SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : CYRIX / NSM -- Ignore unavailable to you. Want to Upgrade?


To: Investor A who wrote (27793)7/2/1998 4:52:00 PM
From: Scumbria  Read Replies (1) | Respond to of 33344
 
Fuchi,

Great finds, once again demonstrating the lousy price-performance tradeoff of Intel CPU's.

It was interesting to note that the full speed L2 on Xeon produces no performance benefit in single processor systems. More evidence that memory bandwidth is relatively unimportant compared to memory latency.

Scumbria



To: Investor A who wrote (27793)7/2/1998 5:14:00 PM
From: Joe NYC  Respond to of 33344
 
Fuchi,

Great article about AGP! This new guy at Tom's web site is great.

Joe



To: Investor A who wrote (27793)7/2/1998 5:25:00 PM
From: Joe NYC  Respond to of 33344
 
Fuchi,

One thing I find strange is that there are so few benchmarks for server performance.

It's not so hard to set up an SQL Server or Oracle Server, load a good size database, and throw some database queries and updates at it.

Or to set up a file server and a couple of worstations that simulate a lot of file related activity.

I have seen some tests where they test # of transactions per second. I was never quite sure what these transactions are. Creations of new records? Those kind of tests will at some point start to stress the disk subsystem. And anyway, the speed of lookups tends to be more important than updates.

I think if you want to know what your CPU(s) is doing for you, you have to eliminate disk activity from equation, and the best way to do it is to have a lot of RAM and do queries on the data that's cached in RAM. This will somewhat isolate the CPU, even though you still have to take into consideration the memory access speed, quality of chipsets, PCI implementation and possibly NIC.

Joe



To: Investor A who wrote (27793)7/2/1998 8:21:00 PM
From: Dale J.  Read Replies (3) | Respond to of 33344
 
Fuchi,

tomshardware is biased against Intel. Isn't he that fellow that was almost sued last year by Intel for using the No-Inel Logo, or do I have my wires crossed.

Dale