Fall 2026
Performance Tools Intern
Etched
201-500 employees
Generates open-world, interactive game environments
No salary listed
San Jose, CA, USA
In Person
Fully in-person in San Jose.
 Software Engineering (1) |
|---|
Get referred to Etched
See people who can refer or advise you
- Strong programming skills in C++ or Rust.
- A solid understanding of computer architecture, including CPUs, GPUs or artificial intelligence accelerators, memory hierarchies, and parallel programming.
- Experience or strong interest in low-level performance analysis, profiling, and performance optimization.
- Experience or strong interest in operating systems, compilers, firmware, drivers, or other low-level systems software.
- Strong problem-solving skills and curiosity to learn quickly in a fast-paced engineering environment.
- Build components of the performance analysis and profiling infrastructure.
- Collect and analyze performance data from custom machine-learning accelerators, including hardware counters, execution traces, and memory behavior.
- Develop tooling to trace host-side runtime activity, system behavior, and accelerator execution.
- Help correlate performance events across central processing units, accelerators, storage, networking, and distributed workloads.
- Build analysis and visualization tools that help engineers identify performance bottlenecks and optimize models.
- Work alongside hardware, compiler, firmware, and inference engineers to understand performance challenges and develop tools that improve developer productivity.
- Implement the data collection framework for hardware performance counters on a custom PCI Express-based accelerator.
- Develop a user-space service for low-overhead tracing of accelerator activity.
- Design and build a correlated timeline view visualizing central processing unit application programming interface calls, driver submissions, PCI Express transfers, and accelerator execution units.
- Create an analysis pass to detect and quantify memory access inefficiencies or PCI Express bandwidth saturation while transacting on a PCI Express-attached accelerator.
- Experience with Python.
- Familiarity with performance analysis tools such as Nsight, VTune, Xprof, Perfetto, or similar tools.
- Direct experience developing performance analysis or debugging tools.
- Experience with machine-learning accelerator architectures such as GPUs or TPUs.
- Experience with kernel-mode driver development for Linux or Windows.
Etched builds Oasis, an AI model that creates playable open-world video games. Oasis generates entire game worlds with interactive, explorable environments, enabling developers to design expansive experiences rather than just producing non-interactive media. Unlike other AI video models that yield only prompts-to-video outputs, Oasis focuses on dynamic game-generation capabilities that deliver a playable world. The company differentiates itself by offering a toolset specifically for game development, licensing technology to studios or providing a platform to create new games, rather than generic AI outputs. The goal is to empower game developers to quickly craft large, living game worlds by using AI-driven world-generation technology.
Company Size
201-500
Company Stage
Series D
Total Funding
$1.6B
Headquarters
San Jose, California
Founded
2022
Get referred to Etched
See people who can refer or advise you
Simplify's Take
What believers are saying
- On August 18, 2026, Etched raised $700 million at a $21 billion valuation.
- Etched says it secured over $1 billion in customer contracts before broad shipment.
- The company shipped its first rack to Jane Street and is ramping production.
What critics are saying
- Nvidia keeps shipping faster GPUs; Etched must prove Sohu beats Blackwell economics by 2027.
- Three-year-old hardware startups miss yield targets; any TSMC delay breaks Etched's 2026 rollout.
- A bad first-rack failure at Jane Street kills Etched's inference-hardware credibility overnight.
What makes Etched unique
- Etched co-designs chips, racks, software, and manufacturing for inference, not general compute.
- Jane Street uses Etched's first rack in production, validating latency for trading workloads.
- Oasis showed Etched can pair hardware with interactive AI experiences, not just benchmarks.
Help us improve and share your feedback! Did you find this helpful?
Benefits
Health Insurance
Dental Insurance
Vision Insurance
Unlimited Paid Time Off
Flexible Work Hours
Remote Work Options
Paid Vacation
Paid Sick Leave
Paid Holidays
Hybrid Work Options
Stock Options
Company Equity
401(k) Retirement Plan
401(k) Company Match
Performance Bonus
Profit Sharing
Employee Stock Purchase Plan
Relocation Assistance
Employee Referral Bonus
Parental Leave
Family Planning Benefits
Fertility Treatment Support
Adoption Assistance
Childcare Support
Elder Care Support
Wellness Program
Mental Health Support
Gym Membership
Phone/Internet Stipend
Home Office Stipend
Conference Attendance Budget
Professional Development Budget
Growth & Insights and Company News
6 month growth
↓ -5%1 year growth
↓ -10%2 year growth
↑ 1%Investor Michael Burry has identified chip startup Etched as serious competition for Nvidia. The company, founded in 2022 by three 24-year-old Harvard dropouts, recently raised $700 million in a funding round led by Jane Street, valuing it at $21 billion. Etched has now raised nearly $2 billion total. The startup focuses on AI inference chips and claims it took just 44 days to run inference workloads after receiving test chips from Taiwan Semiconductor Manufacturing, a process typically taking six months or more. Roughly 15% of Etched's 400 employees previously worked at Nvidia, including 23-year veteran Brian Loiler. Burry, known for his bearish stance on Nvidia, highlighted Etched's rapid development in a post on X, citing a Wall Street Journal report on the startup's progress.
Etched, founded by Harvard dropouts, has its own in-office data center and has signed quant-trading firm Jane Street as its first customer.
Etched, a hardware startup founded by Harvard students in 2022, has raised $300 million in a Series C round at a $10 billion pre-money valuation. Primary Venture Partners led the investment, with participation from Jane Street, Andreessen Horowitz, Diffusion, and SK Hynix. The company builds inference-specific chips and cluster-scale systems for AI models. Etched earlier this year taped out its first generation chip at TSMC, becoming the first post-ChatGPT-era company with a successful full-reticle A0 chip tape-out on TSMC's leading-edge nodes. The company's technology focuses on improving throughput and latency through innovations including low voltage inference and cluster-scale memory. Etched plans to begin shipping its products in 2026, with early customers already testing the system.
Etched, an AI chip startup founded by three Harvard dropouts in 2022, has raised $300 million in Series C funding at a $10.3 billion valuation. The round was led by Sequoia, with participation from Andreessen Horowitz, SK Hynix, Jane Street, and Diffusion Capital. The company has doubled its valuation from $5 billion in December. Last month, Etched announced it had successfully manufactured its chips and booked $1 billion in orders. Etched designs systems specifically for AI inference, creating two new components: a prefill chip using low-voltage technology and a cluster scale memory system. The company aims to deliver faster speeds at lower costs. Despite initial skepticism, Etched has attracted prominent investors including Peter Thiel, Andrej Karpathy, and Geoffrey Hinton through private demos. The startup now employs 400 people and operates a 2 megawatt data centre.
AI chip startup Etched has raised $800 million, with investors including Jane Street and a venture firm linked to Taiwan Semiconductor Manufacturing Co. The company, which aims to compete with Nvidia, plans to begin shipping chips to select customers this summer. The funding marks a significant milestone for Etched as it seeks to establish itself in the competitive AI chip market currently dominated by Nvidia.