applying to NVIDIA for new grad SWE roles this cycle and I've been spiraling a bit. here's what I've actually found out from people who went through it, plus my own prep notes.
first: NVIDIA new grad is not the same pipeline as the full-time lateral hire pipeline. the new grad process is usually: OA -> phone screen -> onsite. simpler than the senior IC loop but still technical.
the OA: 2 leetcode-style questions, 75-90 minutes. medium difficulty. array manipulation and graph traversal were common themes I've heard about. standard FAANG OA tier. you need this to be clean to move forward.
phone screen: one coding question plus some background conversation. 45-60 min. if you pass this, you get the onsite.
onsite (new grad format, usually 3-4 rounds): 2 coding rounds: one data structures/algorithms, one a bit more applied (think: here's a problem a real team solved, show me how you'd approach it). mediums to one hard. no CUDA for new grads, don't panic about that. 1 behavioral: it's real. NVIDIA does STAR behavioral even for new grads. prep 4-5 stories. common questions: a project you're proud of, a time you had to learn something fast, a conflict in a team project. school projects count. 1 technical systems or domain: varies by team. some teams ask light systems design (for new grads they scope it appropriately). other teams skip this.
prep that actually matters: neetcode 150, focus on trees/graphs/DP because those show up one clean behavioral prep session, seriously, don't wing it look up what the team actually does if you know which org you're targeting
timeline for new grad: seems to be faster than senior hiring. I've heard 3-4 weeks to offer from OA.
comp for new grad in Santa Clara 2026: rough ranges I've seen: base $165-190k, signing $30-50k, RSU grants 4-year vest. total first year around $230-270k depending on level within new grad band. competitive with big tech new grad bands.
not going to pretend I'm not stressed. but at least now I know what I'm aiming at.