Nvidia, Menlo Micro collaboration speeds up AI chip testing

Nvidia and startup Menlo Micro said they have deployed Menlo’s MEMS-based metal switching chips to substantially speed testing of Nvidia GPUs—improving test throughput by 30% to 90% depending on the test—addressing a key production and validation bottleneck for high-power, high-speed AI chips. Menlo, a 2016 GE spinout that has raised $227.5 million from backers including Corning and Tony Fadell’s fund, said other major chipmakers are adopting the switches though it would not disclose the size of its Nvidia business; CEO Russ Garcia stressed faster validation is critical to avoid errors in data-center deployments. The timing is notable as Nvidia heads into earnings (analysts expect ~56% sales growth to $56.9 billion), and improved test throughput could materially ease supply-side constraints as demand for AI processors remains strong.

Analysis

Nvidia and Menlo Micro announced deployment of Menlo's MEMS-based metal switching chips to speed Nvidia GPU test throughput by 30% to 90% depending on the test, according to a joint research paper reported by Reuters. Menlo Micro, a 2016 GE spinout that has raised $227.5 million from investors including Corning and Tony Fadell's venture fund, supplies switch technology intended to replace decades-old components on test boards that struggle with the power and speed demands of modern AI GPUs.

Faster validation addresses a concrete production bottleneck because unvalidated GPUs can produce data-center errors; Nvidia has sold millions of AI chips and is reported to be working to iron out production kinks as demand remains strong. The timing coincides with Nvidia's upcoming earnings report where analysts expect roughly 56% sales growth to $56.9 billion, so operational improvements that boost test throughput could materially ease supply-side constraints and support near-term revenue delivery.

Material uncertainties remain: Menlo declined to quantify business with Nvidia and said other chipmakers are adopting the switches, leaving the scale and timing of supply-chain benefit unclear. Investors should weigh the operational upside against noted market caution over stretched AI valuations and monitor management commentary and concrete yield/throughput metrics in the earnings call and subsequent supplier disclosures.

Key Decisions for Investors

• Position bias: adopt a modestly positive stance on NVDA ahead of and immediately after earnings given the potential supply relief from faster test throughput, but size positions to limit downside if guidance disappoints
• Event monitor: prioritize management commentary and explicit metrics on test throughput, yields and timelines in the earnings call and supplier updates before increasing exposure, because Menlo declined to quantify current business with Nvidia
• Supply-chain watch: monitor signs of broader adoption of Menlo's switches and any disclosures from investors such as Corning (GLW) for indirect exposure, but avoid assuming immediate material revenue lift until firm data are released
• Risk management: given elevated AI valuations noted in the article, consider partial profit-taking or hedges if NVDA exposure is large, and reassess on confirmed production and guidance improvements