Question from the Customer: I am using the Cheetah SPI Host Adapter to send large data blocks at maximum speed. I find that I'm getting reliable SPI transmission for long blocks at 40Mb/sec, which corresponds to 5MB/second - these results are OK.
However, the fastest I can queue data to send using ch_spi_queue_array, in blocks of less than 64KB - since num_bytes is only a U16, is about 1.4MB/second.
What causes this difference in speed, and what can I do to improve the performance for queuing data in smaller data blocks?
About my setup:
- The laptop is running Windows 7, 64-bit
- The Cheetah adapter is connected to the laptop via USB.
- The slave device is an FPGA device. I am only blasting transmit data to the FPGA - I am ignoring what the slave might output on the MISO line.
Figure 1: Cheetah SPI Host AdapterThe key to high performance streaming is to asynchronously submit data to the Cheetah adapter, and then create the next queue while the adapter is shifting out the bytes. Then, before the current queue is completely shifted out, would then submit the next queue for delivery. Good luck with your application!
Additional resources that you may find helpful include the following:
We hope this answers your question. If you have other questions about our host adapters or other Total Phase products, feel free to email us at sales@totalphase.com, or if you already own one of our devices and have a technical question, please submit a request for technical support.