Passa al contenutoPassa al piè di pagina
  • Lavori
  • Aziende
  • Stipendi
  • Per le aziende

      Migliora la tua carriera

      Scopri le tue potenzialità di guadagno, trova lavori da sogno e condividi approfondimenti su lavoro e vita privata in forma anonima.

      employer cover photo
      employer logo
      employer logo

      NVIDIA

      Azienda coinvolta

      Chi siamo
      Recensioni
      Stipendi e benefit
      Lavori
      Colloqui
      Colloqui
      Ricerche correlate: Recensioni su NVIDIA | Offerte di lavoro di NVIDIA | Stipendi di NVIDIA | Benefit di NVIDIA
      Colloqui di NVIDIAColloqui per Deep Learning Architect presso NVIDIAColloquio di NVIDIA


      Glassdoor

      • Chi siamo
      • Contattaci

      Aziende

      • Account Business gratuito
      • Spazio per le aziende
      • Blog per le aziende

      Informazioni

      • Aiuto
      • Linee guida
      • Condizioni d'uso
      • Privacy e scelte pubblicitarie
      • Non vendere né condividere le mie informazioni
      • Strumento per l'accettazione dei cookie

      Lavora con noi

      • Inserzionisti
      • Carriere
      Scarica l'app

      • Cerca:
      • Aziende
      • Lavori
      • Località

      Copyright © 2008-2026. Glassdoor LLC. "Glassdoor," "Worklife Pro," "Bowls" e il relativo logo sono marchi registrati di Glassdoor LLC.

      Aziende seguite

      Non lasciarti sfuggire opportunità e informazioni privilegiate seguendo le aziende dove vorresti lavorare.

      Ricerche di lavoro

      Ricevi suggerimenti e aggiornamenti personalizzati avviando le tue ricerche.

      Colloquio per Deep Learning Architect

      16 nov 2017
      Dipendente anonimo
      Offerta accettata
      Esperienza positiva
      Colloquio nella media

      Candidatura

      Ho presentato la mia candidatura tramite segnalazione di un dipendente. Ho sostenuto un colloquio presso NVIDIA

      Colloquio

      Two rounds to three rounds of the phone interview plus one whole day on-site interview. Each round takes around one hour to undergo, covering the project experience and basic knowledge. Different people will interview you to access the technique level and communication skills.

      Domande di colloquio [1]

      Domanda 1

      Efficient ways to parallelize the matrix-multiplication
      Rispondi alla domanda
      3

      Altre recensioni di colloqui per Deep Learning Architect presso NVIDIA

      Colloquio per Deep Learning Architect

      14 dic 2019
      Dipendente anonimo
      Santa Clara, CA
      Offerta accettata
      Esperienza positiva
      Colloquio difficile

      Candidatura

      Ho presentato la mia candidatura tramite l'università. La procedura ha richiesto 2 mesi. Ho sostenuto un colloquio presso NVIDIA (Santa Clara, CA) nel mese di nov 2019

      Colloquio

      Met these guys at a career fair, got a phone call not that long after that for a behavioral/resume questions. Next phone call was a technical interview, it sounded like you can get tailored questions to your language of your liking, and I went with Python since it's easy. TL;DR: Few weeks later, I was flown down to Santa Clara to do the interview, and was given 5 rounds of interviews. Range of ML questions, behavioral questions, Python questions (which I think stemmed from the fact I told them Python is my jam), CI/CD design questions. For the whole story for those interested: Round 1 was given by some PhD. Gave me some simple Python data structure wrangling exercises. While I can't say the questions specifically, it required deep knowledge of how Python dictionaries are built under the hood under specific versions (e.g. 3.6 vs 3.8). We never talked about Python 2. Solution consisted of using builtin functions (e.g. map, filter, sorted), lambdas, and a few for loops. He actually asked the same question that was asked on the second phone interview. Round 2 was given by another PhD. Asked me some tensor questions. Pretty much data cleaning/normalization across 3+ dimensions. I think the secret/trick to solving the question was to solve it using 1 dimension first, and apply it to higher dimensions. End solution was essentially 4 nested for loops, which was promptly followed up by a multi-core parallel question. Required knowledge of knowing when a mutex would be effective, when it wouldn't, and how to parallelize code. Round 3 was given by the hiring manager (lunch). He bought me lunch and we talked about my previous experience. He was especially interested in my Microsoft internships. Round 4 was probably the worst round. He was new to the team and had poorly defined questions. He asked about CI/CD design, and never seemed satisfied with my answers. As if he had difficulty phrasing his questions (I don't think english was his first language). Lots of awkward silences, so I took the opportunity to slam him with a bunch of questions to prevent him from asking me obscure CI/CD questions. e.g. Why don't you like the existing CI/CD system? What market do you think CI/CD poorly targets? What would you change about it? etc. And after that I sort of asked him about company culture/what's with the beef with Apple? He seemed happy at the end of it all. Round 5 was the most rushed round. He came in, gave me an API for some component web-like component and told me to implement some functionality given that API. It was supposed to mimic how things worked in very low level. E.g. when you read a file with system calls, you'll get interrupted a lot and you have to try again a lot. We ran out of time for this one, since I'm a slow coder and he came in late. He didn't care about syntax of my code, so I even used math notation for some of the code. At the end he told me he disliked Python LOL. Flew back. Took some finals in school, got an offer after I bombed a final. Life's good.

      Domande di colloquio [1]

      Domanda 1

      Can't say specifically, but something close would be like: Given this tensor, how would you normalize it, using 32 cores? Is mutex a good idea for this?
      Rispondi alla domanda
      3

      Colloquio per Deep Learning Architect

      30 dic 2018
      Candidato anonimo a colloquio
      Boston, MA
      Nessuna offerta
      Esperienza positiva
      Colloquio nella media

      Candidatura

      Ho presentato la mia candidatura tramite un selezionatore. La procedura ha richiesto 3 settimane. Ho sostenuto un colloquio presso NVIDIA (Boston, MA) nel mese di nov 2018

      Colloquio

      First, I had a Skype interview with an experienced person from the group. He generally asked questions about my resume and my background work. The first interview lasted for an hour. After that I had a technical interview with another person from the group 2 weeks later, it lasted 45 minutes.

      Domande di colloquio [1]

      Domanda 1

      C++ and computer architecture questions
      Rispondi alla domanda
      1