They don't just lie about the capacity, but putting all the rave reviews about how fast it is on the old drive is a complete misrepresentation. The new drives that uses less NAND and thus, less channels don't even come close the the claimed speed in the reviews.
They don't just lie about the capacity, but putting all the rave reviews about how fast it is on the old drive is a complete misrepresentation. The new drives that uses less NAND and thus, less channels don't even come close the the claimed speed in the reviews.
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
at this point actually they address the capacity AND speed. their new 25nm drives use 16 chips, they simply stack fewer die in each. This makes it slightly more expensive to manufacture but returns the lost performance.
The only thing you lose out on is lifespan. and they still end up saving money in the end, just not as much.
I haven't read all three pages. Forgive me if I missed it.... BUT, one thing seems to jump out at me:
OCZ has put ALL of thier eggs into the SSD market (pulling themselves from the RAM market)... and this is how they handle thier publicity?!?
Any good marketing guy will tell you the TRUE cost of aquiring a new customer... Seriously, Is OCZ deliberately trying to make thier own jobs harder?
OCZ has put ALL of thier eggs into the SSD market (pulling themselves from the RAM market)... and this is how they handle thier publicity?!?
Any good marketing guy will tell you the TRUE cost of aquiring a new customer... Seriously, Is OCZ deliberately trying to make thier own jobs harder?
Add to that the growing hostility towards Sandforce's business decisions which lead to a plethora of castrated firmware versions.
For example, OCZ might be selling a SF drives with 60k IOPS but Patriot could be selling the same SF controller drives, but because they didnt make as large an order as OCZ, their firmware will only allow for 20k IOPS.
My next SSD will be either Marvell or Intel controlled (I don't like Indilinx stuff and won't be able to wait for Toshiba and Samsung to release a more potent controller).
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
yea, their new controller has 7 "versions" which use the same chip with different firmware/hardware (eg: super capacitor, more/less NAND, etc). And extra HIDDEN versions with the same name but different write caps due to contracts. We know of at least 3 different firmware version caps already for just one of the so called "models".
I am also not buying their products because of this.
I don't mind the performance caps, but match them to chip number.
flamenko
Senior member
Thats simply smart business. Its like saying we are not buying a Ford because they also have Mercury, Chevy because of GM, Honda...Acura and so on.
By simple engineering alone other drives cannot match the new SandForce because they dont use compression. It gives SF a huge advantage in the marketplace right now and until another starts to use compression, SF will be smiling all the way to the bank.
Other maunfacturers cannot get parallel reads and writes because you need a write ratio or amplification of less than 1. There is only one way to get this.
Can most survive with lesser? Absolutely. In fact, all but the true expert could never see the difference but this will make a huge difference in the enterprise sector where we just may see SandForce as the controller of choice.
Lets see if anything has been done by SandForce without their announcing of such for credit. A few months back we started seeing new firmware appear in ssd refreshes and new drives which increased their performance significantly. We hit them up on it immediately as we had not only received word, but also proved it in test after test. Still, we got no confirmation and there was no price increase whatsoever but now we see it commonplace for SF drives to reach into and exceed PCMark Vantage 40000 mark testing plateau.
Last edited:
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
which part of "hidden versions" do you not get?
Your analogy falls flat on two issues.
1. those are company companies not car models, and if you were listing their models, then you would find they are actually different.
2. It is not the "models" that bother me (as I have plainly stated), I have no issue with intel and AMD and nvidia making a chip and then clocking it at different speeds and selling them as different models; likewise I have no issue with sandforce doing it. It is the hidden contracts that give different firmware caps for identical name products that bother me.
I would be very upset to buy a 2010 honda accord EX edition from Dealer A only to find out that it is capped at half the horsepower of a 2010 honda accord EX edition from Dealer B
Last edited:
flamenko
Senior member
I understand what you are saying but it still relates to simple smart business. Intel puts out the i7 and except for the Extreme, each and every processor is physically identical. the 920/930/ and so on... all 100% identical but you are not after them.
SandForce does the same thing but is a bit more business savy (yes folks the SF 1200 and the SF1500 were as identical physically as the new SF 2000 series whether enterprise or client)and we say we dont like it.
With respect to the OCZ comparison, one needs to understand the relationship between SF and OCZ to understand why what we are discussing even exists. OCZ formed a partnership with SF when SF needed it most. OCZ, even with all of their issues, has gained some 80% marketshare and here was Sandforce not even a year plus ago, the brand new kid on the block.
Should the fact that OCZ can take prototype firmware and complete it on its own be concerning, maybe but any would jump at same. They just released the new drive with ROCK SOLID firmware and the very first thing they said to every one of us reviewers was to make sure we announced that this was prototype and unreleased firmware.
It might sound that I am a huge SF fan but I can assure you that this is not the case. I will praise any accomplishments that are brought out in ssd technology as I did with Intel/Samsung as well. I will also be the first to speak when something needs to be spoken of as was the notification I made that the SF-1200 and SF-1500 were exactly the same physically. People told me I was full of it and I even got a response from SF but they immediately realized that I did all on the cusp of speaking of the difference of the two and their still unannounced release of new firmware to the consumer.
