Summer 2026
Firmware Engineer Intern
Astera Labs
1,001-5,000 employees
Semiconductor connectivity solutions for cloud AI
No salary listed
Vancouver, BC, Canada
In Person
 Software Engineering (1) |
|---|
Get referred to Astera Labs
See people who can refer or advise you
- Strong academic and technical background taking Computer Science, Electrical/Electronics/Software Engineering or equivalent.
- Completed one or more Co-op terms of firmware or embedded software development.
- Completed one or more courses in Real Time Operating Systems (RTOS) or detailed use of an RTOS during a prior coop term or project.
- Authorized to work in Canada and available for 4 or more months.
- Proficiency in C (preferred) or C++, including development of firmware.
- Proficient in Python for scripting.
- Experience with developer workflows, SCM (preferably git), code reviews, CI.
- Experience developing firmware to execute in on-chip microcontrollers.
- Experience debugging firmware/hardware issues using JTAG, trace logs, or other lab tools.
- The mission of this role is to architect and develop firmware and microcontroller subsystems for Astera Labs’ SoC and systems products.
- Knowledge of tools like PyTest or GoogleTest is an asset.
- Knowledge of software build environments, CI/CD, gcc/make is an asset
- Experience with lab hardware/firmware debug is an asset.
- Experience developing firmware for PCIe/CXL/Ethernet is an asset.
Astera Labs provides semiconductor-based connectivity solutions to boost cloud and AI infrastructure. Its products include PCIe, CXL, and Ethernet connectivity ICs and adapters that enable high-speed data transfer between processors, memory, and accelerators. These solutions help alleviate the memory wall by improving memory and I/O bandwidth, enabling faster communication within data centers supporting Generative AI workloads. The company differentiates itself by specializing in memory-centric, high-speed interconnects for cloud AI architectures and by targeting data centers and AI infrastructure customers with a portfolio of PCIe, CXL, and Ethernet products, rather than broad consumer-focused offerings. Astera Labs aims to expand its leadership in the global cloud AI market through continued development and deployment of its connectivity solutions, growing its market share and partnerships in data centers and AI workloads.
Company Size
1,001-5,000
Company Stage
IPO
Headquarters
Santa Clara, California
Founded
2017
Get referred to Astera Labs
See people who can refer or advise you
Simplify's Take
What believers are saying
- Taurus 3.2T and 200G-per-lane products target urgent AI bandwidth bottlenecks.
- Smart Swap reduces server-board redesign and accelerates hyperscaler time-to-market.
- Expansion into Scorpio switches and CXL memory controllers broadens wallet share.
What critics are saying
- Hyperscaler standardization on NVIDIA-led stacks can displace Astera’s attach opportunities.
- Open-standards competition pressures margins because rivals target the same protocols.
- Concentrated AI capex creates revenue downside if hyperscaler buildouts slow.
What makes Astera Labs unique
- Purpose-built connectivity for rack-scale AI across PCIe, CXL, Ethernet, UALink, and NVLink Fusion.
- COSMOS adds telemetry, fleet management, and security across Astera’s connectivity platform.
- Custom connectivity solutions complement standardized products for hyperscaler-specific architectures.
Help us improve and share your feedback! Did you find this helpful?
Benefits
401(k) Retirement Plan
Flexible Work Hours
Growth & Insights and Company News
6 month growth
↓ -8%1 year growth
↓ -11%2 year growth
↓ -8%Astera Labs reported record second-quarter revenue of $392.4 million, up 104% year over year, driven by strong demand for AI infrastructure connectivity products. PCIe 6.0 products accounted for over half of total sales, up from roughly one-third in the previous quarter. Non-GAAP earnings per share rose to $0.80, whilst gross margin reached 73.7%. The company's Scorpio X-Series fabric switches entered volume production and are expected to become Astera's largest product family by revenue in the third quarter, earlier than previously forecast. Management projects third-quarter revenue between $540 million and $560 million, representing approximately 40% sequential growth. Chief executive Jitendra Mohan attributed the growth to continued spending by hyperscalers, sovereign AI initiatives, and expanding enterprise adoption. The company is also developing optical-connectivity products targeted for volume production beginning in 2027.
Jim Cramer has reversed his stance on Astera Labs, telling investors to "buy, buy, buy" the stock during CNBC's "Mad Money" on 29 July. The shift is notable because Cramer previously flagged the semiconductor company as too expensive compared to established chip makers like Broadcom. Astera Labs shares fell more than 40% from an all-time high of $499.48 on 30 June to the mid-$200s in late July, caught in a broader selloff of AI stocks. The decline wasn't driven by business fundamentals — first-quarter revenue rose 93% year over year to $308.4 million. Cramer said the drop reflected market sentiment rather than operational issues. The company reports second-quarter results on 4 August. Astera Labs designs connectivity hardware that moves data between chips in AI data centres.
Astera Labs and Intel present contrasting semiconductor investment opportunities as the AI boom reshapes the industry. Astera Labs specialises in high-performance connectivity solutions for AI data centres, serving major hyperscalers. The company achieved explosive growth in 2025, with revenue reaching approximately $852 million, up 115% year-over-year. It posted net income of nearly $219 million and maintains a debt-to-equity ratio of 0.0x. However, a single customer represented over 70% of 2025 revenue, and stock-based compensation accounted for roughly 50% of operating cash flow. Intel, one of the world's largest semiconductor companies, is executing a turnaround strategy to become a leading global foundry. In 2025, revenue reached close to $53 billion, down 0.5% year-over-year, whilst the company reported a net loss of roughly $267 million.
Astera Labs and Kratos Defense & Security Solutions represent contrasting investment opportunities in technology. Astera focuses on AI infrastructure connectivity, whilst Kratos develops unmanned and hypersonic defence systems. Astera's FY 2025 revenue surged 115.1% to $852.5 million, with net income of $219.1 million yielding a 25.7% net margin. The company carries no debt and maintains a 10.2x current ratio. However, one customer accounts for over 70% of revenue, creating concentration risk. Stock-based compensation represented roughly 50.1% of operating cash flow. Kratos posted FY 2025 revenue of $1.3 billion, up 18.5%, with net income of $22 million and a 1.6% net margin. The US government generates approximately 68% of revenue. Recent wins include $400 million in defence funding and a $156 million Project Solar Shield contract. Astera offers explosive growth tied to AI infrastructure expansion, whilst Kratos provides steady exposure to national security demand.
Semiconductor companies supplying AI data centre networking components are experiencing unprecedented demand as hyperscalers build larger AI clusters. The global semiconductor market reached a record $120.6 billion in May 2026, rising 9.2% sequentially and 104.1% year over year. AI clusters are rapidly migrating from 400G to 800G Ethernet networking, with 1.6-terabit networking expected in 2027. This transition is driving exponential growth in demand for Ethernet switches, optical interconnects, and related components. Four networking semiconductor stocks are highlighted for August 2026: Credo Technology Group Holding, Broadcom, Marvell Technology, and Astera Labs. Credo Technology forecasts over 80% revenue growth in fiscal 2027, with its optical networking business expected to generate more than $600 million. The shift transforms data centres into high-performance AI facilities, creating substantial opportunities for semiconductor suppliers beyond traditional graphics processing units.