{"schemaVersion":"jobsearcher.job.v1","id":"07f300bb5c079596d9ed3abf","url":"https://jobsearcher.com/jobs/07f300bb5c079596d9ed3abf","canonicalUrl":"https://jobsearcher.com/jobs/07f300bb5c079596d9ed3abf","title":"Software Engineer, ML Performance","description":"About Etched\r\nEtched is building AI chips that are hard-coded for individual model architectures. Our first product (Sohu) only supports transformers, but has an order of magnitude more throughput and lower latency than a B200. With Etched ASICs, you can build products that would be impossible with GPUs, like real-time video generation models and extremely deep chain-of-thought reasoning.\r\nSoftware Engineer, ML Performance\r\nRunning millions of tokens per second for large models (e.g Llama-3-70B) means running into new performance bottlenecks. Even with hardware optimization for the operations that usually bottleneck us (attention, kernel parallelism), we encounter novel bottlenecks and must design our own solutions to solve them.\r\nYou will work closely with our hardware and software teams to identify and mitigate performance bottlenecks, enabling our chips to achieve unprecedented throughput and efficiency. Your work will involve a blend of low-level programming, performance profiling, and hands-on debugging, all aimed at maximizing the performance of our custom-built AI hardware.\r\nYou will also play a key role in developing tools and methodologies to help our customers understand the full potential of our hardware.\r\nRepresentative projects\r\nWriting new kernels to improve throughput for LLM embedding\r\nImproving on PagedAttention to prevent fragmentation of the KV cache in memory\r\nDebugging hardware issues on a simulated or emulated chip\r\nProfile transformers running on our hardware, and fix bottlenecks\r\nDevelop ways for customers to work with our chip and understand how their workloads will run on it.\r\nYou may be a good fit if you\r\nHave 5+ years of low-level programming experience\r\nHave a strong understanding of data flow and execution paths within embedded systems\r\nPick up slack, even if it goes outside your job description\r\nAre results-oriented, and bias towards shipping products\r\nUnderstand SoC and computer system architecture, especially for CPU, interconnect, and memory subsystems\r\nWant to learn more about machine learning research\r\nWe encourage you to apply even if you do not believe you meet every single qualification.\r\nStrong candidates may also have experience with\r\nGPU kernel profiling and low-level programming\r\nTransformer optimizations, such as FlashAttention\r\nOngoing research in machine learning\r\nPalladium emulation\r\nHow we're different:\r\nEtched believes in the Bitter Lesson. We think most of the progress in the AI field has come from using more FLOPs to train and run models, and the best way to get more FLOPs is to build model-specific hardware. Larger and larger training runs encourage companies to consolidate around fewer model architectures, which creates a market for single-model ASICs.\r\nWe are a fully in-person team in Cupertino, and greatly value engineering skills. We do not have boundaries between engineering and research, and we expect all of our technical staff to contribute to both as needed.\r\nBenefits\r\nFull medical, dental, and vision packages, with 100% of premium covered, 90% for dependents\r\nHousing subsidy of $2,000/month for those living within walking distance of the office\r\nDaily lunch and dinner in our office\r\nRelocation support for those moving to Cupertino\r\nJ-18808-Ljbffr","company":"Openreq","rawCompany":"openreq","city":"Cupertino","state":"CA","isRemote":false,"isActive":true,"createdAt":"2026-06-25T01:21:43.099Z","occupations":[{"code":"15-1252.00","title":"Software Developers","slug":"software-developers"},{"code":"15-1299.08","title":"Computer Systems Engineers/Architects","slug":"computer-systems-engineers-architects"},{"code":"17-2061.00","title":"Computer Hardware Engineers","slug":"computer-hardware-engineers"}],"industries":[{"code":"541511","title":"Custom Computer Programming Services","slug":"custom-computer-programming-services"},{"code":"513210","title":"Software Publishers","slug":"software-publishers"},{"code":"334111","title":"Electronic Computer Manufacturing","slug":"electronic-computer-manufacturing"}],"jobPosting":{"@context":"https://schema.org","@type":"JobPosting","title":"Software Engineer, ML Performance","description":"About Etched\r\nEtched is building AI chips that are hard-coded for individual model architectures. Our first product (Sohu) only supports transformers, but has an order of magnitude more throughput and lower latency than a B200. With Etched ASICs, you can build products that would be impossible with GPUs, like real-time video generation models and extremely deep chain-of-thought reasoning.\r\nSoftware Engineer, ML Performance\r\nRunning millions of tokens per second for large models (e.g Llama-3-70B) means running into new performance bottlenecks. Even with hardware optimization for the operations that usually bottleneck us (attention, kernel parallelism), we encounter novel bottlenecks and must design our own solutions to solve them.\r\nYou will work closely with our hardware and software teams to identify and mitigate performance bottlenecks, enabling our chips to achieve unprecedented throughput and efficiency. Your work will involve a blend of low-level programming, performance profiling, and hands-on debugging, all aimed at maximizing the performance of our custom-built AI hardware.\r\nYou will also play a key role in developing tools and methodologies to help our customers understand the full potential of our hardware.\r\nRepresentative projects\r\nWriting new kernels to improve throughput for LLM embedding\r\nImproving on PagedAttention to prevent fragmentation of the KV cache in memory\r\nDebugging hardware issues on a simulated or emulated chip\r\nProfile transformers running on our hardware, and fix bottlenecks\r\nDevelop ways for customers to work with our chip and understand how their workloads will run on it.\r\nYou may be a good fit if you\r\nHave 5+ years of low-level programming experience\r\nHave a strong understanding of data flow and execution paths within embedded systems\r\nPick up slack, even if it goes outside your job description\r\nAre results-oriented, and bias towards shipping products\r\nUnderstand SoC and computer system architecture, especially for CPU, interconnect, and memory subsystems\r\nWant to learn more about machine learning research\r\nWe encourage you to apply even if you do not believe you meet every single qualification.\r\nStrong candidates may also have experience with\r\nGPU kernel profiling and low-level programming\r\nTransformer optimizations, such as FlashAttention\r\nOngoing research in machine learning\r\nPalladium emulation\r\nHow we're different:\r\nEtched believes in the Bitter Lesson. We think most of the progress in the AI field has come from using more FLOPs to train and run models, and the best way to get more FLOPs is to build model-specific hardware. Larger and larger training runs encourage companies to consolidate around fewer model architectures, which creates a market for single-model ASICs.\r\nWe are a fully in-person team in Cupertino, and greatly value engineering skills. We do not have boundaries between engineering and research, and we expect all of our technical staff to contribute to both as needed.\r\nBenefits\r\nFull medical, dental, and vision packages, with 100% of premium covered, 90% for dependents\r\nHousing subsidy of $2,000/month for those living within walking distance of the office\r\nDaily lunch and dinner in our office\r\nRelocation support for those moving to Cupertino\r\nJ-18808-Ljbffr","datePosted":"2026-06-25T01:21:43.099Z","dateModified":"2026-06-25T01:21:43.099Z","hiringOrganization":{"@type":"Organization","name":"Openreq","sameAs":"https://jobsearcher.com"},"jobLocation":{"@type":"Place","address":{"@type":"PostalAddress","addressLocality":"Cupertino","addressRegion":"CA","addressCountry":"US"}},"identifier":{"@type":"PropertyValue","name":"JobSearcher","value":"07f300bb5c079596d9ed3abf"},"url":"https://jobsearcher.com/jobs/07f300bb5c079596d9ed3abf"}}