Skip to content

GitLab

Commit 91f3379b authored by Peter Eastman's avatar Peter Eastman
Browse files

Minor optimizations to GBSA

parent d2a5b3bb
Hide whitespace changes
Inline Side-by-side
Showing with 25 additions and 25 deletions
platforms/opencl/src/kernels/gbsaObc2.cl
View file @ 91f3379b
{
float invRSquared = RECIP(r2);
float invRSquaredOver4 = 0.25f*invR*invR;
float rScaledRadiusJ = r+obcParams2.y;
float rScaledRadiusI = r+obcParams1.y;
float l_ijJ = RECIP(max(obcParams1.x, fabs(r-obcParams2.y)));
......@@ -14,12 +14,8 @@
float t1I = LOG(u_ijI*RECIP(l_ijI));
float t2J = (l_ij2J-u_ij2J);
float t2I = (l_ij2I-u_ij2I);
float t3J = t2J*invR;
float t3I = t2I*invR;
t1J *= invR;
t1I *= invR;
float term1 = 0.125f*(1.0f+obcParams2.y*obcParams2.y*invRSquared)*t3J + 0.25f*t1J*invRSquared;
float term2 = 0.125f*(1.0f+obcParams1.y*obcParams1.y*invRSquared)*t3I + 0.25f*t1I*invRSquared;
float term1 = (0.5f*(0.25f+obcParams2.y*obcParams2.y*invRSquaredOver4)*t2J + t1J*invRSquaredOver4)*invR;
float term2 = (0.5f*(0.25f+obcParams1.y*obcParams1.y*invRSquaredOver4)*t2I + t1I*invRSquaredOver4)*invR;
float tempdEdR = select(0.0f, bornForce1*term1, obcParams1.x < rScaledRadiusJ);
tempdEdR += select(0.0f, bornForce2*term2, obcParams2.x < rScaledRadiusJ);
#ifdef USE_CUTOFF
......
platforms/opencl/src/kernels/gbsaObc_nvidia.cl
View file @ 91f3379b
......@@ -7,12 +7,9 @@
typedef struct {
float x, y, z;
float q;
float fx, fy, fz, fw;
float radius, scaledRadius;
float bornSum;
float bornRadius;
float bornForce;
} AtomData;
} AtomData1;
/**
* Compute the Born sum.
......@@ -24,7 +21,7 @@ __kernel void computeBornSum(
__global float* global_bornSum,
#endif
__global float4* posq, __global float2* global_params,
__local AtomData* localData, __local float* tempBuffer,
__local AtomData1* localData, __local float* tempBuffer,
#ifdef USE_CUTOFF
__global ushort2* tiles, __global unsigned int* interactionCount, float4 periodicBoxSize, float4 invPeriodicBoxSize, unsigned int maxTiles, __global unsigned int* interactionFlags) {
#else
......@@ -104,8 +101,8 @@ __kernel void computeBornSum(
float l_ij2 = l_ij*l_ij;
float u_ij2 = u_ij*u_ij;
float ratio = LOG(u_ij * RECIP(l_ij));
bornSum += l_ij - u_ij + 0.25f*r*(u_ij2-l_ij2) + (0.50f*invR*ratio) +
(0.25f*params2.y*params2.y*invR)*(l_ij2-u_ij2);
bornSum += l_ij - u_ij + (0.50f*invR*ratio) + 0.25f*(r*(u_ij2-l_ij2) +
(params2.y*params2.y*invR)*(l_ij2-u_ij2));
if (params1.x < params2.x-r)
bornSum += 2.0f*(RECIP(params1.x)-l_ij);
}
......@@ -161,8 +158,8 @@ __kernel void computeBornSum(
float l_ij2 = l_ij*l_ij;
float u_ij2 = u_ij*u_ij;
float ratio = LOG(u_ij * RECIP(l_ij));
bornSum += l_ij - u_ij + 0.25f*r*(u_ij2-l_ij2) + (0.50f*invR*ratio) +
(0.25f*params2.y*params2.y*invR)*(l_ij2-u_ij2);
bornSum += l_ij - u_ij + (0.50f*invR*ratio) + 0.25f*(r*(u_ij2-l_ij2) +
(params2.y*params2.y*invR)*(l_ij2-u_ij2));
if (params1.x < params2.x-r)
