Showing results for 
Search instead for 
Did you mean: 

PXI-6115 benchmarking

I am trying benchmarking a PXI-6115 DAQ board to determine the maximum sampling rate I can use in an application. I am using a PXI system with a PXI-1044 chasis

and a PXI-8105 controller. According to I should be able to reseach 10MS/s when acquiring from the 4 channels, but I am not capable of that... is it possible to get the code that was used to do this benchmarks?
Kind regards
0 Kudos
Message 1 of 9
You should consider posting code that DOESN'T give you the maximum sampling rate.

Probably you are having trouble transferring the data, not acquiring it. A look at the code should reveal where the problem is.
0 Kudos
Message 2 of 9



I agree with GWD probably you ar trying a continous acquisition or something like that and you are experiencing problems with the data transfer from the card to memory or to your program.

Post the code or provide the error code. But if the specifications show that this is the transfer rate it had been tested by NI.



Jaime Cabrera

NI Applications Engineering Spain
0 Kudos
Message 3 of 9
Sorry, I forgot to attached my code (there it is)... What I am seeing is that if I try to acquire from the 4 input channels at full speed the acqusition will stop at around 23 seconds showing the error shown in the attached picture. Changing the buffer, the block size, or the read wait mode does not change anything... what am I doing wrong?
Download All
0 Kudos
Message 4 of 9
That error message is extremely clear. You are acquiring data faster than you can read it from the DAQ memory. No problem with the board at all.
0 Kudos
Message 5 of 9

Yes, I can understand that

The question is... it's supposed to be possible to acquire data from the four channels at full speed according to

What do I have to do to get this performance?

0 Kudos
Message 6 of 9
It was not clear from the initial post that the problem was with data streaming.

The speed for streaming has nothing to do with the PXI-6115 but will depend on (1) the labview vi (2) setup of the labview vi (3) windows and (4) the PC bus/ hardware.

You will find a fair number of discussions about data streaming on this board. Even assuming you have a streamlined vi: you may need to discipline windows (turn off stuff that wastes cycles) and set up the prority for your vi appropriately.

Your version is too new for me but others may be able to comment on the vi.
0 Kudos
Message 7 of 9



As u can see the problem you are expericiencing means that data transfer is not fast enough for the board rate. That is... the board has an small internal memory (buffer) where the data measurements are stored. These information is loaded into the PC  memory by means of DMA transfer normally. This data transfer from the device to the PC RAM memory is not fast enough, then the on-board buffer becomes full and data is overwritten producing the error.


But you are wrong when you say the board does not work. The board works, the problem is the rest of the system, coz the data cannot be passed to memory as fast as needed.


You can see tests had been done under a PXI-8351 controller:

Also you need to set the Read Wait Time as Poll. Then, free O.S. as mentioned in previous posts, etc... Probably your system does not accomplish the same characteristics (or higher) than the 8351.

Hope this helps,




Jaime Cabrera

NI Applications Engineering Spain
0 Kudos
Message 8 of 9

Hi again

Changing the board from slot 10 to slot 3 was enough to make it possible to acquire the 4 channels at full speed.

It seems that locating a board in a high numbered slot such as 10 means the data acquired has to pass through several PCI bridges until it gets to the controller, slowing bus operations and leading to a higher risk of overflow. Therefore I understand that boards with high data transfer requitements must be placed in lower numbered slots (particulary between 3 and 6 for a PXI-1044

Kind regards


0 Kudos
Message 9 of 9