When and how to use the Multiplier IP core

Hi, I'm Stacey, and in this video I talk about when and how to use the Multiplier IP core instead of inferring one.
Buy me a coffee to support my channel: www.buymeacoff...

[@HarpreetBhatia7]

Glad you are back!

[@susmitamaloji7111]

Where can we get that first code, you have shown in the starting of the code, plz reply me

[@cheneymx]

this video helped me quite a lot. Thx!

[@robertkent2809]

I think I have come across this problem before. When you slice "window_out", the result is interpreted as unsigned, and signed x unsigned multiplication is interpreted as unsigned. My solution to this was something like "mic_data*$signed(window_out[0:-WINDOW_FRAC]);" This has worked for me in the past.

[@zamanizambri832]

This is correct. If you slice one of the multiplicands you need to type cast it. I originally thought this was a Vivado synthesis issue, but I believe this is true even for Synplify/ProtoCompiler.

[@grantyale]

Yup. This should be pinned.

[@RamavathKavitha]

Hi, can you pls share code the for using fft ip core in vhdl

[@ahll2006]

Hi, With the DSP48E1, I can multiply 23-bit signed and 18-bit signed values Now I have a 23-bit unsigned value and my 18-bit value is unsigned (just like a 19-bit signed with only positive values) Is it possible to configure the DSP slice accordingly? If not, is there an elegant way to pre-process the multiplicands or post-process the product in order to arrive at the correct result? Thanks!

[@FPGAsforBeginners]

You should be able to use the same resources for signed or unsigned. I recommend using the multiplier IP core to try out your configuration and see what resources it shows. Usually one of the tabs will allow you to choose to use a DSP slice, and how many your multiplication will use up.

[@ahll2006]

@@FPGAsforBeginners After a few days of trying out all the options, I came out to this conclusion which is DSP 18-bit signed only or 17-bit unsigned. walk around by doing what this thread says. support.xilinx.com/s/question/0D52E00006iHjc5SAC/dsp48e-multiplication-of-25-bit-signed-and-18-bit-unsigned-value-possible?language=en_US

[@nitilpoddar]

Hey Stacey! Could you tell me how can I understand this wizardry. This all looks absolutely bonkers. But I wish to learn.

[@andyashton7061]

Is the ability to have multiple block diagrams in one project a new feature in 2021.1? I am using 2020.1. Thanks for the videos, they're great!

[@russianultraviolet4441]

Hm. I am suspecting the same problem with a code i am writing. Got a lot of multiplications and divisions of signed numbers in an equation and the result comes out wrong. I tried to avoid core ip as much as i can to make things simpler.... Thanks for the hint.

[@MrVerilog]

Nice short useful video

[@siddhantiiitd9523]

After Installing vivado 2019.1 on ubuntu, I have to set up the license and when I go to help>manage license in vivado and then in manage license search paths, I get "Linux users will need to make these settings outside of this application" and I can't edit the license. Can you please help me resolve the issue

[@FPGAsforBeginners]

You probably need to edit one of your environment variables in your bashrc. XILINXD_LICENSE_FILE or LM_LICENSE_FILE usually/

[@ekbhatnagar]

Good videos, very helpful. I started to work in the field as a freshers last month. Please do tell if you need my help in any videos. Help for you, knowledge for me😅