Is Patriot Memory engaging in deceptive advertising tactics?

[TheDarkKnight]

I have been reading lately on some form posts that the only way you can use memory rated at 1.65V with a Sandy Bridge CPU is too run the memory at 1.5V. But I see on NewEgg that "Patriot" sells memory rated at 1.65V describing it as being P67 Platform ready. Why would they advertise their product this way when Intel has stated, according to some forum posts I have read, that 1.65V is too much voltage for the Sandy Bridge platform? Can somebody please clear this up for me once and for all? Making hidden adjustments in the BIOS to the memory voltages to make it "compatible" does not equate, in my opinion to being P67 platform ready. Seems kind of deceptive to me,

 

[Zap]

I think Intel allows for 10% over-voltage just like with socket 1366/1156.

 

[RU482]

what is the purpose of 1.65V DDR3?

 

[bryanl]

what is the purpose of 1.65V DDR3?

To profit by selling defective RAM chips that should have never been allowed to leave the factory.

 

[TheDarkKnight]

To profit by selling defective RAM chips that should have never been allowed to leave the factory.

Thats a pretty dumb statement Bryan. But since you are hungry for attention I'll give you some. Its not defective memory. Memory voltage requirements drop with each new generation of CPUs. But it is wrong for a memory manufacturer to state that their product which runs at 1.65V is P67 Platform Ready. Its really not. 1.65V memory is last generation memory.

 

[bryanl]

Thats a pretty dumb statement Bryan. But since you are hungry for attention I'll give you some. Its not defective memory. Memory voltage requirements drop with each new generation of CPUs. But it is wrong for a memory manufacturer to state that their product which runs at 1.65V is P67 Platform Ready. Its really not. 1.65V memory is last generation memory.

I didn't mean to start an argument, but I have to stand behind my original statement since all DDR3 chips are guaranteed for 1.5V (or less) by their manufacturers:


http://www.samsung.com/global/busine...1&xFmly_id=690

http://www.micron.com/partscatalog.html?categoryPath=products/parametric/DRAM/ddr3_SDRAM

http://www.elpida.com/en/products/ddr3.html

http://www.nanya.com/PageEdition3.aspx?Menu_ID=109&def=210&lan=en-us

http://www.psc.com.tw/english/product/product_1e.html

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=1Gb

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=2Gb

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=4Gb


If I'm wrong, please cite the data sheet of any DDR3 chip rated for 1.65V nominal.

 

[catilley1092]

Probably no more than the HDD manufacturers are. 1TB is NOT 1TB. That's deceptive as it gets.

I don't personally know about Patriot, but if RAM makers does as HDD makers does, that's deceptive, in my opinion.

Cat

 

[TheDarkKnight]

I didn't mean to start an argument, but I have to stand behind my original statement since all DDR3 chips are guaranteed for 1.5V (or less) by their manufacturers:


http://www.samsung.com/global/busine...1&xFmly_id=690

http://www.micron.com/partscatalog.html?categoryPath=products/parametric/DRAM/ddr3_SDRAM

http://www.elpida.com/en/products/ddr3.html

http://www.nanya.com/PageEdition3.aspx?Menu_ID=109&def=210&lan=en-us

http://www.psc.com.tw/english/product/product_1e.html

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=1Gb

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=2Gb

http://www.hynix.com/products/compu...enu3=01&RK=19&RAM_NAME=DDR3 SDRAM&SUB_RAM=4Gb


If I'm wrong, please cite the data sheet of any DDR3 chip rated for 1.65V nominal.

Your original assertion is that any DDR3 memory chip rated at 1.65V is a defective memory chip? Please explain your definition of the word defective, would you? Could we apply your definition to all CPUs sold by AMD and Intel that weren't capable of being sold according to their original design? I wouldn't exactly call memory intentionally designed to run at 1.65V a defective product. You consider anything sold out of JEDEC specs a defective product? Give me a break. It takes time for standards to actually become standards you know. Since when did people start classifying intentionally designed products that didn't conform to a set of standards as "defective" merchandise? Even if DDR 3 memory at 1.65V doesn't become a standard doesn't warrant calling it a defective product. Please choose your words carefully. Words actually have meaning you know.

 

[Hyperlite]

i have yet to purchase a 1.5v DDR3 product and they all seem to work juuuusssst fine.

