Beware of PNY dirty tricks

M

MorgothPl

2[H]4U
Joined
Oct 13, 2008
Messages
3,020
It seems that PNY is playing bad. They sent to Tweaktown for review their Optima SSD, based on Silicon Motion controller. It got stellar reviews, so people started to buy PNY drives. Then, buyers realised, that they were not on SM controller, but instead they got Sandforce stuff.

PNY confirmed the change: "Yes we did ship some Optima SSD's with SandForce controllers, but only if they meet the minimum advertised performance levels (in most of the benchmark tests, LSI controllers outperform SMI controllers). The readers assumption that PNY has abandoned SMI controllers is wrong as we have been shipping mostly SMI controllers, but also utilizing LSI to fill in the gaps."

Kinda cheap trick - send one thing for reviews and sell the other without even telling people what they buy.

Source: http://www.tweaktown.com/blogs/Chris_Ramseyer/93/bait-and-switch-the-sad-state-of-ssds/index.html
 
Do they get the same performance? If so I don't see the problem since they really aren't selling you a controller, just an advertised speed.
 
I've been completely done with PNY for a couple years now. Got a couple SSDs from someone on here that I found out works at PNY. All three (exchanged one) he sent me died, two of them within a couple months. Shady company, shady employees.
 
Since a specific controller isn't listed in the specs for the SSD I don't see an issue here as long there isn't a performance hit. If PNY was using the SMI controller as a selling point I would take issue with it.
 
I don't think it's that simple. If I buy a specific product I expect to get the same each time.

It may produce the advertised numbers but, still there are specific conditions where the replacement performs worse.

I have no problem with it, if they differentiate it say this is Version B that comes with a different controller. That's it. And then I know not to buy it if I'm looking for a specific product.
 
If you buy a car that is tested publicly with a certain type of engine, you expect that type of engine. Buy a lambo with a 4-cylinder? not quite the same, of course, but this is misleading. shady practices.
 
So...do both hit the stated speeds/specs? Do the drives support the same features? If yes then why does it matter? As they said, to fulfill orders, they ship drives using multiple different controllers. As someone stated, they aren't using the controller as a selling point. Nor is it apparent that they cherry picked a drive for reviews.

Has anyone tested these same drives that use different controllers? What are the results?
 
I would be very annoyed too.
Thanks for the heads up, I'll keep my eye on this.
 
teh_chem said:
So...do both hit the stated speeds/specs? Do the drives support the same features? If yes then why does it matter? As they said, to fulfill orders, they ship drives using multiple different controllers. As someone stated, they aren't using the controller as a selling point. Nor is it apparent that they cherry picked a drive for reviews.

Has anyone tested these same drives that use different controllers? What are the results?
Click to expand...

Lets take a look at their "advertised specs". "Up to" 60K IOPS... that's it, that's all they advertise. Even on Amazon it's only 60K read/45K write. There's an awfully huge gap there in controllers that could do 60K IOPS thereby performing "as advertised" while still be massively slower than the review states since as long as whatever garbage they throw in there can reach 60K IOPS at some huge queue depth with some tiny block size (which lets face it isn't very high.)

Basically they throw a single, easily achievable metric to reach on the packaging and they're now free to use anything they have on hand as long as it can reach that not very high hurdle, and send out the faster versions to all the review sites.

Pure. Fucking. Bullshit.

PNY = OCZ = Kingston = Fuck'm all.
 
munkle said:
Do they get the same performance? If so I don't see the problem since they really aren't selling you a controller, just an advertised speed.
Click to expand...

Sandforce is more like a WOD - a write only device.

I mean it's find to put your games on, but...
 
-Dragon- said:
Lets take a look at their "advertised specs". "Up to" 60K IOPS... that's it, that's all they advertise. Even on Amazon it's only 60K read/45K write. There's an awfully huge gap there in controllers that could do 60K IOPS thereby performing "as advertised" while still be massively slower than the review states since as long as whatever garbage they throw in there can reach 60K IOPS at some huge queue depth with some tiny block size (which lets face it isn't very high.)

Basically they throw a single, easily achievable metric to reach on the packaging and they're now free to use anything they have on hand as long as it can reach that not very high hurdle, and send out the faster versions to all the review sites.

Pure. Fucking. Bullshit.

PNY = OCZ = Kingston = Fuck'm all.
Click to expand...
I guess you're right, their stated spec's are very loose, almost meaningless--hadn't seen them until now.

So what are the actual benchmark results between the same drive with different controllers? Is there a real-world perceivable difference? I.e., is there actual reason to be upset about this beyond some academic argument?
 
teh_chem said:
I guess you're right, their stated spec's are very loose, almost meaningless--hadn't seen them until now.

So what are the actual benchmark results between the same drive with different controllers? Is there a real-world perceivable difference? I.e., is there actual reason to be upset about this beyond some academic argument?
Click to expand...

Because getting SandForce instead of real controller is bad? :) They are only fast, until you throw less than like 50% capacity on disk and they can't properly work with TRIM functions.
 
teh_chem said:
I guess you're right, their stated spec's are very loose, almost meaningless--hadn't seen them until now.

So what are the actual benchmark results between the same drive with different controllers? Is there a real-world perceivable difference? I.e., is there actual reason to be upset about this beyond some academic argument?
Click to expand...

Since SF uses compression, uncompressible performance is almost always worse than competing controllers, and even with trim SF performance degrades pretty severely over time, so while fresh out of the box performance may appear similar the non-SF drive will probably be performing much better 6 months down the road. SF GC performance is abysmal so if your intended application can't use trim then forget about it 6 months from now.
 
Computurd said:
If you buy a car that is tested publicly with a certain type of engine, you expect that type of engine. Buy a lambo with a 4-cylinder? not quite the same, of course, but this is misleading. shady practices.
Click to expand...

If the advertisement didn't specifically say 8 cylinders, and the 4 cylinder somehow produced the same horsepower and torque as advertised, it wouldn't matter.

Car analogies almost always fail.
 
The car analogy for this instance would be they sent a V8 to car and driver for a review and advertised the car as "includes engine" and stuck "up to" in front of the torque and HP ratings.

What's funny is with "up to" in front of the specs the only people that can really complain that their drives don't perform as advertised are the ones that have drives that exceed those specs. "Hello PNY, yes I bought one of your SSDs and benchmarked it and it did 80k IOPS, 20k more than the advertised 'up to 60k IOPS' and would like a refund."
 
Guess what, I am a lucky one with SMI

I bought from Newe...(early adopter when released at the first time), total with coupon and rebate $80. mine is identical with optima that tweaktown reviewed.
 
This just furthers my personal "Intel or Samsung only, well Crucial is probably ok too" belief (and hence recommendation to friends, family, and anybody else who will listen). I'm sure Dell and other OEMs will eat this crap up because all they care about advertising is "has SSD", but for me and mine only the best will do. PNY is not the best.
 
You must log in or register to reply here.
Back
Top