Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commit65128bf

Browse files
authored
Update index.html
1 parent85e5966 commit65128bf

File tree

1 file changed

+112
-12
lines changed

1 file changed

+112
-12
lines changed

‎Tools/GPU-Occupancy-Calculator/index.html

Lines changed: 112 additions & 12 deletions
Original file line numberDiff line numberDiff line change
@@ -694,6 +694,108 @@ <h2>Intel® GPU Occupancy Calculator</h2>
694694
"Max_Num_Of_Barrier_Registers":32
695695
}
696696
},
697+
//RPL-P
698+
{
699+
"pci_id":["A7A1","A7A0","A721","A720"],
700+
"name":"Integrated GPU (Xe LP)",
701+
"product_name":"Intel® Iris® Xe Max Graphics",
702+
"code":"gen12",
703+
"device_info":{
704+
"EU_Per_Sub_Slice":16,
705+
"Threads_Per_EU":7,
706+
"EU_Count":96,
707+
"Max_Threads_Per_Sub_Slice":112,
708+
"Large_GRF_Mode":false,
709+
"Subgroup_Sizes":[32,16,8],
710+
"SLM_Size_Per_Sub_Slice":64,
711+
"SLM_Size_Per_Work_Group":64,
712+
"TG_SLM_Sizes":[0,1,2,4,8,16,32,64],
713+
"Max_Work_Group_Size":512,
714+
"Max_Num_Of_Workgroups":112,
715+
"Max_Num_Of_Barrier_Registers":32
716+
}
717+
},
718+
{
719+
"pci_id":["A7A9","A7A8"],
720+
"name":"Integrated GPU (Xe LP)",
721+
"product_name":"Intel® Iris® Xe Max Graphics",
722+
"code":"gen12",
723+
"device_info":{
724+
"EU_Per_Sub_Slice":16,
725+
"Threads_Per_EU":7,
726+
"EU_Count":64,
727+
"Max_Threads_Per_Sub_Slice":112,
728+
"Large_GRF_Mode":false,
729+
"Subgroup_Sizes":[32,16,8],
730+
"SLM_Size_Per_Sub_Slice":64,
731+
"SLM_Size_Per_Work_Group":64,
732+
"TG_SLM_Sizes":[0,1,2,4,8,16,32,64],
733+
"Max_Work_Group_Size":512,
734+
"Max_Num_Of_Workgroups":112,
735+
"Max_Num_Of_Barrier_Registers":32
736+
}
737+
},
738+
//RPL-S
739+
{
740+
"pci_id":["A780","A781","A788","A789"],
741+
"name":"Integrated GPU (Xe LP)",
742+
"product_name":"Intel® UHD Graphics",
743+
"code":"gen12",
744+
"device_info":{
745+
"EU_Per_Sub_Slice":16,
746+
"Threads_Per_EU":7,
747+
"EU_Count":32,
748+
"Max_Threads_Per_Sub_Slice":112,
749+
"Large_GRF_Mode":false,
750+
"Subgroup_Sizes":[32,16,8],
751+
"SLM_Size_Per_Sub_Slice":64,
752+
"SLM_Size_Per_Work_Group":64,
753+
"TG_SLM_Sizes":[0,1,2,4,8,16,32,64],
754+
"Max_Work_Group_Size":512,
755+
"Max_Num_Of_Workgroups":112,
756+
"Max_Num_Of_Barrier_Registers":32
757+
}
758+
},
759+
{
760+
"pci_id":["A782","A78A"],
761+
"name":"Integrated GPU (Xe LP)",
762+
"product_name":"Intel® UHD Graphics",
763+
"code":"gen12",
764+
"device_info":{
765+
"EU_Per_Sub_Slice":12,
766+
"Threads_Per_EU":7,
767+
"EU_Count":24,
768+
"Max_Threads_Per_Sub_Slice":112,
769+
"Large_GRF_Mode":false,
770+
"Subgroup_Sizes":[32,16,8],
771+
"SLM_Size_Per_Sub_Slice":64,
772+
"SLM_Size_Per_Work_Group":64,
773+
"TG_SLM_Sizes":[0,1,2,4,8,16,32,64],
774+
"Max_Work_Group_Size":512,
775+
"Max_Num_Of_Workgroups":112,
776+
"Max_Num_Of_Barrier_Registers":32
777+
}
778+
},
779+
{
780+
"pci_id":["A783","A78B"],
781+
"name":"Integrated GPU (Xe LP)",
782+
"product_name":"Intel® UHD Graphics",
783+
"code":"gen12",
784+
"device_info":{
785+
"EU_Per_Sub_Slice":16,
786+
"Threads_Per_EU":7,
787+
"EU_Count":16,
788+
"Max_Threads_Per_Sub_Slice":112,
789+
"Large_GRF_Mode":false,
790+
"Subgroup_Sizes":[32,16,8],
791+
"SLM_Size_Per_Sub_Slice":64,
792+
"SLM_Size_Per_Work_Group":64,
793+
"TG_SLM_Sizes":[0,1,2,4,8,16,32,64],
794+
"Max_Work_Group_Size":512,
795+
"Max_Num_Of_Workgroups":112,
796+
"Max_Num_Of_Barrier_Registers":32
797+
}
798+
},
697799
//ARC
698800
{
699801
"pci_id":["56A5","5694"],
@@ -1256,23 +1358,21 @@ <h2>Intel® GPU Occupancy Calculator</h2>
12561358
}
12571359

12581360
// Calculate GPU Occupancy
1361+
12591362
functioncompute_gpu_occupancy(wg,num_wg,ss_occ,num_ss,global_range){
12601363
vargpu_occ;
12611364
// Calculate max num of work-items in all ss of gpu
12621365
varnum_wi=num_ss*num_wg*wg;
12631366
// Calculate occupancy for all ss in gpu
1264-
if(global_range>num_wi){
1265-
varnum_pass=parseInt(global_range/num_wi);
1266-
gpu_occ=ss_occ*num_pass;
1267-
if(global_range%num_wi!==0){
1268-
num_pass+=1;
1269-
}
1270-
gpu_occ=gpu_occ/num_pass;
1271-
return{gpu_occ, ss_occ}
1272-
}else{
1273-
gpu_occ=(global_range/num_wi)*ss_occ;
1274-
return{gpu_occ, ss_occ}
1275-
}
1367+
varnum_pass=parseInt(global_range/num_wi);
1368+
vargpu_pass=ss_occ*num_pass;
1369+
varnum_wi_left=global_range%num_wi;
1370+
if(num_wi_left!==0){
1371+
gpu_pass+=(num_wi_left/num_wi)*ss_occ;
1372+
num_pass+=1;
1373+
}
1374+
gpu_occ=gpu_pass/num_pass;
1375+
return{gpu_occ, ss_occ}
12761376
}
12771377

12781378
// Generate Graphs and Optimal Occupancy Config table

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp