Kingston and PNY Caught Bait-and-Switching Cheaper Components After Good Reviews 289
An anonymous reader writes Over the past few months, we've seen a disturbing trend from first Kingston, and now PNY. Manufacturers are launching SSDs with one hardware specification, and then quietly changing the hardware configuration after reviews have gone out. The impacts have been somewhat different, but in both cases, unhappy customers are loudly complaining that they've been cheated, tricked into paying for a drive they otherwise wouldn't have purchased.
Re:And another on the ban pile (Score:3, Informative)
PNY has been on my ban pile since 2007, when I discovered that they wouldn't process a RMA on a SD card with a lifetime warranty without the original proof of purchase (verified by actually contacting support.)
It's sad to hear about Kingston, though. I've always trusted them and never had poor results.
Many, many companies do this ... (Score:5, Informative)
Good advice - when checking reviews for a product (e.g. on Amazon), always sort them by time and check how the ratings change. Many products get good reviews first, then it dives. You won't see this otherwise.
Re:And another on the ban pile (Score:4, Informative)
A bit more subtle than you think (Score:5, Informative)
It's a bit more subtle a scam than you think. Kingston/PNY haven't changed the specs of the product at all, all they did was ship hardware that's cheaper/closer to spec. That is, they never promised the crazy performance reviewers were getting, they just overbuilt the first run of components and then switched to something cheaper that still met spec requirements. Hardware manufacturers reserve the right to reformulate product all the time without indicating as much, so long as the spec is still the same. So basically, they spec a $100 box, but put $200 worth of components in it for the first few customers and review units. Once the good reviews go out, they pull the expensive components out of the box. But it's "technically not a scam" because they "technically never promised such a good deal", they just accidentally happened to give reviewers a good deal.
From a reviewer's point of view, however, I'd be incredibly skeptical of parts that perform too good compared to what they should be doing on paper. If you have something that is supposed to get 200mb/sec writes, but is actually getting 400 or more, then you should probably question the manufacturer and perhaps even score the product lower for being overbuilt, on the expectation that future hidden product revisions will stop overbuilding it.
Re:And another on the ban pile (Score:5, Informative)
Actually, if you read the article...
None of the drives died at their 200TB rated endurance, although the Samsung DID fail a data retention test. The Intel let go at 700+ TB of writes along with two other drives, but did so with plenty of advance warning and died in a way as to allow for one last read off of the data without corrupting it with a bad write. Hard to fault them there.
Re:And another on the ban pile (Score:5, Informative)
Both of you are mistaken - in a lot of cases this is simply because production runs are different.
Kingston and PNY are well known brands that buy a lot of excess parts. They build their storage using what stuff they have available. If Samsung overproduced flash chips that Apple can't soak up, they can either idle their factories (expensive), sell the excess on the market (depressing prices), or sell it to a company who does wholesale purchase of excess, like Kingston or PNY.
Option 3 is generally preferred because option 2 can impact contracts (i.e., if Apple sees Samsung is selling the same product for far less than they paid, they're going to demand a refund).
So basically, Kingston and PNY build products based on what they have on hand - perhaps today it's slower flash chips from Toshiba, tomorrow Samsung had an overrun and they can put in super speedy Samsung chips, etc.
While most electronics manufacturers generally try to go for the same parts over and over again (or with few substitutions - e.g., Apple buys hard drives from Seagate and WD (and their acquired companies like Toshiba and HGST), flash from Toshiba and Samsung, etc), there are other companies that build product based on what's on hand.
And heck, it's also one reason why Kingston and PNY product is so damn cheap - because by taking the excess stock and building what's on hand, they get parts at a good discount, but the variability in parts is much greater. Part manufacturers are happy because it means they don't have to dump product on the open market where their customers may demand the discounts as well, and they have someone to absorb overruns.
It's just like the McRib, really. McDonalds brings it back when pork prices are low and there's an excess they can obtain far cheaper than the open market (but they can take it all rather than buy it in small batches).
The downside is, of course, that product variability is high. Perhaps they get a stock of superfast Samsung, decide to use it to launch a new line, then Samsung has better supply management and the source of cheap excess disappears. Then they're now handling excess of a slower chip some other manufacturer has excess.
Heck, you can buy several different seemingly identical products and they'd all be different inside - the only way to guarantee would be to check the batch numbers.
And this applies to their products as well - RAM, SD cards, etc.
Andrew "bunnie" Huang actually did an analysis of this when they were buying SD cards in bulk from Kingston and getting issues. On Micro SD problems [bunniestudios.com]. It's a very detailed analysis of what REALLY happens with Kingston. PNY is probably extremely similar in behavior as well.
If you want consistency, you need to go with someone who builds it in, like Sandisk (Toshiba), Lexar, etc. who order parts direct, rather than an aggregator who builds simply based on what they were sold.
It's less a bait and switch, and more of "well, we had these parts today, and when we run out tomorrow we'll use those parts".
Re:And another on the ban pile (Score:5, Informative)
It's surprising. Kingston? I thought they were a good brand.
Kingston is a fairly serious company; but it's unfortunately not too surprising to see them involved in this story(and, specifically, with a NAND downgrade, rather than a controller swap). The company has its fingers in just about every step of the flash and DRAM supply chain, except actually fabbing the stuff(they do testing, they do IC packaging, they assemble DIMMs and the various USB, SSD, SD, CF, etc. flavors that people want flash in, they do support and logistics for PC outfits that want memory to shove into their products, and so on).
Unfortunately for them, the companies that do fab flash tend to have SSD interests of their own at this point. This puts Kingston in a slightly tricky position: too much on the line to just go full OCZ; but always having to scrape around to get flash at prices that they can still make a living on.
There's a very neat piece [bunniestudios.com] about the...interesting issues... that this causes with some of their SD products.
Re:Kingston selling shit USB3 flash keys (Score:5, Informative)
SanDisk does that same crap. There's a huge difference between the write speeds for Ultra and Extreme models even though they are both rated USB3. Learned that lesson the hard way.
Re:remember when Walmart was all "Buy American!" ? (Score:4, Informative)
I keep hearing about this so called race to the bottom (most often espoused by self proclaimed communists) yet my computer equipment today is a lot better than that which I owned 10 years ago (around the time I first started hearing about this race to the bottom.)
You're clearly not a laptop user. I fondly remember the days of 16x10 screens, caps lock and num lock LEDs, standard and stable keyboard layouts, inaudible CPU fans, etc.