WACI (Wild and Crazy Ideas) Session

Title: Computing Foundations: Knitting Together a WACI Purl
Abstract: The WACI community lives at the bleeding edge by charting a course for the hardware, programming languages, compilers, and design tools that could support  more unconventional possibilities for computing. However, to best shape the future, we need to deeply understand foundations going all the way back to the first programmable devices — looms for textile production. This talk will take a quick journey through the entire computing stack, tying together disparate threads to weave a narrative that loops through the past, present, and future of computing. We present two core arguments: (1) everything is a program and (2) we should be able to program everything.
We illustrate these arguments by exploring the frontier of hardware programming, which has yet to fully exploit the potential of unconventional hardware devices for physical fabrication. As a case study, we will consider whether modern industrial knitting machines should be programmed with “one more or one fewer dimensions” and how the recent trend of scheduling languages could unlock new forms of creativity and efficiency in textiles. We will tie up loose ends by sketching a vision for the future of computing that embraces the full spectrum of hardware devices and call for WACI to continue leading the charge by putting the field on even firmer foundations.
Zachary Tatlock (University of Washington)
Title: BoltBleed – Security from an Aerial Perspective
Abstract: You may think the data on your phone is safe when it falls out of a Boeing 737 going at a cruising speed of 525 miles per hour. However, what you didn’t anticipate is the malicious hackers on the ground or below the plane waiting for that phone. This is BoltBleed, a dangerous security vulnerability in an aerial age. You may think your phone falling out of the door plug of a passenger plane is an absurd hypothetical, but this exact situation occurred on and below Alaska Airlines flight 1281. We propose a solution to this vulnerability with both a detection (with sample microarchitectural implementations) and tolerance mechanism. We further examine the power constraints of such a system and what further research can be done to better realize it.
Guy Wilks (UC Santa Barbara)
Title: Apparate?: Evading Memory Hierarchy with GodSpeed Wireless on Chip
Abstract: The rapid advancements in memory systems, CPU technology, and emerging technologies herald a transformative potential in computing, promising to revolutionize memory hierarchies. Innovations in DDR memory are delivering unprecedented bandwidth, while advancements in on-chip wireless technology are reducing size and increasing speed. The introduction of godspeed wireless transceivers on chip, alongside near high-speed DRAM, is poised to directly facilitate memory requests. This integration suggests the potential for eliminating traditional memory hierarchies, offering a new paradigm in computing efficiency and speed. These developments indicate a near-future where computing systems are significantly more responsive and powerful, leveraging direct, high-speed memory access mechanisms.
Nitesh Narayana Gondlyala Sathya and Abhijit Das (Universitat Politècnica de Catalunya Barcelona)
Title: Bureaucracy in Systems
Abstract: Managing complexity in a systems architecture is often regarded as a point on a spectrum between a monolithic system and a microkernel-based system. On a monolith, all complexity and functionality is collected in a few components or — in the extreme case — in a single one. In a microkernel-based system, components are small and have reduced purpose, simplifying the individual components. However, we argue that complexity was not reduced but merely moved from a single component with inherent complexity to the complex interactions between many small components. In order to make a fair argument, we must not only consider the size of components, but also the complexity manifested in the interfaces between them. We therefore propose to learn from bureaucratic process and apply the experienced bureaucracy of component interactions as a novel complexity metric we call Trans-Compartmental Bureaucracy (TCB). We show how this metric leads to different characterizations of system architectures compared to existing complexity metrics.
Michael Roitzsch (Barkhausen Institut)
Title: BRAINSTORM: Supercharging Innovation with AI-Driven Ideation
Abstract: The Wild and Crazy Ideas (WACI) session at ASPLOS thrives on out-of-the-box, audacious ideas that push the boundaries of research. Generating such ideas can be challenging and time-consuming. This paper introduces BRAINSTORM, a self-contained AI system designed to automatically generate WACI-worthy research ideas and papers. BRAINSTORM utilizes advanced natural language processing and generation techniques, combined with knowledge of existing research and trends, to synthesize novel and unconventional concepts within the ASPLOS domain. Beyond idea generation, BRAINSTORM can also assist researchers in the brainstorming process, providing inspiration and prompting creative thinking. This paper explores the capabilities and potential impact of BRAINSTORM, while also addressing the ethical considerations of utilizing AI in the research process.
Deniz Altınbüken, Martin Maas, Phitchaya Mangpo Phothilimthana (Google DeepMind)