Thursday, June 17th 2010

Samsung Introduces High-speed 512GB SSD Utilizing New Toggle-mode DDR NAND Memory

Samsung Electronics Co., Ltd., the world leader in advanced semiconductor technology solutions, today introduced the first solid state drive (SSD) utilizing high-performance toggle-mode DDR NAND. The new 512 gigabyte (GB) SSD provides electronic data processing application designers with advanced performance and reliability for notebooks with premium value.

“The highly advanced features and characteristics of our new SSD were obtained as a direct result of an aggressive push for further development of our NAND flash technology, our SSD controller and our supportive SSD firmware,” said Dong-Soo Jun, executive vice president, memory marketing, Samsung Electronics. “Early introduction of this state-of-the-art toggle DDR solution will enable Samsung to play a major role in securing faster market acceptance of the new wave of high-end SSD technology,” he added.

The new 512GB SSD makes use of a 30 nanometer-class 32 gigabit chip that the company began producing last November. The toggle-mode DDR structure together with the SATA 3.0Gbps interface generates a maximum sequential read speed of 250 Megabyte per second (MBps) and a 220MBps sequential write speed, both of which provide three-fold the performance of a typical hard disk drive. At these speeds, two standard length (apprx. 4GB each) DVD movies can be stored in just a minute.

Samsung provides further gains in power efficiency by having developed a low-power controller specifically for toggle-mode DDR NAND. The resulting power throttling capability enables the drive’s high-performance levels without any increase in power consumption over a 40nm-class 16Gb NAND-based 256GB SSD. The controller also analyzes frequency of use and preferences of the user to automatically activate a low-power mode that can extend a notebook’s battery life for an hour or more.

The Samsung 512GB SSD makes use of reinforced 256-bit AES (advanced encryption standard) encryption to ensure higher security, protecting personal data against online hackers or undesired access when its host PC is misplaced and lost.

Samsung also provides streamlined boot time and application access with this new SSD, showing an approximately nine-fold improvement in random performance over HDDs. Also, an intelligent operation management function optimizes the SSD’s background working environment. Coupled with the popular Windows 7 TRIM feature the operation management function secures the reliability of the drive in write mode.

Samsung plans to begin volume production of the 512GB SSD next month. The new capacity extends Samsung’s range of SSD densities from 64GB to 512GB.
Add your own comment

17 Comments on Samsung Introduces High-speed 512GB SSD Utilizing New Toggle-mode DDR NAND Memory

#2
Animalpak
512 GB now can worth to buy SSD... but the price is still to much
Posted on Reply
#3
RejZoR
WHERE THE F IS THE PRICE EH? Stop bragging about how fast they are, we know that already. But wwe NEVER get to know the damn price. Geez is it so hard? Stupid vendors and their crappy PR...
Posted on Reply
#4

Well, i can give you an estimate, it's going to cost over 1k... so yah, take your mind off it.
Posted on Edit | Reply
#5
Completely Bonkers
Strange "case study" for its use; copying full length DVDs. Now who would have thought Samsung would use DVD ripping as a prima facie reason for buying a SSD :pimp:
Posted on Reply
#6
Jstn7477
I wonder how "durable" it is compared to a SandForce-based SSD.
Posted on Reply
#7
BazookaJoe
What about the IOPS?

Will somebody PLEASE think of the IOPS?!

Srsly tho - read/write looks good but means not much if the iops suck.. what ARE they?
Posted on Reply
#8
mstenholm
Did I hear 3 years warrenty? After one dead SSD I'm a bit reluctant going down that road again.
Posted on Reply
#9
btarunr
Editor & Senior Moderator
Most press releases will not talk about pricing, because they differ from market to market. Most certainly you won't get the 512 GB drive for $100.

by: BazookaJoe
Srsly tho - read/write looks good but means not much if the iops suck.. what ARE they?
IOPS and MB/s are different units for the same thing.
Posted on Reply
#10
jsfitz54
English Press Editor Needed

Yes, I agree with others that some important details were left out; price, availability in what markets and when an end consumer can expect availability (or is it being offered "to the trade", only? or initially?) as well as other technical facts. The rush to smallness is overrated. What is usefull now? In six months will I regret my decision? Should I wait? I think others, may begin to see my point.

After reading many press releases here on techPowerUp and other sites I find it increasingly hard to read Buzz Word written press releases that leave out these pertinent facts. Is price left out because Samsung has not finalized a "Value Pricing" scheme? They must know the production cost and other hard tangibles. By the time they do reach street level I will have forgotten this as an item I may have had an interest in. The Well is very full of eager pitchmen, so the price better be darn good or you lost me. Anyone know the famous P.T. Barnum quote I am thinking of?

Best Regards, All.
Posted on Reply
#11
BazookaJoe
by: btarunr
IOPS and MB/s are different units for the same thing.
Forgive me, but they MOST CERTAINLY ARE NOT - an ssd vs a mechanical hard drive may see real world IOP difference from ranges as great as 10 000 vs 70!!

Even if these two device had exact same read write speeds, the SSD with 10 000 IOPS would load and perform massively faster than the mechanical drive , in some cases loading applications more than 500% faster in ideal conditions.

