Silicon Investor (SI) -- The First Internet Community

STOCKTALK

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.

Pastimes : Computer Learning -- Ignore unavailable to you. Want to Upgrade?

To: bosquedog who wrote (13686)	12/4/2000 12:58:33 PM
From: tanstfl	Respond to of 110652

It may be wishful thinking, since I'm looking at dual processor for my next Computer (probably around July when the replacements for the current P4 come out); but I would think that multiple processors would take advantage of the fact that the CPU is the racehorse often not pulling a cart. By that I mean that the CPU can process data at least twice as fast as it is capable of getting it from any device on the motherboard, so with any sort of halfway decent scheduling algorithm you should see an improvement using 2 750 MHz processors as opposed to one 1.5 GHz processor. However, being lazy, this is strictly surmise on my part. Along the same lines, I just picked up an Adaptec 133U2 RAID controller and three 9 GIG SCSI drives to run as a D-Drive on my Win2kServer. I'll run RAID 0 striping and see what kind of improvement I see over the current (UDMA 2 66MHz IDE) configuration. steve

To: bosquedog who wrote (13686)	12/4/2000 5:06:56 PM
From: PMS Witch	Read Replies (1) \| Respond to of 110652

I suppose I didn't express my thoughts about dual processors very clearly. Yes, if the work can be divided so that each processor can work independently, two will double productivity. But much of PC computing doesn't lend itself to workload splitting. The majority of computer work is serial in nature: Working on the results of previous work; and completing current work in anticipation for the next process. Since today's CPU speeds result in considerable idleness, overall performance deteriorates little when secondary threads execute during idle moments of CPU capacity that otherwise would be unused as the primary process waits on slower devices. Interestingly enough, the internal micro-code in processors seems ahead of the capabilities of OS or applications in exploiting simultaneous instruction execution. Where one may perceive a meaningful improvement is speed is dealing with those cases where the primary processor is totally consumed with the task at hand and another thread is forced to wait for execution. In this case, a second processor can alleviate the contention and provide the resources needed to change a waiting system into a performing one. Another advantage of multiple processors is the reduction of time wasted for context switching. However, at today's CPU speeds, I'd be really surprised if humans could discern the difference unless this was a major bottle-neck. So, it really depends on the application. Some highly specific systems may show improvement, and in rare cases, approach a doubling of performance. But the improvement brought to most general-purpose PC computing through multiple processors will be limited. I don't know if it's a common practice, but software developers, who should know if the applications they produce can deliver improved performance from multiple processors, should recommend this configuration when advisable. That would be a real assistance to consumers. Cheers, PW.