That would explain it. The 989-150 doubles the cache sizes again,
to get a little more out of the processor. But then the question
is why did they cripple these new processors by going back
to so much less cache?

I am certain the new I/O system is much faster, but a surprising amount
of the time I see my systems CPU bound, even though our I/O
systems (20mb/sec SCSI and 32mb/sec PB interfaces) are so much
slower than the new PCI stuff.

-----Original Message-----
From: Marco Suerig [mailto:[log in to unmask]]
Sent: Friday, February 16, 2001 12:25 AM
To: [log in to unmask]
Subject: Re: [HP3000-L] Performance HP989-100 versus N4000-100-220


With a look to the Instruction/Data Cache sizes of the CPUs you can get an
idea why the performance units are like that:
The 989KS/100 has an Instruction Cache size of 2048KB and a Data Cache Size
of 2048KB.
The N4000-100-220 has an I-Cache size of 512KB and a D-Cache size of 1024KB.

I do not know how the HP e3000 performance Units are measured but the
N-Class should be faster regarding I/O activities because of higher PCI BUS
Speed in comparision to HP-PB Bus speed.