Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
seminars [2025-03-19] – [Accelerating Operator Calculus] Martin Zieglerseminars [2025-08-13] (current) – [XpLUTo: Modelling Bulk Parallel Processing in RAM via Lookup Tables] Martin Ziegler
Line 14: Line 14:
  
 ===== 2025 ===== ===== 2025 =====
 +
 +====XpLUTo: Modelling Bulk Parallel Processing in RAM via Lookup Tables====
 +  * August 12, 4pm KST
 +  * E3-1 #4420 and online
 +  * Nguyên Trần Bảo (HCMUT and KAIST)
 +
 +Processing-in-memory (PIM) has been investigated for its ability to
 +perform bulk data operations while eliminating data movement, which is a major
 +performance bottleneck. However, existing designs, regardless of how minimal,
 +still require modifications to the physical memory circuitry. Moreover, each
 +proposed operation introduces different primitives, inherently hindering the development
 +of a general design capable of supporting all operations. In this work,
 +we propose XPLUTO, a new parallel architecture model that leverages the capabilities
 +of PIM. Our key observation is that, in the worst case, any complex
 +operation can be implemented via a lookup table (precomputation and query),
 +which can be viewed as a SIMD (single-instruction multiple-data) operation.
 +Based on this insight, we focus on designing algorithms built upon SIMD operations,
 +with asymptotic costs estimated according to lookup table performance.
 +So far, XPLUTO has demonstrated the ability to emulate various problems,
 +including sorting, addition, and prefix operations.
 +
  
 ====Integer factorization of matrices and 4-dimensional genera of knots==== ====Integer factorization of matrices and 4-dimensional genera of knots====