And and an SSD that read/writes (just for example) 120/80 and IOP's 1000 will load software & cause your machine to run SLOWER than a SSD that read/writes 70/50 and IOPS 8 000 - as almost all files that are actually used to RUN/LOAD software are tiny and constitute virtually no drive bandwidth what-so-ever, but rather hundreds of thousands of individual read/writes, where IOPS are the only measurement of performance.

A typical Windows XP WINDOWS folder having ~ 4000 files UNDER 4Kb - and anyone familiar with drive benchmarking knows that puts you SLAP in the middle of your WORST read performance band for random reads - other wise known as LOADING YOUR OS...

At times like this IOPS are the only thing that can save you.
Posted on Reply
#12
Wile E
Power User
by: BazookaJoe
Forgive me, but they MOST CERTAINLY ARE NOT - an ssd vs a mechanical hard drive may see real world IOP difference from ranges as great as 10 000 vs 70!!

Even if these two device had exact same read write speeds, the SSD with 10 000 IOPS would load and perform massively faster than the mechanical drive , in some cases loading applications more than 500% faster in ideal conditions.

And and an SSD that read/writes (just for example) 120/80 and IOP's 1000 will load software & cause your machine to run SLOWER than a SSD that read/writes 70/50 and IOPS 8 000 - as almost all files that are actually used to RUN/LOAD software are tiny and constitute virtually no drive bandwidth what-so-ever, but rather hundreds of thousands of individual read/writes, where IOPS are the only measurement of performance.

A typical Windows XP WINDOWS folder having ~ 4000 files UNDER 4Kb - and anyone familiar with drive benchmarking knows that puts you SLAP in the middle of your WORST read performance band for random reads - other wise known as LOADING YOUR OS...

At times like this IOPS are the only thing that can save you.
BTA is correct. IOPS can be used to measure speeds at any byte size, just like MB/S. You aren't factoring in file size at all. IOPS does not mean random access, or exclusively small file access.

Measure both IOPS and MB/s at the same file size, and you'll see that they correspond perfectly.
Posted on Reply
#13
ToigaT
by: Wile E
BTA is correct. IOPS can be used to measure speeds at any byte size, just like MB/S. You aren't factoring in file size at all. IOPS does not mean random access, or exclusively small file access.

Measure both IOPS and MB/s at the same file size, and you'll see that they correspond perfectly.
BazookaJoe is correct. Sorry.

IOPS are not primarily or ordinarily used to measure transfer speeds. Consider the IOPS and transfer rates of HDD's and then consider the difference between normal HDD's and SSD's.

Yes, IOPS -DO- factor into the final end measurement of transfer speeds but that is not what BazookaJoe is talking about. It's a factor, not the end result. Being ignorant enough not to understand that a factor may in fact affect another area of performance or application other than the minimum/average/maximum/mean transfer rates is no excuse. Broaden your horizons.

I think you had both better go back to Wikipedia or Google and learn what IOPS actually are and what the significance of the difference between normal HDD's and SSD's are and just what this means to the end-user.
Posted on Reply
#14
shevanel
Welcome to TPU, why the harsh tone?
Posted on Reply
#15
ToigaT
by: shevanel
Welcome to TPU, why the harsh tone?
I apologise/apologize.
Posted on Reply
#16
DaedalusHelios
Heated debate for such a boring topic as storage. How long are sandforce or these units suppose to last? I am in the middle of considering using multiple in Raid in a corporate environment so durability and speed are a serious concern. Sorry to interrupt the debate but I know it will be fast enough. But will it be reliable with a low risk of corruption, or are traditional mechanical drives a better bet? :confused:

My other options are tradional Raid SAS setups but my capacity only needs to be 2TB total capacity. The data is going to be worth millions to the company so corruption cannot be an option. It will be accessed constantly and opened as a shared raid drive array on the network by multiple users.
Posted on Reply
#17
Wile E
Power User
by: ToigaT
BazookaJoe is correct. Sorry.

IOPS are not primarily or ordinarily used to measure transfer speeds. Consider the IOPS and transfer rates of HDD's and then consider the difference between normal HDD's and SSD's.

Yes, IOPS -DO- factor into the final end measurement of transfer speeds but that is not what BazookaJoe is talking about. It's a factor, not the end result. Being ignorant enough not to understand that a factor may in fact affect another area of performance or application other than the minimum/average/maximum/mean transfer rates is no excuse. Broaden your horizons.

I think you had both better go back to Wikipedia or Google and learn what IOPS actually are and what the significance of the difference between normal HDD's and SSD's are and just what this means to the end-user.
So you are focusing on my choice of terminology, instead of the point I was making that both are correct? That's called arguing semantics, and it doesn't meaningfully add to the conversation.

Reread my post and my choice of words.

IOPS and Read/write Speed correspond perfectly when the same files sizes are used. I did leave out that you have to use the same type of test as well, like sequential on both tests, and on the same drive, but I figured that could go without saying.

I wasn't comparing SSD and HDD, Bazooka brought that up, and it was not the point BTA was making. Although, Bazooka also made some correct points.

IOPS and MB/s are both fully useless without knowing the parameters of the testing, however. So this whole argument is actually completely moot anyway.
Posted on Reply
Add your own comment