search

neurosim / DNN_NeuroSim_V2.0

Benchmark framework of compute-in-memory based accelerators for deep neural network (on-chip training chip focused)
44 stars 18 forks source link

I have some problem when I read the code in CHIP.CPP. #4

Closed springluo closed 4 years ago

springluo commented 4 years ago

Recently ,I read the file Chip.cpp.I have some doubt. line 652 in chip.cpp define the variable int weightMatrixRow = netStructure[l][2]*netStructure[l][3]*netStructure[l][4]*numRowPerSynapse; line 840 in Chip.cpp define the variable int numRowMatrix = min(desiredPESizeNM, weightMatrixRow-i*desiredPESizeNM); the numRowMatrix which I think is the row number of weight that in a tile,then I think line 840 should be bellow: int numRowMatrix = min(desiredPESizeNM*9, weightMatrixRow-i*desiredPESizeNM*9); becasue there is 9 PE in each tile.
If the line 840 is correct. the numRowMatrix transmit to the variable weightMatrixRowof functionTileCalculatePerformane(). the functionTileCalculatePerformane in file Tile.cpp line576:int location = i*MIN(peSize, (int) weightMatrixRow/numPE); line578:pEMemoryOld = CopyPEArray(oldMemory, location, 0, (int)(weightMatrixRow/numPE), weightMatrixCol);the same case in line 581,583. so here are some problem. If the weightMatrixRow=desiredPESizeNM,thenweightMatrixRow/numPE stand for what? can you explain it ? Thank you.

neurosim commented 4 years ago

You are right, the numRowMatrix should be int numRowMatrix = min(desiredPESizeNM*numPENM, weightMatrixRow-i*desiredPESizeNM*numPENM); We have updated the codes accordingly.

springluo commented 4 years ago

hello，Very glad to receive your reply.Thank you very much for writing such an excellent tool so we can use it to evaluation our design. recent I'm try to use NeuriSim to benchmark some work.but i have some problem. could you  please give me your contact information if possible convenient . Thank you very much

------------------ 原始邮件 ------------------ 发件人: "NeuroSim"<notifications@github.com>; 发送时间: 2020年5月6日(星期三) 上午7:42 收件人: "neurosim/DNN_NeuroSim_V2.0"<DNN_NeuroSim_V2.0@noreply.github.com>; 抄送: "爱玩"<834843018@qq.com>;"Author"<author@noreply.github.com>; 主题: Re: [neurosim/DNN_NeuroSim_V2.0] I have some problem when I read the code in CHIP.CPP. (#4)

You are right, the numRowMatrix should be int numRowMatrix = min(desiredPESizeNMnumPENM, weightMatrixRow-idesiredPESizeNM*numPENM); We have updated the codes accordingly.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

neurosim commented 4 years ago

Please let us know your affiliation and the scope of your work, so we can decide how we can help.

On Wed, May 6, 2020 at 4:04 AM changhang luo notifications@github.com wrote:

hello，Very glad to receive your reply.Thank you very much for writing such an excellent tool so we can use it to evaluation our design. recent I'm try to use NeuriSim to benchmark some work.but i have some problem. could you  please give me your contact information if possible convenient . Thank you very much

------------------ 原始邮件 ------------------ 发件人: "NeuroSim"<notifications@github.com>; 发送时间: 2020年5月6日(星期三) 上午7:42 收件人: "neurosim/DNN_NeuroSim_V2.0"< DNN_NeuroSim_V2.0@noreply.github.com>; 抄送: "爱玩"<834843018@qq.com>;"Author"<author@noreply.github.com>;

主题: Re: [neurosim/DNN_NeuroSim_V2.0] I have some problem when I read the code in CHIP.CPP. (#4)

You are right, the numRowMatrix should be int numRowMatrix = min(desiredPESizeNMnumPENM, weightMatrixRow-idesiredPESizeNM*numPENM); We have updated the codes accordingly.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/neurosim/DNN_NeuroSim_V2.0/issues/4#issuecomment-624501999, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHDWYRVMBC6BT4VLS24LPPDRQEKYDANCNFSM4MCDMFAQ .

-- Shimeng Yu