Author: Jeremiah Penery
Date: 19:48:35 02/23/03
Go up one level in this thread
On February 23, 2003 at 21:25:04, Robert Hyatt wrote: >On February 23, 2003 at 01:50:31, Jeremiah Penery wrote: > >>On February 23, 2003 at 00:53:51, Robert Hyatt wrote: >> >>>On February 22, 2003 at 20:06:44, Jeremiah Penery wrote: >>> >>>>If you don't like SPEC, look at TPC numbers. >>> >>>Let's back up to reality a minute. Servers have one type of requirement. >>>workstations have another. I'm not about to talk about the PC platform as >>>a high-performance server. It is _not_. The raw PCI bus is completely >>>unable to sustain the kind of bandwidth that the high-end server market is >>>all about, and that is a problem. Doesn't matter how fast the CPU goes. >>> >>>Just compare the I/O bandwidth of "server machines" with the I/O bandwidth >>>of a PC and you will see what I mean. >> >>Do you have any clue what TPC benchmarks are all about? No, I thought not. TPC >>are SERVER benchmarks, heavily stressing on I/O performance! > >And look at the machines you quote. Who cares about a machine that is a big >cluster of slow machines that can handle a large TPC? Just because such >cluster architectures dominate some particular niche, doesn't mean they work >for all large server applications. > >Dunn and Bradstreet in NY comes to mind. But so do others. > > >Any time you want to compare I/O throughput on any PC of your choice against >an alpha server here, I'll provide some benchmark data. But we are talking >about a database that is _not_ split into tiny parts on a pile of PCs. One >database, on one machine, spread over say 64 drives... > >The PC simply doesn't address that market. The NUMA Itaniums are up there >of course but they aren't PCs. You seem to be ignoring that TPC-W has non-clustered x86 machines in the lead.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.