FluidX3D v3.0 upgrade

Author: ProjectPhysX

DOCUMENTATION.md

@@ -25,7 +25,7 @@
     sudo apt update && sudo apt upgrade -y
     sudo apt install -y g++ git make ocl-icd-libopencl1 ocl-icd-opencl-dev
     mkdir -p ~/amdgpu
-    wget -P ~/amdgpu https://repo.radeon.com/amdgpu-install/6.1.3/ubuntu/jammy/amdgpu-install_6.1.60103-1_all.deb
+    wget -P ~/amdgpu https://repo.radeon.com/amdgpu-install/6.2.3/ubuntu/noble/amdgpu-install_6.2.60203-1_all.deb
     sudo apt install -y ~/amdgpu/amdgpu-install*.deb
     sudo amdgpu-install -y --usecase=graphics,rocm,opencl --opencl=rocr
     sudo usermod -a -G render,video $(whoami)
@@ -60,12 +60,12 @@
 
   - Option 1: Download and install the [oneAPI DPC++ Compiler](https://github.com/intel/llvm/releases?q=oneAPI+DPC%2B%2B+Compiler) and [oneTBB](https://github.com/oneapi-src/oneTBB/releases) with:
     ```bash
-    export OCLV="2024.18.6.0.02_rel"
-    export TBBV="2021.13.0"
+    export OCLV="2024.18.10.0.08_rel"
+    export TBBV="2022.0.0"
     sudo apt update && sudo apt upgrade -y
     sudo apt install -y g++ git make ocl-icd-libopencl1 ocl-icd-opencl-dev
     sudo mkdir -p ~/cpurt /opt/intel/oclcpuexp_${OCLV} /etc/OpenCL/vendors /etc/ld.so.conf.d
-    sudo wget -P ~/cpurt https://github.com/intel/llvm/releases/download/2024-WW25/oclcpuexp-${OCLV}.tar.gz
+    sudo wget -P ~/cpurt https://github.com/intel/llvm/releases/download/2024-WW43/oclcpuexp-${OCLV}.tar.gz
     sudo wget -P ~/cpurt https://github.com/oneapi-src/oneTBB/releases/download/v${TBBV}/oneapi-tbb-${TBBV}-lin.tgz
     sudo tar -zxvf ~/cpurt/oclcpuexp-${OCLV}.tar.gz -C /opt/intel/oclcpuexp_${OCLV}
     sudo tar -zxvf ~/cpurt/oneapi-tbb-${TBBV}-lin.tgz -C /opt/intel

README.md

@@ -1,9 +1,9 @@
 # FluidX3D
 
-The fastest and most memory efficient lattice Boltzmann CFD software, running on all GPUs via [OpenCL](https://github.com/ProjectPhysX/OpenCL-Wrapper "OpenCL-Wrapper"). Free for non-commercial use.
+The fastest and most memory efficient lattice Boltzmann CFD software, running on all GPUs and CPUs via [OpenCL](https://github.com/ProjectPhysX/OpenCL-Wrapper "OpenCL-Wrapper"). Free for non-commercial use.
 
 <a href="https://youtu.be/-MkRBeQkLk8"><img src="https://img.youtube.com/vi/o3TPN142HxM/maxresdefault.jpg" width="50%"></img></a><a href="https://youtu.be/oC6U1M0Fsug"><img src="https://img.youtube.com/vi/oC6U1M0Fsug/maxresdefault.jpg" width="50%"></img></a><br>
-<a href="https://youtu.be/XOfXHgP4jnQ"><img src="https://img.youtube.com/vi/XOfXHgP4jnQ/maxresdefault.jpg" width="50%"></img></a><a href="https://youtu.be/clAqgNtySow"><img src="https://img.youtube.com/vi/clAqgNtySow/maxresdefault.jpg" width="50%"></img></a>
+<a href="https://youtu.be/XOfXHgP4jnQ"><img src="https://img.youtube.com/vi/XOfXHgP4jnQ/maxresdefault.jpg" width="50%"></img></a><a href="https://youtu.be/K5eKxzklXDA"><img src="https://img.youtube.com/vi/K5eKxzklXDA/maxresdefault.jpg" width="50%"></img></a>
 (click on images to show videos on YouTube)
 