 

[bryanl]

Your original assertion is that any DDR3 memory chip rated at 1.65V is a defective memory chip? Please explain your definition of the word defective, would you? Could we apply your definition to all CPUs sold by AMD and Intel that weren't capable of being sold according to their original design? I wouldn't exactly call memory intentionally designed to run at 1.65V a defective product. You consider anything sold out of JEDEC specs a defective product? Give me a break. It takes time for standards to actually become standards you know. Since when did people start classifying intentionally designed products that didn't conform to a set of standards as "defective" merchandise? Even if DDR 3 memory at 1.65V doesn't become a standard doesn't warrant calling it a defective product. Please choose your words carefully. Words actually have meaning you know.

A chip that doesn't meet all of its manufacturer's specifications is defective, and chip manufacturers reject all their DDR3 chips that fail testing at 1.5V or less. None are designed to require 1.65V for reliability. Failed parts are either destroyed or sold on the secondary market where they're bought by companies that then apply more lax testing standards, often using nothing but ordinary PC motherboards.

I'm not saying standards are the end-all, but I'd trust those of chip manufacturers much more than those of module manufacturers.

Why can't you provide a DDR3 chip datasheet that gives 1.65V as a nominal operating condition?

 

[zokudu]

Can we get a link to these "shady" practices? Also I believe AM3 board default to 1.65 volts.

 

[bradley]

1.65V DDR3 isn't defective, just binned differently. Plenty of high frequency, low voltage DDR3 modules (beyond JEDEC spec) have been released as the process improves. In theory power consumption is greatly reduced, in practice, not so much. The true reason to move from DDR2 to DDR3 is price, especially if you want 4GB modules.

 

[bryanl]

1.65V DDR3 isn't defective, just binned differently.

In what bin do chip manufacturers place DDR3 chips that need 1.65V to operate reliably when not overclocked?

 

[bradley]

Downgraded and uTT DDR are unbranded/untested and aren't meant for enthusiasts/overclockers, only eTT and major brand chips. Silicon is binned all the time.

As long as it's priced appropriately, advertised as such, and guaranteed at those rated speeds, I don't see a concern. Everyone wants to save a buck. Let the buyer beware though. Those fancy heatsinks only help to keep any silicon stencilling obscured.

 

[Voo]

A chip that doesn't meet all of its manufacturer's specifications is defective, and chip manufacturers reject all their DDR3 chips that fail testing at 1.5V or less. None are designed to require 1.65V for reliability. Failed parts are either destroyed or sold on the secondary market where they're bought by companies that then apply more lax testing standards, often using nothing but ordinary PC motherboards.

So you're saying all those Intel/AMD CPUs that have to binned because they fail some tests (ie needing more voltage than allowed to reach a certain frequency) are faulty and should be destroyed? Because that's the same process at work with RAM.
If you lowered the timings of the RAM you could probably downvolt it to 1.5V just fine - just a different way of binning.

As long as its priced reasonably and passes all tests at the advertised voltage I don't see a problem with that - people who want the higher binned RAM can just pay a few bucks more.

 

[WhoBeDaPlaya]

To profit by selling defective RAM chips that should have never been allowed to leave the factory.

Are you daft? You can run 1.65V DDR3 at 1.5V at a lower speed grade, ie. it is basically pre-OCed RAM, eg. I am running a bunch of Kingston DDR3-1600 1.65V sticks at 1333-1.5V on my Phenom II systems.
As long as one understands that, it is fine to purchase.

 

[Zap]

To profit by selling defective RAM chips that should have never been allowed to leave the factory.

Defective RAM is RAM that does not work. Operational voltage has nothing to do with it. If it did, then what does that make 1.35v DDR3?

But it is wrong for a memory manufacturer to state that their product which runs at 1.65V is P67 Platform Ready. Its really not. 1.65V memory is last generation memory.

Does any one remember when DDR3 first came out for the higher end socket 775 platform? There were kits on the market that were rated at 1.8v - yes, DDR3 and not DDR2!

Intel allows for a 10% voltage deviation on their IMC. "P67 ready" is just a marketing term, since in theory any memory within Intel's allowed voltage of 1.65v and otherwise following JEDEC specifications should work with anything from X58 to P67.

Memory IC (not to be confused with memory module) manufacturers will rate for 1.5v because that is the JEDEC standard.

Note that manufacturers of other computer components will vary voltage as needed. For instance, even Intel varies the voltage their CPUs require - some require more and some require less.

Memory modules sold at a voltage higher than 1.5v is marketing to overclockers and enthusiasts. In the past memory was sold at the nominal JEDEC standard. Then, overclockers figured out that more voltage = higher clocks. However, not all modules could go faster. Memory module manufacturers just capitalized on this by pre-testing (AKA binning) modules so the ones capable of higher speeds can be sold at a premium because they were better (AKA guaranteed).