Business is business. New manufacturers will charge huge premium to recoup business costs and then keep the price up because cost and demand. Business is business and we cant change that. What gives us the right to question business practice that is found to be within legal practice? Nada IMHO.
Love these exchanges with you Talt. They are historical I think by now.
Oh nd I would love to know the diff between the Yukon and Suburban
SandForce does the same thing but is a bit more business savy (yes folks the SF 1200 and the SF1500 were as identical physically as the new SF 2000 series whether enterprise or client)and we say we dont like it.
With respect to the OCZ comparison, one needs to understand the relationship between SF and OCZ to understand why what we are discussing even exists. OCZ formed a partnership with SF when SF needed it most. OCZ, even with all of their issues, has gained some 80% marketshare and here was Sandforce not even a year plus ago, the brand new kid on the block.
Should the fact that OCZ can take prototype firmware and complete it on its own be concerning, maybe but any would jump at same. They just released the new drive with ROCK SOLID firmware and the very first thing they said to every one of us reviewers was to make sure we announced that this was prototype and unreleased firmware.
It might sound that I am a huge SF fan but I can assure you that this is not the case. I will praise any accomplishments that are brought out in ssd technology as I did with Intel/Samsung as well. I will also be the first to speak when something needs to be spoken of as was the notification I made that the SF-1200 and SF-1500 were exactly the same physically. People told me I was full of it and I even got a response from SF but they immediately realized that I did all on the cusp of speaking of the difference of the two and their still unannounced release of new firmware to the consumer.
Business is business. New manufacturers will charge huge premium to recoup business costs and then keep the price up because cost and demand. Business is business and we cant change that. What gives us the right to question business practice that is found to be within legal practice? Nada IMHO.
Love these exchanges with you Talt. They are historical I think by now.
Oh nd I would love to know the diff between the Yukon and Suburban
Last edited:
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
You claim to understand but you are still clearly missing the point....
If I buy an intel i7 875K, from ANY MANUFACTURER, it is IDENTICAL both in design AND in performance cap.
There is no favorable reseller. I don't care that ALL i3, i5, and i7 chips are actually identical in design with different parts disabled and a different performance cap. Intel gives them DIFFERENT NAMES. (ok, I do care, but I do not condemn it... that IS smart business practice and it is actually honest/fair because the consumer is told exactly what they are buying)
Having 7 "model names" for the same chip (rather then sub names, eg: Athlon X2 ####) is confusing and I would prefer they use clearer sub-names... but overall it is actually a GOOD THING. That lets us know of different configuration. If they want to make exclusivity contracts, fine! All they have to do is come up with a few more model names. OCZ can have an exclusivity to the 2584 controller while others will get the 2583 and 2582 (depending on their tier) controller. (They only actually use 2582 of those 3 names)
Although a analogy would be if I buy an intel P55 mobo from ASUS or MSI it will perform the same, there is no performance cap that they must do a jig to mitigate, with varying caps across different manufacturers.
Or how about we cut to the chase... if I buy an INDILINX based SSD, I don't have to deal with this crap.
Business is business, fraud is fraud, fraud is neither smart business practices nor normal business practice.
Last edited:
I agree with you. I wanted to buy a sandforce drive and at the same time avoid the OCZ ones. However, I found that other sandforce SSDs are somehow inferior despite having the same controller model. To say the least, I was disappointed.
Ok flamenko, would you buy a vertex 3 or a patriot wildfire if they were the same price. I bet you would research whether the patriot had the 20k IOPS cap first. You can't just say, "this SSD uses a SF 1200 so it's all good".
Even Corsair was not happy with Sandforce when it was found out that sandforce were shipping different firmware versions to different vendors. Everyone at the time was advising Corsair F owners not to update their firmware because they might get capped.
Even Corsair was not happy with Sandforce when it was found out that sandforce were shipping different firmware versions to different vendors. Everyone at the time was advising Corsair F owners not to update their firmware because they might get capped.
I guess this is the dilemma when you have no close competition. So SF is delivering different FW to different manufacturers, but due to the demand for it, manufacturers will have to bite the bullet and release them since they are the fastest and NAND friendly controller makers.
Hi,
About "..right now, SF being a choice for enterprise.."... Are we talking about SF1200 (which is "right now"). If that's the case, then it's opposite: SF1200 has no value at all in enterprise (or business) segment. Why? because SF's "great specs" are based (hyped) on compressible data! Encrypt your disk (to protect valuable data) and write specs are slower than 5200rpm HDD.
IMO, SF1200 is ment for average consumer (performance- and price-wise). But even then, it's ment for having only your OS on (to speedy OS and app's startup). Because, most of the data "average" consumer is dealing with, are compressed (zip, mp3, jpg, docx, avi, cab, msi...) here, writting to HDD is faster again. Or are we dealing with txt files only?
I have SF1200 based SSD (supposed to be 120G
), so I'm talking from experience. And I still recommend to put OS & Apps on (SF) SSD. However, one should be carefull about SSD "brand". Because, as it is now, specs can be twisted at will. IMO, it's not SF to blame... it's just, some are not telling the truth...
