At its best, computer architecture is a dialogue between science and engineering—informed both by the possibility of transformative new applications and architectures and grounded in practical implementation and deployment. This track aims to educate the wider computer architecture community on the challenges facing industry and to encourage them to investigate solutions. The session will include a small number of papers selected based on depth and relevance to the HPCA audience. Architects, designers, and developers involved in some aspect of industrial computer system design and development are invited to submit papers describing existing, planned, or canceled products, as well as new challenges, issues, and opportunities in next-generation computer systems. Submitted papers need to provide insights and results that are unique to industry and that make it clearly distinguishable from a regular paper.
Authors should submit an abstract by September 12, 2025. They should submit the full version of the paper by September 19, 2025. No extensions will be granted. The full version should be a PDF file following the submission guidelines that will be made available at the submission website.
Papers should be submitted for double-blind review. The company or product name in question need not be obscured, but no references should be made to individual author identities.
Please take note of the following guidelines for the industry track:
The goal of the HPCA industry track is to publish papers written by industry authors with content related to an industrial product/process. This includes both existing products and planned products, even if the product was cancelled, as long as interesting insights/learned lessons are presented in the paper.
Examples are papers that describe a product, papers that characterize a product, papers that describe some interesting component of a product, papers that provide some insights on challenges faced by industry, etc.
Papers must have the first author and most authors be from industry. Papers may include co-authors from academia, provided that the content is clearly related to an industrial product/process. On the other hand, research papers whose scope is not clearly related to a product are not appropriate for this track and will be rejected from the industry track. Such papers should be submitted to the regular track even if authors are from industry.
Papers will be evaluated with the same standards of quality as papers submitted to the regular track. Papers must be blinded in that they may not reveal individual authorship.
Topics of interest include, but are not limited to:
New Topic: Applications of generative AI for HPC
Domain specific architectures and accelerators
Architectures for cloud, HPC, and data centers
GPUs, their architecture, microarchitecture, and domain adaptations such as Tensor Cores
Algorithm/IP co-design and co-optimization
Processor, cache, and memory architectures
Parallel/Multi-core architectures
Power-efficient architectures
Dependable/secure architectures
Analysis and exploitation of product security vulnerabilities
High-performance I/O systems
Embedded, IoT, reconfigurable, and heterogeneous architectures
Interconnect and network interface architectures
Innovative hardware/software trade-offs
Impact of compilers and system software on architecture
Performance/Power modeling and evaluation
Architectures for emerging technology and applications
At its best, computer architecture is a dialogue between science and engineering—informed both by the possibility of transformative new applications and architectures and grounded in practical implementation and deployment. This track aims to educate the wider computer architecture community on the challenges facing industry and to encourage them to investigate solutions. The session will include a small number of papers selected based on depth and relevance to the HPCA audience. Architects, designers, and developers involved in some aspect of industrial computer system design and development are invited to submit papers describing existing, planned, or canceled products, as well as new challenges, issues, and opportunities in next-generation computer systems. Submitted papers need to provide insights and results that are unique to industry and that make it clearly distinguishable from a regular paper.
Authors should submit an abstract by September 12, 2025. They should submit the full version of the paper by September 19, 2025. No extensions will be granted. The full version should be a PDF file following the submission guidelines that will be made available at the submission website.
Papers should be submitted for double-blind review. The company or product name in question need not be obscured, but no references should be made to individual author identities.
Please take note of the following guidelines for the industry track:
The goal of the HPCA industry track is to publish papers written by industry authors with content related to an industrial product/process. This includes both existing products and planned products, even if the product was cancelled, as long as interesting insights/learned lessons are presented in the paper.
Examples are papers that describe a product, papers that characterize a product, papers that describe some interesting component of a product, papers that provide some insights on challenges faced by industry, etc.
Papers must have the first author and most authors be from industry. Papers may include co-authors from academia, provided that the content is clearly related to an industrial product/process. On the other hand, research papers whose scope is not clearly related to a product are not appropriate for this track and will be rejected from the industry track. Such papers should be submitted to the regular track even if authors are from industry.
Papers will be evaluated with the same standards of quality as papers submitted to the regular track. Papers must be blinded in that they may not reveal individual authorship.
Topics of interest include, but are not limited to:
New Topic: Applications of generative AI for HPC
Domain specific architectures and accelerators
Architectures for cloud, HPC, and data centers
GPUs, their architecture, microarchitecture, and domain adaptations such as Tensor Cores
Algorithm/IP co-design and co-optimization
Processor, cache, and memory architectures
Parallel/Multi-core architectures
Power-efficient architectures
Dependable/secure architectures
Analysis and exploitation of product security vulnerabilities
High-performance I/O systems
Embedded, IoT, reconfigurable, and heterogeneous architectures
Interconnect and network interface architectures
Innovative hardware/software trade-offs
Impact of compilers and system software on architecture
Performance/Power modeling and evaluation
Architectures for emerging technology and applications