 <details><summary>Update History</summary>
@@ -193,6 +193,13 @@ The fastest and most memory efficient lattice Boltzmann CFD software, running on
   - fixed maximum buffer allocation size limit for AMD GPUs and in Intel CPU Runtime for OpenCL
   - fixed wrong `Re<Re_max` info printout for 2D simulations
   - minor fix in `bandwidth_bytes_per_cell_device()`
+- [v3.0](https://github.com/ProjectPhysX/FluidX3D/releases/tag/v3.0) (16.11.2024) [changes](https://github.com/ProjectPhysX/FluidX3D/compare/v2.19...v3.0) (larger CPU/iGPU simulations)
+  - reduced memory footprint on CPUs and iGPU from 72 to 55 Bytes/cell (fused OpenCL host+device buffers for `rho`/`u`/`flags`), allowing 31% higher resolution in the same RAM capacity
+  - faster hardware-supported and faster fallback emulation atomic floating-point addition for `PARTICLES` extension
+  - hardened `calculate_f_eq()` against bad user input for `D2Q9`
+  - fixed velocity voxelization for overlapping geometry with different velocity
+  - fixed Remaining Time printout during paused simulation
+  - fixed CPU/GPU memory printout for CPU/iGPU simulations
 
 </details>
 
@@ -759,6 +766,8 @@ section Orange Pi 5 Mali-G610 MP4
 	232 :active, 0, 232
 section Samsung Mali-G72 MP18 (S9+)
 	230 :active, 0, 230
+section 2x EPYC 9754
+	5179 :crit, 0, 5179
 section 2x EPYC 9654
 	1814 :crit, 0, 1814
 section 2x EPYC 7352
@@ -767,6 +776,12 @@ section 2x EPYC 7313
 	498 :crit, 0, 498
 section 2x EPYC 7302
 	784 :crit, 0, 784
+section 2x 6980P
+	7875 :done, 0, 7875
+section 2x 6979P
+	8135 :done, 0, 8135
+section 2x Platinum 8592+
+	3135 :done, 0, 3135
 section 2x CPU Max 9480
 	2037 :done, 0, 2037
 section 2x Platinum 8480+
@@ -993,10 +1008,14 @@ Colors: 🔴 AMD, 🔵 Intel, 🟢 Nvidia, ⚪ Apple, 🟡 ARM, 🟤 Glenfly
 | 🟡&nbsp;Mali-G610&nbsp;MP4 (Orange&nbsp;Pi&nbsp;5) |             0.06 |          16 |           34 |              130 (58%) |               232 (52%) |                93 (21%) |
 | 🟡&nbsp;Mali-G72&nbsp;MP18 (Samsung&nbsp;S9+)    |               0.24 |           4 |           29 |              110 (59%) |               230 (62%) |                21 ( 6%) |
 |                                                  |                    |             |              |                        |                         |                         |
+| 🔴&nbsp;2x&nbsp;EPYC&nbsp;9754                   |              50.79 |        3072 |          922 |             3276 (54%) |              5077 (42%) |              5179 (43%) |
 | 🔴&nbsp;2x&nbsp;EPYC&nbsp;9654                   |              43.62 |        1536 |          922 |             1381 (23%) |              1814 (15%) |              1801 (15%) |
 | 🔴&nbsp;2x&nbsp;EPYC&nbsp;7352                   |               3.53 |         512 |          410 |              739 (28%) |               106 ( 2%) |               412 ( 8%) |
 | 🔴&nbsp;2x&nbsp;EPYC&nbsp;7313                   |               3.07 |         128 |          410 |              498 (19%) |               367 ( 7%) |               418 ( 8%) |
 | 🔴&nbsp;2x&nbsp;EPYC&nbsp;7302                   |               3.07 |         128 |          410 |              784 (29%) |               336 ( 6%) |               411 ( 8%) |
+| 🔵&nbsp;2x&nbsp;Xeon&nbsp;6980P                  |              98.30 |        6144 |         1690 |             7875 (71%) |              5112 (23%) |              5610 (26%) |
+| 🔵&nbsp;2x&nbsp;Xeon&nbsp;6979P                  |              92.16 |        3072 |         1690 |             8135 (74%) |              4175 (19%) |              4622 (21%) |
+| 🔵&nbsp;2x&nbsp;Xeon&nbsp;Platinum&nbsp;8592+    |              31.13 |        1024 |          717 |             3135 (67%) |              2359 (25%) |              2466 (26%) |
 | 🔵&nbsp;2x&nbsp;Xeon&nbsp;CPU&nbsp;Max&nbsp;9480 |              27.24 |         256 |          614 |             2037 (51%) |              1520 (19%) |              1464 (18%) |
 | 🔵&nbsp;2x&nbsp;Xeon&nbsp;Platinum&nbsp;8480+    |              28.67 |         512 |          614 |             2162 (54%) |              1845 (23%) |              1884 (24%) |
 | 🔵&nbsp;2x&nbsp;Xeon&nbsp;Platinum&nbsp;8380     |              23.55 |        2048 |          410 |             1410 (53%) |              1159 (22%) |              1298 (24%) |

