7.
 
Re: Morning Mobilization
Mar 12, 2017, 00:09
7.
Re: Morning Mobilization Mar 12, 2017, 00:09
Mar 12, 2017, 00:09
 
Agent-Zero wrote on Mar 11, 2017, 17:38:
Im not a Windows or OS programmer, so I have no idea how true this may be - but I would have suspected the lack of performance of new AMD chips in terms of 1080p gaming and other gaming related tasks would be more likely due to the way games themselves are coded...

It's an interplay between how parallelizable the computation is, i.e. how much work can profitably be done simultaneously, how much of the computation is actually parallelized and how well that parallelization is implemented, i.e. how the game/application is programmed, and how the operating system spreads the parallelized computation, e.g. processes and threads, across the multiple cores, and how the OS schedules those processes and threads that have been tasked to a given core.

If the same program on the same hardware produce different benchmarks for different operating systems, then you are probably seeing the results of different operating system process/thread scheduling algorithms. After all, the purpose of an operating system is to A) provide a clean abstraction of the hardware for programmers to access through an API, and, more apropos to this discussion, B) maximize the utilization of the hardware. In other words, it's an OS responsibility to schedule the CPUs workload to make best use of the hardware.

However, parallel programming is a really difficult and sticky business. Even though we've long had the ability to parallelize computation, first through software abstraction and later with the hardware itself, we still haven't found an effective, generalized approach to parallel programming that allows us to consistently yield substantial performance gains while simultaneously avoiding a set of very subtle and tricky bugs specific to parallel execution, e.g. race conditions, bad data synchronization, and deadlocks. In other words, parallel programming presents a series of unsolved computer science problems for which we only have haphazard solutions. If you ever solve this stuff, you'll definitely earn yourself an ACM Turing prize (the Nobel of computer science).

If I had to guess, the Windows 10 version of the NT kernel is using a different process scheduling algorithm than that of Windows 7, and that while that different algorithm almost certainly yielded gains in some key areas, it has led to performance loss with respect to Ryzen and certain game benchmarks.
Date
Subject
Author
1.
Mar 11, 2017Mar 11 2017
2.
Mar 11, 2017Mar 11 2017
5.
Mar 11, 2017Mar 11 2017
6.
Mar 11, 2017Mar 11 2017
 7.
Mar 12, 2017Mar 12 2017
 Re: Morning Mobilization
8.
Mar 12, 2017Mar 12 2017
9.
Mar 12, 2017Mar 12 2017
10.
Mar 12, 2017Mar 12 2017
11.
Mar 12, 2017Mar 12 2017
14.
Mar 12, 2017Mar 12 2017
15.
Mar 12, 2017Mar 12 2017
16.
Mar 12, 2017Mar 12 2017
17.
Mar 12, 2017Mar 12 2017
18.
Mar 12, 2017Mar 12 2017
19.
Mar 12, 2017Mar 12 2017
20.
Mar 12, 2017Mar 12 2017
        Re: Morning Mobilization
21.
Mar 12, 2017Mar 12 2017
23.
Mar 12, 2017Mar 12 2017
24.
Mar 12, 2017Mar 12 2017
28.
Mar 13, 2017Mar 13 2017
3.
Mar 11, 2017Mar 11 2017
4.
Mar 11, 2017Mar 11 2017
12.
Mar 12, 2017Mar 12 2017
13.
Mar 12, 2017Mar 12 2017
22.
Mar 12, 2017Mar 12 2017
25.
Mar 12, 2017Mar 12 2017
27.
Mar 12, 2017Mar 12 2017
26.
Mar 12, 2017Mar 12 2017