Re: [Discuss-gnuradio] asymmetric USRP throughput

[Eric Blossom]

On Thu, Sep 13, 2007 at 02:03:55AM +0200, Vincenzo Pellegrini wrote:
> AMD sempron 3000+, 512KiB RAM
> USB2.0 on board

> 00:03.0 USB Controller: Silicon Integrated Systems [SiS] USB 1.0
> Controller (rev 0f)
> 00:03.1 USB Controller: Silicon Integrated Systems [SiS] USB 1.0
> Controller (rev 0f)
> 00:03.2 USB Controller: Silicon Integrated Systems [SiS] USB 1.0
> Controller (rev 0f)
> 00:03.3 USB Controller: Silicon Integrated Systems [SiS] USB 2.0
> Controller

Vincenzo, 

I'm not familiar with this chipset.
It's not inconceivable that it's got asymmetric throughput.  We've
seen that on many PPC host controllers.

Eric


> On Wed, 2007-09-12 at 16:35 -0700, Eric Blossom wrote:
> > On Wed, Sep 12, 2007 at 11:55:08PM +0200, Vincenzo Pellegrini wrote:
> > > I'm beginning to see some light over my 8MHz tx problems..
> > > 
> > > the throughput towards my USRP is significantly reduced compared to the
> > > one from it...
> > > 
> > > ./test_usrp_standard_tx
> > > xfered 1.34e+08 bytes in 5.46 seconds.  2.457e+07 bytes/sec.  cpu time =
> > > 0.492
> > > 0 underruns
> > > 
> > > 
> > > ./test_usrp_standard_rx
> > > xfered 1.34e+08 bytes in 4.19 seconds.  3.2e+07 bytes/sec.  cpu time = 
> > > 0.384
> > > noverruns = 0
> > > 
> > > what could be the cause for this? any hint..? ( benchmark_usb.py reports 
> > > a good throughput up to 32M)
> > > 
> > > thanks 
> > > vincenzo
> > 
> > What h/w platform are you on?'
> > What USB host controller do you have?
> > Is it on a PCI card, or on the motherboard?
> > 
> >  $ lspci
> > Eric