[{"data":1,"prerenderedAt":389},["ShallowReactive",2],{"project-3513165":3},{"id":4,"name":5,"line":6,"line_source":7,"title":8,"url":9,"domain":10,"maker":11,"posted_at":12,"year":13,"points":14,"comments":15,"rank_in_year":16,"shown_in_year":17,"top":18,"tier":19,"category":20,"subcategory":21,"format":22,"keywords":23,"reception":28,"sentiment":29,"stance_mix":30,"scored_comments":35,"aspects":36,"strip":42,"post_text":45,"what":46,"said":78,"maker_in_thread":192,"shape":199,"grid":203,"attempts":351,"same_week":352,"maker_shows":18,"maker_hidden":381,"reader_version":382,"card":383},"3513165","Alenka","my database engine for GPU","title","Show HN: my database engine for GPU","https:\u002F\u002Fsourceforge.net\u002Fprojects\u002Falenka\u002Ffiles\u002F","sourceforge.net","antonmks",1327560950,2012,364,60,24,1677,2,"legend","data","database engine","other",[24,25,26,27],"gpu database","cuda","column store","sql engine","mixed",0.132,{"positive":31,"neutral":32,"mixed":33,"critical":34},6,10,1,4,21,{"docs":37,"performance":39,"execution":40,"idea":41},{"positive":38,"critical":33},0,{"positive":38,"critical":33},{"positive":38,"critical":38},{"positive":31,"critical":38},[43,33,43,43,38,43,38,43,38,43,38,38,38,38,44,44,38,33,43,44,38,43,44,43,33,44,44,38,43,38,38,38,38,38,38,38,38,38,38,38],3,5,null,{"problem":47,"audience":45,"pricing":45,"motivation":45,"time_to_build":45,"stack":59,"openness":77,"stage":77,"team_size":77},{"value":48,"source":49,"evidence":50},"Running database scans and aggregations faster by using the GPU.","inferred",[51,55],{"id":52,"by":53,"text":54},"3513324","corysama","Very, very cool. I'm always happy to see more ways to utilize the tremendous performance\u002Fprice of GPUs. So, what are the pros and cons of this approach? For a given hardware and…",{"id":56,"by":57,"text":58},"3513255","ericflo","Interesting. I wonder how a GPU could best be put to use in a database engine. Joins, I suppose? Would be cool to read more about this implementation, however the readme is a bit…",[60,70],{"value":61,"source":62,"evidence":63},"CUDA","maker",[64,67],{"id":65,"by":11,"text":66},"3513391","It is 6 GB for Tesla GPU. PCI Express transfer speed is 5GB\u002Fs, with average compression rate averaging 4\u002F1 the speed becomes 20GB\u002Fs so transfer speed is not a bottleneck anymore.",{"id":68,"by":11,"text":69},"3513289","Thank you ! The data are stored on a disk and only the columns necessary for transformations are brought to GPU. In case the data does not fit into a limited GPU memory, it is…",{"value":71,"source":49,"evidence":72},"C",[73],{"id":74,"by":75,"text":76},"3513629","6ren","github would be better for browsing. But the source is only 117k - I'm surprised how much of a barrier a few clicks and a few seconds was for me. (Partly, I expected it to be…","unknown",{"praise":79,"objections":94,"requests":128,"questions":139,"alternatives":166},[80],{"value":81,"source":82,"evidence":83},"Impressive GPU database idea and implementation","commenter",[84,85,89,93],{"id":52,"by":53,"text":54},{"id":86,"by":87,"text":88},"3513300","jrmg","I love this kind of novel thinking. I'd definitely like to read more detailed articles on how this works.",{"id":90,"by":91,"text":92},"3513345","jdf","If you look up columnar databases, you can find a whole host of research on how these things store data. Storing values column by column to do vector processing is pretty common,…",{"id":74,"by":75,"text":76},[95,102,113],{"value":96,"source":82,"evidence":97},"Column stores are poor for writes and transactional work",[98],{"id":99,"by":100,"text":101},"3515345","alecco","Column store, compare it with MonetDB (open source) and similar. This design is suitable for read-mostly databases. It's a trade-off, faster reads and much better compression.