bornSum += 2.0f*(RECIP(params1.x)-l_ij);
}
......@@ -173,8 +170,8 @@ __kernel void computeBornSum(
float l_ij2 = l_ij*l_ij;
float u_ij2 = u_ij*u_ij;
float ratio = LOG(u_ij * RECIP(l_ij));
float term = l_ij - u_ij + 0.25f*r*(u_ij2-l_ij2) + (0.50f*invR*ratio) +
(0.25f*params1.y*params1.y*invR)*(l_ij2-u_ij2);
float term = l_ij - u_ij + (0.50f*invR*ratio) + 0.25f*(r*(u_ij2-l_ij2) +
(params1.y*params1.y*invR)*(l_ij2-u_ij2));
if (params2.x < params1.x-r)
term += 2.0f*(RECIP(params2.x)-l_ij);
tempBuffer[get_local_id(0)] = term;
......@@ -220,8 +217,8 @@ __kernel void computeBornSum(
float l_ij2 = l_ij*l_ij;
float u_ij2 = u_ij*u_ij;
float ratio = LOG(u_ij * RECIP(l_ij));
bornSum += l_ij - u_ij + 0.25f*r*(u_ij2-l_ij2) + (0.50f*invR*ratio) +
(0.25f*params2.y*params2.y*invR)*(l_ij2-u_ij2);
bornSum += l_ij - u_ij + (0.50f*invR*ratio) + 0.25f*(r*(u_ij2-l_ij2) +
(params2.y*params2.y*invR)*(l_ij2-u_ij2));
if (params1.x < params2.x-r)
bornSum += 2.0f*(RECIP(params1.x)-l_ij);
}
......@@ -232,8 +229,8 @@ __kernel void computeBornSum(
float l_ij2 = l_ij*l_ij;
float u_ij2 = u_ij*u_ij;
float ratio = LOG(u_ij * RECIP(l_ij));
float term = l_ij - u_ij + 0.25f*r*(u_ij2-l_ij2) + (0.50f*invR*ratio) +
(0.25f*params1.y*params1.y*invR)*(l_ij2-u_ij2);
float term = l_ij - u_ij + (0.50f*invR*ratio) + 0.25f*(r*(u_ij2-l_ij2) +
(params1.y*params1.y*invR)*(l_ij2-u_ij2));
if (params2.x < params1.x-r)
term += 2.0f*(RECIP(params2.x)-l_ij);
localData[tbx+tj].bornSum += term;
......@@ -313,6 +310,13 @@ __kernel void computeBornSum(
} while (pos < end);
}
typedef struct {
float x, y, z;
float q;
float fx, fy, fz, fw;
float bornRadius;
} AtomData2;
/**
* First part of computing the GBSA interaction.
*/
......@@ -324,7 +328,7 @@ __kernel void computeGBSAForce1(
__global float4* forceBuffers, __global float* global_bornForce,
#endif
__global float* energyBuffer, __global float4* posq, __global float* global_bornRadii,
__local AtomData* localData, __local float4* tempBuffer,
__local AtomData2* localData, __local float4* tempBuffer,
#ifdef USE_CUTOFF
__global ushort2* tiles, __global unsigned int* interactionCount, float4 periodicBoxSize, float4 invPeriodicBoxSize, unsigned int maxTiles, __global unsigned int* interactionFlags) {
#else
......
platforms/opencl/src/kernels/nonbonded_nvidia.cl
View file @ 91f3379b
......@@ -159,7 +159,7 @@ __kernel void computeNonbonded(
else {
// Compute only a subset of the interactions in this tile.
for (unsigned int j = 0; j < TILE_SIZE; j++) {
for (j = 0; j < TILE_SIZE; j++) {
if ((flags&(1<<j)) != 0) {
bool isExcluded = false;
int atom2 = tbx+j;
......@@ -230,7 +230,7 @@ __kernel void computeNonbonded(
excl = (excl >> tgx) | (excl << (TILE_SIZE - tgx));
#endif
unsigned int tj = tgx;
for (unsigned int j = 0; j < TILE_SIZE; j++) {
for (j = 0; j < TILE_SIZE; j++) {
#ifdef USE_EXCLUSIONS
bool isExcluded = !(excl & 0x1);
#endif
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment