Question Real world performance vs theoretically performance

[Wing901]

There is a big difference of theoretically performance among SATA III, PCIe 2.0 x 4, 3.0 x4 and 4.0 x4 SSDs, but in the real world, is there any significant difference between them?

 

[kanewolf]

There is a big difference of theoretically performance among SATA III, PCIe 2.0 x 4, 3.0 x4 and 4.0 x4 SSDs, but in the real world, is there any significant difference between them?

It depends on many factors. But, the real world differences are getting so small that they are difficult or impossible to perceive.

 

[Ralston18]

A long time expression comes to mind : "Measure with a micrometer, mark with a piece of chalk, cut with an axe."

= = = =

My answer is no.

Many tests are designed to achieve specific levels of results that are not viable in the real world.

I.e., those who wish to "demonstrate" that Product X is better than Product Y can easily bias the test results.

Plus it must be considered what is a "significant difference" with respect to some objective and quantifiable measurement.

I.e., Margin of error, statistical considerations, and so forth.

Some differences are not even noticeable by product end users.

Overall, performance criteria and measurements must be carefully and critically studied.

Just my thoughts on the matter.

 

[kanewolf]

A long time expression comes to mind : "Measure with a micrometer, mark with a piece of chalk, cut with an axe."

= = = =

My answer is no.

Many tests are designed to achieve specific levels of results that are not viable in the real world.

I.e., those who wish to "demonstrate" that Product X is better than Product Y can easily bias the test results.

Plus it must be considered what is a "significant difference" with respect to some objective and quantifiable measurement.

I.e., Margin of error, statistical considerations, and so forth.

Some differences are not even noticeable by product end users.

Overall, performance criteria and measurements must be carefully and critically studied.

Just my thoughts on the matter.

Can you tell the difference between 0.1 seconds and 0.12 seconds consistently? I can't. Those are the types of "real world" difference that you have to decide if they are "significant'.

 

[Ralston18]

No - I cannot tell the difference. And probably not much better these days telling the difference between 10 seconds and 12 seconds.

Likewise for things such as fps, Mbps, MBps, etc..

Agree - there may be theoretical performance test differences that are not truly end user identifiable in the real world.

I would indeed be skeptical and wary of anything touting an end user difference between 0.1 seconds and 0.12 seconds.

 

[geofelt]

Potentially big difference, depending on the app.
But minimal perceived difference most of the time.

These experts could not tell which was which:

View: https://www.youtube.com/watch?v=4DKLA7w9eeA

 

[USAFRet]

There is a big difference of theoretically performance among SATA III, PCIe 2.0 x 4, 3.0 x4 and 4.0 x4 SSDs, but in the real world, is there any significant difference between them?

Some videos:

View: https://www.youtube.com/watch?v=4YoRKQy-UO4

View: https://www.youtube.com/watch?v=4DKLA7w9eeA

View: https://www.youtube.com/watch?v=GQ9LyNXpsOo

 

[Bob.B]

There is a big difference of theoretically performance among SATA III, PCIe 2.0 x 4, 3.0 x4 and 4.0 x4 SSDs, but in the real world, is there any significant difference between them?

Depends on the stuff you run.

I can't tell the diff between a sata ssd and a gen3x4 ssd.

 

[punkncat]

Potentially big difference, depending on the app.
But minimal perceived difference most of the time.

I think this is a big part of that equation. Perception. Seat of the pants.

Are there certain places or actions you can tell? Yes.
Are you going to be able to tell for day to day web surfing? No.

One of the biggest factors right now inside the PC case alone comes down to things such as what are you writing to and from? Many motherboards have one Gen4x4 slot and then lesser additional slots. Your data transfer even on a local level is going to be the slowest drive in the chain. Even if it was the same speed and as pointed out above, you are talking about a few seconds here or there. Might see some difference in load to OS times. Perhaps a bit quicker load into a new map on a game, things like that.

edit- keep in mind as well that you tend to get pretty used to what you are working on and with. It can often be very difficult to "unsee" that when you use a system notably newer or faster. This is often more to do with other hardware than the drives or even graphics card particularly with age as a factor.

 

[thestryker]

Personal experience:
I couldn't tell the difference between SATA and PCIe 3.0 x4 as an OS drive. I could tell the difference between NAND and Optane, but I'd never actually recommend Optane to anyone unless they had money to burn.

SATA vs PCIe 4.0 x4 drives for gaming/applications is a bit different and very hit and miss. Things like transfer, verification and installation rates are very obvious. Game loading can be different, but that is super hit and miss because a lot of game loading isn't sequential. When I tested my drives the NVMe ones were all about twice as fast as SATA at low QD so that's going to be the peak difference for a lot of these situations despite sequential being 10-12x faster.

 

[USAFRet]

The move from spinning HDD to solid state was huge.

The various flavors of solid state?
Not so much.

I have 6x SSDs in my main system.
1x 980 Pro
1x Intel 660p (slow in the NVMe world)
4x SATA III.

Sometimes, it is hard to tell.