AI Software Lead – PyTorch & CUDA Runtime (Next-Gen Accelerator)
SanDisk is seeking an AI Software Lead to own the runtime and neural network layer for a next-generation AI accelerator platform. The role involves designing and optimizing performance-critical NN operators and driving runtime-level scheduling improvements. The candidate will work closely with compiler and framework teams to maximize hardware utilization and efficiency. This position requires deep expertise in C++, CUDA, and deep learning frameworks.
50k new jobs listed every day. Install TAL to find more jobs like this.
Experience
8+ years
Function
Engineering
Work mode
Onsite, India
Company
Tier 2
What you will work on
SanDisk is seeking an AI Software Lead to own the runtime and neural network layer for a next-generation AI accelerator platform. The role involves designing and optimizing performance-critical NN operators and driving runtime-level scheduling improvements. The candidate will work closely with compiler and framework teams to maximize hardware utilization and efficiency. This position requires deep expertise in C++, CUDA, and deep learning frameworks.
TAL's take
High-impact technical role at a global hardware company, specifically focused on next-generation AI accelerators.
Very clear responsibilities focused on runtime, operator optimization, and hardware acceleration, with a well-defined stack.
Salaries at SanDisk
16.3 LPA average
Based on 3 Grapevine salary entries for SanDisk.
Other roles
2 - 4 years
10 LPA average
Range: 6 - 15 LPA
Other roles
8 - 10 years | Staff Engineer
28 LPA average
Range: 28 - 28 LPA
Must haves
- 8+ years in systems software, runtime, or performance engineering
- Experience with PyTorch, TensorFlow, or JAX
- NN operator or kernel development and optimization
- Proficiency in C/C++ and CUDA
- Understanding of memory hierarchy and parallel execution
Tools and skills
Nice to have: gpu, npu, custom ai accelerators, xla, mlir, compiler-runtime interaction.
About the company
Global established technology company, though currently positioned as a hardware-centric memory provider rather than a pure-play AI infrastructure firm.
Posts mentioning SanDisk
Ultimate Companies List
# Companies in India with highest pay ## Good WLB - Linkedin - Microsoft - Google - Atlassian - Adobe - Salesforce - Intuit - Apple - Twitter - Indeed - Expedia - Nvidia - VMWare - Intel - Flipkart - Inmobi - Nutanix - Morgan Stanley - JP Morgan ## Bad WLB - Uber - Tower Research - Amazon - Goldman Sachs - Codenation - Zomato - DE Shaw - Sprinklr - Arcesium - Harness - Coinbase ## Unknown WLB (Didn't find much data) - Rippling - Rubrik - Udaan - Sumologic - Cure Fit - Swiggy - Ola - Directi - ServiceNow - Stripe - Sharechat - Postman - Oracle (OCI) - Compass - HealthifyMe - Aviso - Target - Palo Alto Networks - Compass - Vizury - Qubole - Practo - Whatfix - World Quant - Alphonso - App Dynamics - Citrix - Cohesity - MotorQ - Hasura - Quadeye - Bloomreach - Instabase - AirBnB ------------------------------------------------ # Lower paying companies (AFAIK; please let me know if I am wrong) ## Good WLB - Cisco - Samsung - Walmart - Slack - Blackbuck - Oracle (other than OCI) - Visa - Intel - Qualcomm - ARM - Broadcom - Texas Instruments - Mentor Graphics ## Unknown WLB - Cred - Upgrad - Paypal - Dunzo - Unacademy - Oyo - BigBasket - Cloudera - Twilio - Box8 - MyGate - Jio - MakeMyTrip - Dream11 - BrowserStack - Razorpay - Juniper Networks - SanDisk - Redhat - Chowbotics (DoorDash) ## Bad WLB - Hotstar - Paytm - Wissen ------------------------------------------------ # Unknown compensation - Groupon - Gojek - Zoom - Grab - Truecaller - Snowflake (Should be high considering it pays well in US?) ------------------------------------------------ # Visa-sponsors outside US - Facebook London - Yelp London - Booking Amsterdam - Databricks Amsterdam - Spotify Stockholm - Zalando Berlin - ByteDance Singapore - Plenty of other options on Stackoverflow jobs. ------------------------------------------------ # Remote-first companies - Atlassian - Slack - Stackoverflow - Facebook - Datadog - Twitter - Square - Shopify - Github - Gitlab - Auth0 - Coinbase
Is sandisk a good oppurtunity?
Is it good to shift from samsung to sandisk for rtl design role? What chips do sandisk work on -FPGA or ASIC?