tromp
commented
5 years ago What executable did you run? src/cuckoo/cuda29 ? What are your first lines of output? I get
$ ./cuda29 -d 0 -E 0 -h "" -m 176 -n 0 -r 1 -U 4096 -u 256 -v 128 -w 512 -y 1024 -Z 1024 -z 1024 GeForce GTX 1080 Ti with 10GB @ 352 bits x 5505MHz Looking for 42-cycle on cuckoo30("",0) with 50% edges, 64*64 buckets, 176 trims, and 64 thread blocks. Using 6976MB of global memory. nonce 0 k0 k1 k2 k3 a34c6a2bdaa03a14 d736650ae53eee9e 9a22f05e3bffed5e b8d55478fa3a606d Seeding completed in 545 + 307 ms 4-cycle found 2-cycle found 16-cycle found 26-cycle found 40-cycle found findcycles edges 70214 time 16 ms total 2010 ms Time: 2010 ms 0 total solutions
Note that this needs 7 GB to run. Does your GPU have that much?
nevata
timolson
-d 0 -E 0 -h "" -m 176 -n 0 -r 1 -U 4096 -u 256 -v 128 -w 512 -y 1024 -Z 1024 -z 1024
The thread 0x1c84 has exited with code 0 (0x0). The thread 0x1160 has exited with code 0 (0x0). The thread 0x2a30 has exited with code 0 (0x0). First-chance exception at 0x01049349 in vs2013.exe: 0xC0000005: Access violation reading location 0x08812770. Unhandled exception at 0x01049349 in vs2013.exe: 0xC0000005: Access violation reading location 0x08812770.
class cuckoo_hash { public: u64 *cuckoo;
cuckoo_hash() { cuckoo = new u64[CUCKOO_SIZE]; } ~cuckoo_hash() { delete[] cuckoo; } void set(node_t u, node_t v) { u64 niew = (u64)u << NODEBITS | v; for (node_t ui = u >> IDXSHIFT; ; ui = (ui+1) & CUCKOO_MASK) { u64 old = cuckoo[ui]; //error at if (old == 0 || (old >> NODEBITS) == (u & KEYMASK)) { cuckoo[ui] = niew; return; } } } node_t operator[](node_t u) const { for (node_t ui = u >> IDXSHIFT; ; ui = (ui+1) & CUCKOO_MASK) { u64 cu = cuckoo[ui]; //error at if (!cu) return 0; if ((cu >> NODEBITS) == (u & KEYMASK)) { assert(((ui - (u >> IDXSHIFT)) & CUCKOO_MASK) < MAXDRIFT); return (node_t)(cu & NODEMASK); } } } };