I guess PNY hates someone or the other way around :P :P :P

[jiyang1018]

Don't take the word "hate" seriously. I am joking.

 

I did some SSD benchmark on my Rampage IV Extreme. 2 SSDs tested, OCZ Vector 256GB and PNY Optima 480GB (SMI). I noticed something very strange. Asmedia SATA 3 6gbps ports are slower than the ones from X79 chipset, which is understandable. PNY Optima 480GB suffers if Asmedia driver is installed, v2.0.8.0, v2.0.8.1, or v2.0.9.1. 

 

Take OCZ Vector 256GB results as a reference

 

Note that "AS SSD Seq Write" speed here, with any driver tested, is much slower than no driver, X79 SATA3 port, or even X79 SATA2 port.

 

 

I am not sure it is an AS SSD issue because one other tool I used has the same result as posted below. The strange thing is other 2 tools show no difference (very minor difference) with different drivers or without driver.

 

Large amount of data screenshots below:

AS SSD data:













Crystal DiskMark data:













Anvil's Storage Utilities data:


OCZ Vector was also tested on Asmedia port with v2.0.8.0 driver but I pasted the wrong text in the note section.

[SuperPug]

huh.. that is strange. but probably just a bug. i cant imagine thats purposfully done

[jiyang1018]

huh.. that is strange. but probably just a bug. i cant imagine thats purposfully done

 

Using "hate" is a joke. I agree with you. That must be a bug. I tried 8GB single file transfer, and it was always above 300MB/s from another SSD on X79 SATA3 port.

[Dimwitted]

The Optima is PNY's "basic" drive. It's low end. Relative to that it could be a controller issue in combination with the Asmedia chip like the old Sandforce/Haswell issues in the last gen drives. Hard to say for sure but it's definitely not on the same tier as the OCZ drive.

 

I know that when I was building my system that Gigabyte was insistent that I use Sata addresses 0 + 1 for ssd's which is the Intel chipset. Maybe a known issue?

[jiyang1018]

The Optima is PNY's "basic" drive. It's low end. Relative to that it could be a controller issue in combination with the Asmedia chip like the old Sandforce/Haswell issues in the last gen drives. Hard to say for sure but it's definitely not on the same tier as the OCZ drive.

 

I know that when I was building my system that Gigabyte was insistent that I use Sata addresses 0 + 1 for ssd's which is the Intel chipset. Maybe a known issue?

The problem is Asmedia controller has benchmark software performance issue with PNY Optima 480GB when driver installed. Without driver, PNY and OCZ perform closely on asmedia controller. I copied an 8GB file from another SSD to PNY on Asmedia port, and it was well over 300MB/s.

[Dimwitted]

Yep. Clearly you need to run the drive(s) without the drivers which is counterintuitive. If it's running "driverless" I would imagine it's using the Intel chipset driver as a default which doesn't have those issues. It could be that the Asmedia driver uses compression to boost performance which screws up the controller on the PNY drive and cause it to slow down. Pure WAG tho. It may not even be the driver or controller but the benchmarking software itself. Bugs abound everywhere.

[jiyang1018]

Yep. Clearly you need to run the drive(s) without the drivers which is counterintuitive. If it's running "driverless" I would imagine it's using the Intel chipset driver as a default which doesn't have those issues. It could be that the Asmedia driver uses compression to boost performance which screws up the controller on the PNY drive and cause it to slow down. Pure WAG tho. It may not even be the driver or controller but the benchmarking software itself. Bugs abound everywhere.

 

2 tools show no difference, 2 show significant difference with or without driver. Funny part is ATTO showed slow write speed once, but never happened again in the retest.

My guess is they use different writing procedures, and you are probably right on the money about compression side of things.