Just my 2c,
Bogdan
About "..right now, SF being a choice for enterprise.."... Are we talking about SF1200 (which is "right now"). If that's the case, then it's opposite: SF1200 has no value at all in enterprise (or business) segment. Why? because SF's "great specs" are based (hyped) on compressible data! Encrypt your disk (to protect valuable data) and write specs are slower than 5200rpm HDD.
IMO, SF1200 is ment for average consumer (performance- and price-wise). But even then, it's ment for having only your OS on (to speedy OS and app's startup). Because, most of the data "average" consumer is dealing with, are compressed (zip, mp3, jpg, docx, avi, cab, msi...) here, writting to HDD is faster again. Or are we dealing with txt files only?
I have SF1200 based SSD (supposed to be 120G
), so I'm talking from experience. And I still recommend to put OS & Apps on (SF) SSD. However, one should be carefull about SSD "brand". Because, as it is now, specs can be twisted at will. IMO, it's not SF to blame... it's just, some are not telling the truth...Just my 2c,
Bogdan
taltamir
Lifer
- Mar 21, 2004
- 13,576
- 6
- 76
I think he means the new 2xxx series. Those are currently (until next month) the fastest for even compressed / encrypted data.
You are right that the SF1200 is worthless for enterprise solutions.
It seems Anand won't be covering the story at all now.
The OCZ V2 are still the only 25nm SSDs retailing at the moment. AFAIK, other companies have only announced or paper launched their versions. Why is this not newsworthy? Am I the only one who has been anticipating 25nm NAND nodes since IMFT announced their plans and especially after the Oct 2010 supposed launch date?
http://www.anandtech.com/show/4202/the-intel-ssd-510-review?all=true
The OCZ V2 are still the only 25nm SSDs retailing at the moment. AFAIK, other companies have only announced or paper launched their versions. Why is this not newsworthy? Am I the only one who has been anticipating 25nm NAND nodes since IMFT announced their plans and especially after the Oct 2010 supposed launch date?
I purchased my first SSD this week and after reading this thread OCZ is off the list forever.
Can't believe that they tried to pull this off and only came up with a solution when people dragged OCZ though the hot coals.
Sorry, never want to do business with a company that ever does this. I don't care how good Vertex 3 looks it's out of the race from my perspective.
I think that there is definitely a story here....... people are interested in how companies handle their mistakes.
Look at all the interest in the SB chip set issue. Intel set a very good example of what to do and people like me take notice (I have one fault MB by the way)
Not sure why an article with all the facts would not be of interest to many readers (and perhaps I don't have the facts right?)
Can't believe that they tried to pull this off and only came up with a solution when people dragged OCZ though the hot coals.
Sorry, never want to do business with a company that ever does this. I don't care how good Vertex 3 looks it's out of the race from my perspective.
I think that there is definitely a story here....... people are interested in how companies handle their mistakes.
Look at all the interest in the SB chip set issue. Intel set a very good example of what to do and people like me take notice (I have one fault MB by the way)
Not sure why an article with all the facts would not be of interest to many readers (and perhaps I don't have the facts right?)
Last edited:
jmet
Member
- Feb 19, 2001
- 36
- 0
- 76
I was one of the people who received one of the 25nm "bait and switch" drives, I returned it to Newegg once OCZ started the "Pay up" program and will never consider one of their products again because of the way they initially handled things.
Elixer
Lifer
- May 7, 2002
- 10,371
- 762
- 126
The only reason to care about 25nm is that it should bee cheaper than 3x nm. But it's not. 25nm Vertex 2 drives were initially a little cheaper with some rebates if you bought from specific e-tailers, but not really anymore.
Performance is no better with 25nm. There's no story with 25nm, because there's no benefit to the end user. Half node should cut prices dramatically, what we're seeing is that NAND is selling at capacity and demand for that NAND means the fabs can keep prices flat and earn their "Return on Investment" for their 25nm facilities that much quicker.
Which if Anand posts a story on it now, he would get as much negative feedback as positive for because it's so late.
That's not a story now. It's old news. He missed the window on that one.
That's not a story now. It's old news. He missed the window on that one.
jmet
Member
- Feb 19, 2001
- 36
- 0
- 76
Elixer
Lifer
- May 7, 2002
- 10,371
- 762
- 126
It is only old news for people that look at the forums, everyone else has no clue.
I have to say, that I did not realize that OCZ was doing this until very recently and recommended the drive to others. I check the front page regularly and this is one of those stories that I would have expected Anand to be completely on top of. He is kind of the go-to guy when it comes to SSDs.
I don't want to go all conspiracy theory, but I feel there has to be a reason why there wasn't a post about it other than just that he forgot.
I don't want to go all conspiracy theory, but I feel there has to be a reason why there wasn't a post about it other than just that he forgot.
You hit the nail on his head. It is because anandtech was the place to visit for in-depth information on ssds, which makes his silence even more jarring. What is strange other tech sites have kept quiet on the subject as well. To their credit, StorageReview did mentioned the problem on their site, http://www.storagereview.com/ocz_vertex_2_25nm_review_oczssd22vtxe60g
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter