46 lines
2.5 KiB
Org Mode
46 lines
2.5 KiB
Org Mode
|
|
* Introduction plan
|
|
- First paragraph looks good to me.
|
|
- 2nd paragrah also looks fine.
|
|
- 3rd paragraph needs work
|
|
- [ ] Rewrite to mention
|
|
|
|
|
|
* Literature review plan
|
|
|
|
|
|
** Robs feedback
|
|
|
|
"In the dynamic landscape of computing, the pursuit of optimal performance is a constant endeavor" - this is far too wordy and
|
|
grandiose. Reviewers of scientific paper do no like phrases like "landscape of computing" or "constant endeavour". Also "utilization
|
|
of resources is paramount" - there are smaller words than "paramount", you should stick to small words.
|
|
|
|
I suggest you just stick a clear message in the introduction: (1) state the problem, (2) summarise existing approaches for this
|
|
problem (with citations), (3) summarise the limitations of those problems (with citations of papers that demonstrate those limitations),
|
|
(4) "We propose <NAME>, which <DESCRIBE>.", (5) "It overcomes these problems by <EXPLAIN>.". (6) "It performs <STATE RESULTS>.".
|
|
(7) "The contributions of this paper are...".
|
|
|
|
"leading to increased TLB misses and subsequent performance bottlenecks" - this would be better if you could quantify this
|
|
and cite a paper that reports this number. "E.g. 40% increased TLB misses resulting in 25% slowdowns [CITE]".
|
|
|
|
General comment: the "WHY?" you are doing this question is unclear from the introduction. And you need to get to the
|
|
"WHY?" question much earlier, i.e. there is too much verbose discussion huge pages and TLB and CHERI.
|
|
|
|
The *key* "WHY?" paraphrase you've written is:
|
|
|
|
"In this context, the integration of huge pages into memory management strategies alongside capability-based
|
|
addressing in architectures like CHERI offers a compelling synergy. By optimizing TLB utilization through the
|
|
utilization of huge pages and leveraging the security features of capability-based addressing, significant
|
|
performance improvements can be realized."
|
|
|
|
But there are several important problems with this paragraph:
|
|
|
|
1) It comes too late in the introduction, it should be stated much earlier.
|
|
|
|
2) Avoid wordy phrases like "compelling synergy" and long words like "leveraging".
|
|
|
|
3) "significant performance improvements can be realized" is far too non-committal! If the "WHY?" is "performance",
|
|
then this needs to be really (really) clear: state how existing huge pages + TLB approach suffer performance issues;
|
|
state why exactly your technique will outperform existing huge pages + TLB approaches; state what performance metric
|
|
you are referring to; state; then state that they "*are* realised" and give the performance numbers to back up this claim.
|