…",{"value":103,"source":82,"evidence":104},"GPU transfer overhead and disk streaming limit performance on large data",[105,109],{"id":106,"by":107,"text":108},"3513365","mrich","Congratulations, well done! Our team has built an in-memory DB (hybrid, both column and row store available, also using advanced compression). We have investigated GPUs for…",{"id":110,"by":111,"text":112},"3516168","grundprinzip","Looks like a very interesting approach to me given the recent developments. Still I have a few questions regarding your numbers. Can you specify how much data you actually stream…",{"value":114,"source":82,"evidence":115},"SourceForge hosting is hard to browse",[116,120,124],{"id":117,"by":118,"text":119},"3514373","breckinloggins","Excellent work! I have a very small (marketing-related) nitpick: Would you consider hosting your project on github instead of SourceForge? I don't say this to be trendy, github…",{"id":121,"by":122,"text":123},"3515725","davvid","Another practical consideration: I clicked on the \"manual.txt\" file and SF redirected me to a download page. I would have preferred to read the file in-place (like on github).",{"id":125,"by":126,"text":127},"3515786","astrodust","I second this. I'd rather be at the dentist than on a SourceForge page. I swear it's the \"myspace.com\" of open source hosting.",[129,134],{"value":130,"source":82,"evidence":131},"Move the project to GitHub",[132,133],{"id":117,"by":118,"text":119},{"id":121,"by":122,"text":123},{"value":135,"source":82,"evidence":136},"More detailed documentation and overviews",[137,138],{"id":56,"by":57,"text":58},{"id":74,"by":75,"text":76},[140,147,158],{"value":141,"source":82,"evidence":142},"Does it work with ATI\u002FAMD GPUs or only NVIDIA?",[143],{"id":144,"by":145,"text":146},"3514340","2AM","nvidia gpus only, right? does it work with ati as well? CUDA is probably more advanced, so it's no surprise.",{"value":148,"source":82,"evidence":149},"Why CUDA rather than OpenCL?",[150,154],{"id":151,"by":152,"text":153},"3515810","pavelkaroukin","antonmks, did you consider using OpenCL instead? If yes - why CUDA outweighed it?",{"id":155,"by":156,"text":157},"3517246","mappu","Whether it's still relevant or not CUDA has a lot of mindshare in the academic community. Partially this is due to marketing and having their product out first, but also OpenCL…",{"value":159,"source":82,"evidence":160},"Whether the database is stored on disk or in memory",[161,165],{"id":162,"by":163,"text":164},"3513259","chubs","That looks absolutely brilliant. So is the database all stored in the GPU memory? Or in main memory? Or does your engine handle storing to disk as well? Pretty novel use of a GPU…",{"id":68,"by":11,"text":69},[167,172,178,185],{"value":168,"source":82,"evidence":169},"MonetDB",[170,171],{"id":99,"by":100,"text":101},{"id":90,"by":91,"text":92},{"value":173,"source":82,"evidence":174},"SAP HANA",[175],{"id":176,"by":107,"text":177},"3516265","You can find info here (unfortunately not just an easy download) http:\u002F\u002Fwww.sap.com\u002Fhana\u002Findex.epx",{"value":179,"source":82,"evidence":180},"PGStrom",[181],{"id":182,"by":183,"text":184},"3513781","foobarbazetc","This is awesome. :) Someone did something similar for PostgreSQL here: http:\u002F\u002Fwiki.postgresql.org\u002Fwiki\u002FPGStrom",{"value":186,"source":82,"evidence":187},"kdb\u002FQ and Kona",[188],{"id":189,"by":190,"text":191},"3514000","beagle3","Super cool, and thanks for sharing! Are you familiar with APL\u002FK\u002FQ\u002FJ ? These implement column store databases which a programming language significantly more expressive and…",{"comments":34,"answered":34,"key":193},[194,195,198],{"id":65,"text":66},{"id":196,"text":197},"3513434","By column. FOR(frame of reference) compression for integers and decimals, FOR-DELTA for already