>
> Ok. I didn't want to underestimate Your knowledges. I actually wanted
> You to take a closer look at the architectures before judging them. The
> page I mentioned was:
>
http://docs.sun.com:80/ab2/coll.28.19/SBUSPCICMP/@Ab2PageView/215?Ab2Lang=C&;
Ab2Enc=iso-8859-1
I'm a little curious who wrote this article as clearly they are biased
toward the SBUS. They didn't bother to look at the actual difference before
writing None in the differences column. The following strikes me as a good
indicator of a bad report.
Bus bandwidth:
SBUS 25MHz/32-bit, max.
100MBytes/sec. for 32-bit
200MBytes/sec for 64-bit (which they don't even support as show above)
PCI 33MHz (rev 2.0) or 66MHz for PCI/66, 32/64bit
132MByte/second (33MHz/32-bit)
to 528MByte/second (66 MHz/64-bit).
Difference according to the reviewer? NONE
That's a patently misleading conclusion. Clearly the PCI bus is far superior
based on the numbers shown above. I'm not sure how someone can look at those
numbers and conclude that there is no difference. The PCI bus performs over
5X better for data transfer than the SBUS. There are other, similar
inconsistencies in the report.
As to your question about why they continue to use SBUS for high-end server
architecture? Hard to say but don't immediately conclude that it's because
of performance or technical characteristics. Think back to the argument over
ISA vs MCA. MCA was clearly the superior bus architecture at the time but
rather than going with that, PC manufacturers continued to cling to an
underperforming ISA architecture. Never, ever assume that all decisions
related to computers are made based strictly on the technological advantages
of one system over another. In fact, often, the inferior system wins out.
For a wide number of reasons.
Ric
-
[To unsubscribe, send mail to [EMAIL PROTECTED] with
"unsubscribe firewalls" in the body of the message.]
