128GB Kingston SSDNow V+200 SSD performing poorly

[mikeymikec]

The ATTO benchmark is giving figures at the low end that I would class as low for a HDD (2-4MB/sec), let alone an SSD. I found a benchmark for that exact drive on a review site back in the day that it was 'current' (2012) which suggested a ATTO result more along the lines of 20MB/sec a the lowest end should be expected.

It has a Sandforce controller, which IIRC were notorious for reliability issues, but long-term performance issues as well?

I tried TRIMming it, but Win10 had done it recently already and doing it again made no difference to the ATTO benchmark. At the highest end it was saturating the SATA 3Gbps connection (laptop btw, somewhat dated, Celeron T3300).

SMART readings for the drive are all fine (reckons 100% health). ~50% capacity.

 

[bbhaag]

I'd get anything important off it right now. Last time the hdd in my wifes laptop had similar symptoms it lasted about 12 hours and then crashed.

 

[dave_the_nerd]

Post full benchmark results please.

4MB/sec on 4k random write (usually the slowest) would be 1k IOPS, which is actually ok for SSDs of that vintage "in the wild" (not a review, which is probably benchmarking an empty drive.)

If your sequential reads (usually the fastest) can still saturate a SATA-2 connection (3Gbps) you're probably fine.

So I suspect you're probably alright, although any piece of consumer electronics sold in 2012 is approaching my threshold for "wait... do I want to replace this because I have an upgrade itch, or because it's actually slow and annoying?"

 

[mikeymikec]

 

[mxnerd]

Celeron T3300 FSB speed is 800MHz, already forgot whether it has anything to do with SATA speed?

Or it has anything to do with chipset?

Post your laptop model spec and it might help.

===

Seems your SSD benchmark is within the ballpark.

http://www.legitreviews.com/kingston-ssdnow-v200-128gb-ssd-review_1980/4

 

[mikeymikec]

Asus X5DIJ (surprising lack of specs when googling, esp. on asus site)
celeron t3300
2GB RAM (doing surprisingly well, 38-48% memory usage when idle)
WIn10-64 1803
intel gma4500 graphics I think
probably correct: http://mobilespecs.net/laptop/ASUS/ASUS_X5DIJ.html


This is the benchmark that made me wonder if this particular ssd is up the creek):
https://www.tweaktown.com/reviews/4...00_120gb_solid_state_drive_review/index5.html

I'm considering disconnecting the SSD from the laptop and connecting it to my PC via USB3 to see whether the SSD benchmark is CPU intensive at the low end. I'm a little reticent about it because it involves taking the underside completely off, and it does seem like it would be for naught.

 

[mxnerd]

Wow, why the benchmark results from different sites differ so much?

Kingston seems have V series and V+ Series SSDs, are your sure you have the V+ SSD?

==

Found Kingston's discontined SSDNOW V+ specs.

https://www.kingston.com/datasheets/snvp325-s2_us.pdf

• Sequential Read Throughput — 230 MB/sec.
• Sequential Write Throughput — 180MB/sec

How in the world Tweaktown achieved such high performance that's way beyond vendor's own claim?

==

Your ASUS laptop apparently use one of the 3 mobile chipsets: GM45, GS45 or GL40, which only supports SATA II 3.0 Gb/s. That's definitely a limiting factor.

https://www.mouser.com/ds/2/612/mobile-gm45-intel-gs45-gl40-express-chipsets-brief-257542.pdf

So it could be that SSDNow drives can actually perform better than SATA 2 speed but was limited by the SATA technology at the time.

 

[mikeymikec]

I took the drive ID from Device Manager and googled it, it came up with that.

I plugged the drive into USB3 on my machine and got comparable ATTO stats to the ones I posted. The drive appears to only do SATA 3Gbps as well.

 

[mxnerd]

According to legitreviews, Kingston have 2 very similarly named SSDs with different size and very big performance difference.

Pull out the drive and look at the label. Drive ID / Device ID could be wrong.

 

[dave_the_nerd]

That honestly looks pretty good to me, for a SATA-2 system. It might not be the theoretical maximum of which the drive is capable in perfect benchmarking conditions, but it's way better than "broken."

Wow, why the benchmark results from different sites differ so much?

Different testing methodologies and testbench hardware. You can't really compare across different sites/test labs, you just have to find a tester whose equipment is similar to yours and who seems legit.