{"id":112354,"date":"2023-12-20T03:10:28","date_gmt":"2023-12-20T01:10:28","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/reliz-pocl-5-0-s-nezavisimoj-realizacziej-standarta-opencl"},"modified":"2023-12-20T03:10:28","modified_gmt":"2023-12-20T01:10:28","slug":"reliz-pocl-5-0-s-nezavisimoj-realizacziej-standarta-opencl","status":"publish","type":"post","link":"https:\/\/prohoster.info\/sq\/blog\/novosti-interneta\/reliz-pocl-5-0-s-nezavisimoj-realizacziej-standarta-opencl","title":{"rendered":"L\u00ebshimi PoCL 5.0 me implementimin e pavarur t\u00eb standardit OpenCL","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
\u00cbshte publikuar versioni i projektit PoCL 5.0 (Portable Computing Language OpenCL), i cili zhvillon implementimin e standardit OpenCL, t\u00eb pavarur nga prodhuesit e acelerator\u00ebve grafik\u00eb dhe lejon p\u00ebrdorimin e backend-eve t\u00eb ndryshme p\u00ebr ekzekutimin e kernel-eve OpenCL n\u00eb tipe t\u00eb ndryshme procesor\u00ebsh grafik\u00eb dhe central\u00eb. Kodi i projektit \u00ebsht\u00eb n\u00eb dispozit\u00eb n\u00ebn licenc\u00ebn MIT. Mb\u00ebshtetet funksionimi n\u00eb platformat X86_64, MIPS32, ARM v7, AMD HSA APU, NVIDIA GPU dhe n\u00eb procesor\u00ebt ASIP (Application-Specific Instruction-set Processor) dhe TTA (Transport Triggered Architecture) me arkitektur\u00ebn VLIW. <\/p>\n
Implementimi i kompilatorit t\u00eb kernel-eve OpenCL \u00ebsht\u00eb i bazuar n\u00eb LLVM, nd\u00ebrsa si frontend p\u00ebr OpenCL C p\u00ebrdoret Clang. P\u00ebr t\u00eb siguruar portabilitet dhe performanc\u00eb t\u00eb duhur, kompilatori i kernel-eve OpenCL mund t\u00eb gjeneroj\u00eb funksione t\u00eb kombinuara, t\u00eb cilat mund t\u00eb p\u00ebrdorin burime t\u00eb ndryshme hardware p\u00ebr t\u00eb b\u00ebr\u00eb paralelizimin e ekzekutimit t\u00eb kodit, si VLIW, supercalaritetin, SIMD, SIMT, shum\u00ebcore dhe multi-threading. Mb\u00ebshteten drejtuesit ICD (Installable Client Driver). Jan\u00eb gjithashtu t\u00eb pranish\u00ebm back-end p\u00ebr t\u00eb siguruar funksionimin p\u00ebrmes CPU, ASIP (TCE\/TTA), GPU-t\u00eb e bazuara n\u00eb arkitektur\u00ebn HSA dhe GPU-t\u00eb NVIDIA (p\u00ebrmes libcuda). <\/p>\n
N\u00eb k\u00ebt\u00eb version t\u00eb ri: <\/p>\n
\n
\u00cbsht\u00eb realizuar nj\u00eb backend i ri «Remote», i cili \u00ebsht\u00eb i destinuar p\u00ebr organizimin e llogaritjeve t\u00eb shp\u00ebrndara p\u00ebrmes transferimit t\u00eb p\u00ebrpunimit t\u00eb komandave OpenCL n\u00eb hoste t\u00eb tjer\u00eb n\u00eb rrjet, ku \u00ebsht\u00eb aktivizuar procesi mbrapa pocld.\n
N\u00eb drejtorin\u00eb CUDA jan\u00eb realizuar mund\u00ebsi t\u00eb reja dhe zgjerime p\u00ebr OpenCL 3.0, si operacione atomike, variabla me fush\u00eb t\u00eb kufizuar, n\u00ebngrupe intel_sub_group_shuffle, intel_sub_group_shuffle_xor, get_sub_group_local_id, sub_group_barrier dhe sub_group_ballot.\n
\u00cbsht\u00eb p\u00ebrmir\u00ebsuar mb\u00ebshtetja e CPU-ve bazuar n\u00eb arkitektur\u00ebn RISC-V. Funksionimi i PoCL \u00ebsht\u00eb testuar n\u00eb pllak\u00ebn Starfive VisionFive 2 me ambientin e ngarkuar Ubuntu 23.10 me LLVM 17 dhe GCC 13.2.\n
\u00cbsht\u00eb realizuar zgjerimi cl_ext_float_atomics me mb\u00ebshtetje p\u00ebr FP32 dhe FP64.\n
Realizimi i zgjerimit cl_khr_command_buffer \u00ebsht\u00eb p\u00ebrdit\u00ebsuar n\u00eb versionin 0.9.4.\n