src/info.cpp

@@ -42,7 +42,7 @@ void Info::print_logo() const {
 	print("|                                  ");                 print("\\  \\ /  /", c);                print("                                  |\n");
 	print("|                                   ");                 print("\\  '  /", c);                 print("                                   |\n");
 	print("|                                    ");                 print("\\   /", c);                 print("                                    |\n");
-	print("|                                     ");                 print("\\ /", c);                 print("               FluidX3D Version 2.19 |\n");
+	print("|                                     ");                 print("\\ /", c);                 print("                FluidX3D Version 3.0 |\n");
 	print("|                                      ");                 print( "'", c);                 print("     Copyright (c) Dr. Moritz Lehmann |\n");
 	print("|-----------------------------------------------------------------------------|\n");
 }
@@ -61,8 +61,17 @@ void Info::print_initialize(LBM* lbm) {
 #else // FP32
 	collision += " (FP32/FP32)";
 #endif // FP32
-	cpu_mem_required = (uint)(lbm->get_N()*(ulong)bytes_per_cell_host()/1048576ull); // reset to get valid values for consecutive simulations
-	gpu_mem_required = lbm->lbm_domain[0]->get_device().info.memory_used;
+	bool all_domains_use_ram = true; // reset cpu/gpu_mem_required to get valid values for consecutive simulations
+	for(uint d=0u; d<lbm->get_D(); d++) {
+		all_domains_use_ram = all_domains_use_ram&&lbm->lbm_domain[d]->get_device().info.uses_ram;
+	}
+	if(all_domains_use_ram) {
+		cpu_mem_required = lbm->get_D()*lbm->lbm_domain[0]->get_device().info.memory_used;
+		gpu_mem_required = 0u;
+	} else {
+		cpu_mem_required = (uint)(lbm->get_N()*(ulong)bytes_per_cell_host()/1048576ull);
+		gpu_mem_required = lbm->lbm_domain[0]->get_device().info.memory_used;
+	}
 	const float Re = lbm->get_Re_max();
 	println("|-----------------.-----------------------------------------------------------|");
 	println("| Grid Resolution | "+alignr(57u, to_string(lbm->get_Nx())+" x "+to_string(lbm->get_Ny())+" x "+to_string(lbm->get_Nz())+" = "+to_string(lbm->get_N()))+" |");

src/lbm.cpp

@@ -1259,8 +1259,8 @@ void LBM_Domain::allocate_transfer(Device& device) { // allocate all memory for
 	if(Dy>1u) Amax = max(Amax, (ulong)Nz*(ulong)Nx); // Ay
 	if(Dz>1u) Amax = max(Amax, (ulong)Nx*(ulong)Ny); // Az
 
-	transfer_buffer_p = Memory<char>(device, Amax, max(transfers*(uint)sizeof(fpxx), 17u)); // only allocate one set of transfer buffers in plus/minus directions, for all x/y/z transfers
-	transfer_buffer_m = Memory<char>(device, Amax, max(transfers*(uint)sizeof(fpxx), 17u));
+	transfer_buffer_p = Memory<char>(device, Amax, max(transfers*(uint)sizeof(fpxx), 17u), true, true, 0, false); // only allocate one set of transfer buffers in plus/minus directions, for all x/y/z transfers
+	transfer_buffer_m = Memory<char>(device, Amax, max(transfers*(uint)sizeof(fpxx), 17u), true, true, 0, false); // these transfer buffers must not be zero-copy!
 
 	kernel_transfer[enum_transfer_field::fi              ][0] = Kernel(device, 0u, "transfer_extract_fi"              , 0u, t, transfer_buffer_p, transfer_buffer_m, fi);
 	kernel_transfer[enum_transfer_field::fi              ][1] = Kernel(device, 0u, "transfer__insert_fi"              , 0u, t, transfer_buffer_p, transfer_buffer_m, fi);

src/main.cpp

@@ -77,7 +77,7 @@ void main_label(const double frametime) {
 			draw_label(ox, oy+i, "Steps "          +alignr(31u, /************************************/ alignr(10u, info.lbm->get_t())+" ("+alignr(5, to_uint(1.0/info.runtime_lbm_timestep_smooth))+" Steps/s)"), c); i+=FONT_HEIGHT;
 			draw_label(ox, oy+i, "FPS "            +alignr(33u, /************************************************************/ alignr(4u, to_uint(1.0/frametime))+" ("+alignr(5u, camera.fps_limit)+" fps max)"), c);
 		}
-		draw_label(2, camera.height-1*(FONT_HEIGHT)-1, "FluidX3D v2.19 Copyright (c) Dr. Moritz Lehmann", c);
+		draw_label(2, camera.height-1*(FONT_HEIGHT)-1, "FluidX3D v3.0 Copyright (c) Dr. Moritz Lehmann", c);
 		if(!key_H) {
 			draw_label(camera.width-16*(FONT_WIDTH)-1, 2, "Press H for Help", c);
 		} else {