Brad Wardell's site for talking about the customization of Windows.

We’ve had some great discussions on Hard drive speeds over the years.

One of my favorites was the one about whether SSDs was worth it.

I’ve also been benchmarking my PC’s for a very, very long time:

https://forums.wincustomize.com/168111

Here is a brief recap:

image

Most of the CPU speed improvements comes from adding more CPU cores.  My 2019 machine has 18 cores, 32 logical threads.  My 2008 box was the last machine with only a single core in it.  That isn’t to say that the CPU experience isn’t much better these days.  I’m just saying that unless your app is using all those cores, you could argue that you could divide these scores by the number of cores to get a better approximation of what a single threaded app would perform like.

Anyway, here are the results from my box today:

Samsung 970 Pro NVME:

image

Sabrent Rocket  NVME 2280:

image

As a practical matter, they’re both incredibly fast. 

In real world practice, an NVME (like these) will have roughly the same load times as a SATA SSD despite scoring a lot lower on the benchmarks.

Feel free to use this thread to post your Hard drive specs and scores to share with others.


Comments (Page 2)
2 Pages1 2 
on Feb 28, 2020

starkers

Yeah, there's quite some difference between ALMonty's and our default scores, but the thing that I don't get is why Peak and Real World scores for Azdude and myself are through the roof.  I understand that tests under Peak are probably measuring optimum levels, but the Real World scores are equally confusing.

The reason is you are not comparing like with like.  The Peak and Real world scores are showing uops and us vs MB/Sec which the other ones show.

ALMonty's is due to some sort of ram caching some manufacturers offer.  This makes the benchmark scores look great but the OS was already caching when needed most of the time and the difference is the OS cache gets disabled by the benchmark (intentionally), but a third party one does not.

on Feb 28, 2020

I just remembered I have Rapid mode turned on through my Samsung software. If that answers anyone's questions.

on Feb 28, 2020

Neil Banfield

The reason is you are not comparing like with like. The Peak and Real world scores are showing uops and us vs MB/Sec which the other ones show.

No, I was curious as to why Peak AND Realworld readings were so different to the default ones..... and unlike ALMonty, I do not have any firmware or software installed to boost my OS drive.  I do have IObit's Smart Defrag installed, but it merely Optimises and trims, that's all.

2 Pages1 2