sorted columns, dictionary for strings.",{"id":68,"text":69},{"commenters":200,"max_depth":44,"maker_replies":34,"first_comment_mins":201,"lifespan_hours":202},48,38,29,[204,206,208,210,212,213,217,218,220,223,226,229,233,236,239,240,242,243,246,248,250,253,255,258,260,263,266,269,271,272,273,274,276,278,280,282,283,285,289,291,293,294,297,301,304,307,310,313,316,319,322,325,328,331,334,337,340,343,346,349],[117,38,43,205,118,38],"project\u002Fneutral\u002Fnone",[121,38,33,207,122,38],"project\u002Fcritical\u002Fdocs",[125,38,33,209,126,38],"project\u002Fcritical\u002Fnone",[106,33,43,211,107,38],"project\u002Fneutral\u002Fperformance",[65,33,38,211,11,33],[214,33,43,215,216,38],"3513422","project\u002Fneutral\u002Fexecution","alexchamberlain",[196,33,38,215,11,33],[219,33,43,211,107,38],"3513485",[221,33,38,222,216,38],"3513565","commenter\u002Fneutral\u002Fnone",[224,33,38,222,225,38],"3513588","orcadk",[227,33,43,211,228,38],"3513803","binarymax",[230,33,38,231,232,38],"3515566","other\u002Fneutral\u002Fnone","marshray",[234,33,38,231,235,38],"3515970","Symmetry",[237,33,43,238,111,38],"3516083","project\u002Fneutral\u002Fdocs",[176,33,38,231,107,38],[241,33,38,222,111,38],"3517282",[189,18,38,231,190,38],[244,18,38,231,245,38],"3514448","silentbicycle",[247,18,38,222,190,38],"3514751",[249,18,38,222,245,38],"3514883",[251,18,38,231,252,38],"3517112","vessenes",[52,43,44,254,53,38],"project\u002Fpositive\u002Fidea",[256,43,44,254,257,38],"3513521","liquidcool",[259,43,38,222,216,38],"3513562",[261,43,38,222,262,38],"3514541","Impossible",[264,43,38,222,265,38],"3517989","maximilian",[267,34,33,209,268,38],"3513237","joshu",[74,34,43,270,75,38],"project\u002Fmixed\u002Fdocs",[182,44,38,231,183,38],[162,31,44,254,163,38],[68,31,38,215,11,33],[110,275,43,211,111,38],7,[277,275,38,211,11,33],"3517383",[56,279,44,254,57,38],8,[151,281,43,215,152,38],9,[155,281,38,231,156,38],[99,32,33,284,100,38],"project\u002Fcritical\u002Fperformance",[286,287,44,254,288,38],"3514787",11,"toblender",[144,290,43,205,145,38],12,[86,292,44,254,87,38],13,[90,292,38,231,91,38],[295,292,38,231,296,38],"3513688","cshesse",[298,299,43,205,300,38],"3513271",14,"TobbenTM",[302,299,38,222,303,38],"3513297","maggit",[305,299,38,222,306,38],"3513546","beaumartinez",[308,299,38,222,309,38],"3517507","snaky",[311,299,38,222,312,38],"3513315","Tuna-Fish",[314,299,38,222,315,38],"3513330","tricolon",[317,299,38,222,318,38],"3513371","joelthelion",[320,299,38,222,321,38],"3514895","lambda",[323,299,38,222,324,38],"3513381","qw",[326,299,38,222,327,38],"3514056","phamilton",[329,299,38,222,330,38],"3513407","nakkiel",[332,299,38,222,333,38],"3513421","ajuc",[335,299,38,222,336,38],"3513433","michael_dorfman",[338,299,38,222,339,38],"3513436","petsos",[341,299,38,222,342,38],"3513438","killedbydeath",[344,299,38,222,345,38],"3513477","oozcitak",[347,299,38,222,348,38],"3513484","tzury",[350,299,38,222,225,38],"3513591",[],[353,358,363,368,373,377],{"id":354,"name":355,"points":356,"category":357},"3539792","Screenleap",402,"dev_tools",{"id":359,"name":360,"points":361,"category":362},"3518051","SocialFolders",187,"productivity",{"id":364,"name":365,"points":366,"category":367},"3540481","My afternoon project",185,"web_saas",{"id":369,"name":370,"points":371,"category":372},"3512445","Cheekswab",120,"consumer_social",{"id":374,"name":375,"points":376,"category":367},"3526929","TrollBlock",88,{"id":378,"name":379,"points":380,"category":372},"3519864","Bellbot",85,false,"v2",{"id":384,"name":5,"line":6,"ai":381,"cat":20,"year":13,"points":14,"comments":15,"rec":28,"s":385,"top":18,"kw":386,"maker":11,"domain":10,"tier":19,"attempts":38,"strip":387,"url":9,"mix":388,"praise":81,"objection":96,"sub":21},3513165,0.13,[24,25,26,27],"3133030303000055013503531550300000000000",[31,32,33,34],1793990878273]