From the article you referenced, near the beginning of the discussion, 5.39 log10 units[viral copies]/mL = 250,000 IU/mL. That will allow you to get a conversion for viral copies to IU. I think you'd need to look at whatever the source of the ng/mL figure is to see how they converted between viral copies and ng. Then you can convert ng back to viral copies and then into IU.