CmpSci 635 Lecture 24
Slides from the last lecture are below — catching up on a backlog, so two sets.
Slides for lecture 24
Slides for lecture 25
CmpSci 635 Lecture 22
Catching up a bit today, and looking ahead to next time, so two sets of slides.
Slides for lecture 22
Slides for lecture 23
CmpSci 635 Lecture 21
Vectror processing, from Cray to Xeon Phi to GPUs
Slides are here
CmpSci 635 Lecture 20
Shared memory multiprocessing, including directory-based sharing and tradeoffs of interconnect area versus additional cache.
Slides are here
CmpSci 635 Lecture 19
Building on the Metldown approach, Foreshadow exploits abandoned cache state changes associated with accessing contents of the Intel SGX secure enclave....
CmpSci 635 Lecture 18
Security is generally assured at the ISA level in an architecture, and it assumed that the microarchitecture preserves it through a faithful implementation....
CmpSci 635 Lecture 17
If we can achieve a non-volatile solid-state memory that approaches the performance of DRAM, such that we can hide the latency with a small amount of...
CmpSci 635 Lecture 16
This is the first part of our exploration of issues related to secondary storage — why are we shifting from disk to solid state non-volatile memory,...
CmpSci 635 Lecture 15
More on virtualization and virtual memory. We will wrap this up next week.
Slides are here
CmpSci 635 Lecture 14
We’re about a day behind now. Just looking at virtual memory, which will continue with the next clas as we also get into virtualization.
Slides...
CmpSci 635 Lecture 13
We got through the material on pipelining today, and will not get to the material in the slides below until next Thursday, after the midterm.
Remember...
CmpSci 635 Lecture 12
We caught up a bit today, getting through Prefetching, Multithreading, and Predication. Next time we will cover pipelining, which is covered in the slides...
CmpSci 635 Lecture 11
Still running behind. We took a look at the structure of a modern processor today as background for how prefetching and multithreading fit into an overall...
CmpSci 635 Lecture 10
We’re falling behind as part of filling in the background for some of the topics. The slides today will be covered next time. We focused today mainly...
CmpSci 635 Lecture 9
We wrapped up our exploration of transactional memory and started a refresher on the architecture of cache memory. We did not get to the papers for today,...
CmpSci 635 Lecture 8
We went back to consider a bit more of the context for the memory wall paper, then considered the supercomputer memory study, with some explanation of...
CmpSci 635 Lecture 7
We wrapped up our exploration of wearout issues and took a look at PCM before we began reviewing issues associated with memory, and just got started...
CmpSci 635 Lecture 6
We are shifting toward looking at factors that affect the design choices of architects, and that are driving some changes in how architectrues are built....
CmpSci 635 Lecture 5
We’ve been looking at the challenges of using good methodology in doing architecture research, which can include benchmark selection, simulator validation,...
CmpSci 635 Lecture 4
Today we went back and discussed the prior papers a bit more, then looked at the ones by Lizzy John (about subsetting SPEC2006) and Xi Yang (about the...
CmpSci 635 Lecture 3
We took a look at the problem of benchmark engineering, then at the issues of simulator accuracy and how commercial workloads are not well represented...
CmpSci 635 Lecture 2
Today we had an overview of the history of architecture, instruction set architecture, microarchitecture, the diversification of architectures, and how...
CmpSci 635 Lecture 1
Today we went over the course syllabus, talked about how the readings will be done, began looking at the paper for next time, and took a quick tour...