You assemble PCs and servers for a living?
By the way, I have asked vmware people too who work with servers for a living, and been recommended a particular complete server. To which one can install ESX as you are suggesting and control it from another computer. Missing the video hardware acceleration that Workstation gives would probably not hurt too much nowadays that CPUs are so fast, but why might I need video acceleration on a VM? To play high definition online videos from youtube, vimeo etc. And to preview my edited videos before publishing them back to youtube etc. Since even encoders can be infected and malware can hide in videos, I better do the entire video edit on the VM.
By the way, my hosts do not have internet access at all. That's for security - infected VMs are disposable, you just duplicate a clean VM and back to step one.
Quote:
I guess if your end goal is a system to do something else and you don't care about technology that explains everything.
|
Well said.
Quote:
If the end goal is a system capable of copying a 10GB file in 1 second, then I will seek out the hardware and software that will let me do that. If it ends up being to large/expensive/complex then I need to re think the initial requirement.
|
Even if it wasn't for my avoidance for technical knowledge and not caring about technology, I urge you to reconsider this part of your rationale:
Say you successfully come up with a combination of hardware components that achieves the 1 GB/s transfer. And say someone else like you comes up with a combination of hardware that also achieves the 1 GB/s transfer. Then you both go out shopping online (lots of shop prices are automatically compared for you so you get the optimal price for each component and give it to the client). Now what are the chances that your total price is exactly the same as the other assembler's? I think very small. One of the two will achieve the 1 GB/s goal for less money.
Therefore for a particular client who only cares about disk transfers and not the cpu or anything else, one of the two solutions is better value for money. Another ten assemblers could do the same, and the winner would probably come up with an even lower price.
Now should this client build your combination of hardware that achieves 1.1 GB/s for $4000, or the winner of the competition of 12 assemblers that achieves 1.15 GB/s for $3500?
In short, an optimal shopping decision requires lots of hardware combinations to be benchmarked and compared, which can only be done by someone paid for it. Such as a magazine. Or whoever else can do benchmarking for a living.
Manufacturers probably do it. And they can assemble PCs with robots on a large scale for economies of scale, possibly beating all competitors in value for money if they want to. Ultimately only benchmarks plus price comparisons make the winner.