple: Q4 revenue in range, ACV plus royalties in range, and positive free cash flow would validate that the market overreacted to the Q3 revenue miss. The break point is equally clear: any reset to the $68.8 million to $69.2 million full-year revenue guide, or a failure to hold RPO near the $104.7 million milestone, would say the contract-base thesis is not converting fast enough to own.