weft - CPEN 511 Project - Weft: Improving SIMD Utilization through MIMD-like Co-issue and Thread Compaction

Age	Commit message (Collapse)	Author
2018-06-29	Allowing modern GPU configrations to properly run PTXPLUS.	tgrogers
	There are several problems with this: 1) To get proper occupancy, based on register file usage, you must querry ptxas with the sm version that corresponds to the register usage in the config file. To enable this, a new config option has been added that determines what sm version you pass to ptxas to compute reg-usage. This configuration option is always required in the gpgpusim.config file 2) If you are running PTXPLUS with a modern card (i.e. volta/pascal), you need ptxas from CUDA 9.1. However since PTXPLUS only supports sm_13 - you need a version of CUDA where cudaobjectdump supports sm_1x. This ended at CUDA 5 - and PTXPLUS requires CUDA 4.2. Therefore, to run PTXPLUS on a modern card, you need CUDA 4.2 + modern CUDA installed. To fascilitate this, a new envronment varaible is added and the setup envrionment script prints an appropraite warning if you are using a newer CUDA. We have tried to make this as fail-proof as possible - and die appropraitely when something is wrong.
2016-06-03	Added support for %laneid SFR. Also added a notice clarifying that power ↵	speverel
	modeling for GTX750Ti is currently completely untested and should not be considered supported.
2016-06-02	Updated parser and config file to support compute versions up to 5.2. Full ↵	speverel
	support is NOT claimed; however, it has been tested to work on a number of CUDA version 7.5 benchmarks such as matrix multiply and simpleMultiGPU.
2016-06-02	Added handling of .cc option for arithmetic instructions. NOTE: Only made ↵	speverel
	changes to parse instructions. Carry functionality NOT fully implemented; .cc instructions function like their unmodified ueqivelents. Also modified GTX750Ti config to model L1 data cache as simply not being used for global loads (instead of not existing at all). Changed ptxinfo parsing to avoid crashing when info includes texture information.
2016-06-02	Made additional improvements to Maxwell correlation in config file such as ↵	Scott Peverelle
	merger of L1/Texture cache. [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 21837]
2016-06-02	Tweaked Maxwell config to improve correspondence in a number of areas; clock ↵	Scott Peverelle
	speed, DRAM bus width, L2 cache size etc. Adjusted parse_and_compare to match new clock frequency as per config file. [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 21818]
2016-06-02	Modified config files to better represent Maxwell architecture; ↵	Scott Peverelle
	specifically, changed the number of clusters, controllers, and interconnect nodes. Also modified parse_and_compare to fix an occasional division by zero bug. [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 21812]
2016-06-02	GeForceGTX750Ti correlation setup. Modified config file to reflect GTX750Ti ↵	Scott Peverelle
	architecture. Modified shader.h to allow for larger CTA per warp, to accomodate Maxwell specs. [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 21798]
2016-06-02	Branch of GTX480 configs for Maxwell board	Scott Peverelle
	[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 21797]