Of course it is now getting a bit out of hand because I've seen even mediocre DDR3-1333 with higher voltage, but still that has nothing to do with anything being "defective" or "last generation."

This all being said, as I've recommended countless times before, buy 1.5v DDR3. Especially with Sandy Bridge, there is no benefit to buying expensive RAM. Just buy cheap $40-for-4GB DDR3-1333 at 1.5v.

 

[AdamK47]

Especially with Sandy Bridge, there is no benefit to buying expensive RAM. Just buy cheap $40-for-4GB DDR3-1333 at 1.5v.

I disagree. I'm willing to bet that $40 memory is not going to get anywhere near DDR3-2000+ when overclocking. Of course you already know this, but for your individual wants/needs YOU see no benefit in the more expensive RAM. You may not care about the 5% increase in performance, but others like me do. To sum up, there is a benefit.

 

[Zap]

I disagree. I'm willing to bet that $40 memory is not going to get anywhere near DDR3-2000+ when overclocking. Of course you already know this, but for your individual wants/needs YOU see no benefit in the more expensive RAM. You may not care about the 5% increase in performance, but others like me do. To sum up, there is a benefit.

Do you really get 5% more overall system performance? I'd rather spend the extra $35 (cost difference between 1333MHz 1.5v and 2000MHz 1.5v) on something else like a bigger SSD or the next step up GPU (or one with a better cooler) or just a faster CPU - one Micro Center promotion was bundling a Core i7 2600K with motherboard but not a 2500K, making the upgrade from 2500K to 2600K essentially around $40 for that week. Heck, right now there's an 8GB (2x4GB) DDR3-1600 1.5v kit that is $5 less than the 4GB (2x2GB) DDR3-2000 1.5v kit. I'd rather do that. (hot deal 8GB $70 with free 2GB USB)

"Fast" RAM used to be necessary BITD when overclocking meant pushing up RAM speeds as well due to stuff like how memory multipliers interacted with BSEL back in the Core 2 Duo heyday.

These days, IMO it is no longer necessary, especially for the budget overclocker.

 

[Voo]

I disagree. I'm willing to bet that $40 memory is not going to get anywhere near DDR3-2000+ when overclocking. Of course you already know this, but for your individual wants/needs YOU see no benefit in the more expensive RAM. You may not care about the 5% increase in performance, but others like me do. To sum up, there is a benefit.

But you won't see 5% increases in most applications (apart from SuperPI and whatnot) from using 2000+ instead of 1600 (the standard today, 1333 isn't really noticeable cheaper most of the time) memory. AT had tests on that for nehalem and the results were quite lacking really - site down atm so I'll hand in the benchmarks later..

 

[bryanl]

Defective RAM is RAM that does not work. Operational voltage has nothing to do with it. If it did, then what does that make 1.35v DDR3?

Better chips than for 1.5V.

1.65V operational voltage for chips that came off a production line producing parts meant for 1.5V proves voltage and defects are related.

 

[killster1]

Better chips than for 1.5V.

1.65V operational voltage for chips that came off a production line producing parts meant for 1.5V proves voltage and defects are related.

i dunno i agree with bryanl, defective doesnt mean doesnt work flat out just means it doesnt work as it was designed. i have no idea what the ram was designed for but i do know that patriot isnt the best brand =P me personally i would buy ram that is 5% faster for a few bux more sure 5% isnt much but come on now you goign to upgrade other parts anyway not like you are on a 5$ budget and a computer nut.

 

[Zap]

defective doesnt mean doesnt work flat out just means it doesnt work as it was designed.

Here is a simple flowchart:

. . . . Does it work?
. . . . . / . . . . \
. . . yes . . . . . no
. . . . | . . . . . . |
. . enjoy . . . defective

It isn't that difficult. If a manufacturer says "this memory needs X voltage to run at Y settings" and it works exactly as stated, why is it defective? It's like arguing that an 8oz can of soda is defective and the company is involved in deceptive marketing because all soda cans should be 12oz. Ridiculous.

Besides, why all the hate for Patriot?

Name one brand of enthusiast memory (AKA memory with heatspreaders/heatsinks) that has never marketed memory using voltages above JEDEC standards.

 

[bradley]

If lower binned silicon was defective, then so too would be anything not an AMD Black or Intel Extreme Edition product. The big guys do it as much as the little ones.

 

[aigomorla]

1.65V @ what memory clock?

This argument is moot without knowing memory clock.

Because roughly most memory is all same things.. and to get the higher speed, they just push more voltage, which is why they are 1.65V.