backtop

Print E-mail del.icio.us 56 comment(s) - last by Clauzii.. on Jun 14 at 9:27 PM
64GB flash discs will be available in PATA and SATA flavors

We are finally starting to see some real technological breakthroughs in the area of mobile storage after a long period of stagnation. 2004 saw the rise of speedy 7200RPM hard drives while this year saw the introduction of perpendicular recording which allows data to be recorded in a smaller area. Just yesterday, DailyTech reported on Seagate's hybrid solution which pairs a traditional hard drive with perpendicular recording technology to 256MB of non-volatile flash for better performance, increased battery life and faster booting in Windows Vista.

Today, PQI is showing off new drives that mimic Samsung's 32GB Flash-SSD.  PQI, with the help of Samsung NAND flash memory chips, has new 64GB IDE and 64GB SATA 2.5" storage solutions for mobile users. The drives, which are due for release in August, are by nature more rugged, lighter, cooler and more efficient than traditional hard drives with a spinning disc. And best of all, there are absolutely no moving part so no more listening to your hard drive whir while you’re typing away and no more clicking and thrashing as you open up Photoshop or perform other disk-intensive operations.

Pricing has not been announced on the new 64GB IDE and SATA 2.5" drives, but rest assured that the new drives will be many times more expensive than even the fastest 7200RPM hard drives on the market today. As the market matures and more players enter the fray, we are sure to see a steady fall in prices. In fact, Samsung predicts that the global market for NAND flash based drives will increase from $540M USD in 2006 to over $4.5 billion USD in 2010. With growth like that, there will always be a premium for NAND-based disks over traditional hard drives, but the price differential should be much more manageable than it is today.

Comments     Threshold


This article is over a month old, voting and posting comments is disabled
By animedude on 6/8/2006 5:46:45 AM , Rating: 1
I think they have taken this fatal flaw into consideration before releasing it.


By animedude on 6/8/2006 6:09:07 AM , Rating: 4
Addon:
Longevity/Lifespan

"Unlike DRAM, flash memory chips have a limited lifespan. Further, different flash chips have a different number of write cycles before errors start to occur. Flash chips with 300,000 write cycles are common, and currently the best flash chips are rated at 1,000,000 write cycles per block (with 8,000 blocks per chip). Now, just because a flash chip has a given write cycle rating, it doesn't mean that the chip will self-destruct as soon as that threshold is reached. It means that a flash chip with a 1 million Erase/Write endurance threshold limit will have only 0.02 percent of the sample population turn into a bad block when the write threshold is reached for that block. The better flash SSD manufacturers have two ways to increase the longevity of the drives: First, a "balancing" algorithm is used. This monitors how many times each disk block has been written. This will greatly extend the life of the drive. The better manufacturers have "wear-leveling" algorithms that balance the data intelligently, avoiding both exacerbating the wearing of the blocks and "thrashing" of the disk: When a given block has been written above a certain percentage threshold, the SSD will (in the background, avoiding performance decreases) swap the data in that block with the data in a block that has exhibited a "read-only-like" characteristic. Second, should bad blocks occur, they are mapped out as they would be on a rotating disk. With usage patterns of writing gigabytes per day, each flash-based SSD should last hundreds of years, depending on capacity. If it has a DRAM cache, it'll last even longer. "

http://www.bitmicro.com/press_resources_flash_ssd....


By Pops on 6/8/2006 12:25:53 PM , Rating: 2
I have not had a hard drive last me more then 5-7 years before they fail. So the deminish rate on a HDD seems to go from 100% to 0%. If they dont outright fail in 5-7 years you need to replace them anyways due to size and speed limitations.

So if these flash drives can last hundreds of years I think we are ok, even if their capacity drops a little.


By CorrND on 6/13/2006 1:59:33 PM , Rating: 1
quote:
I have not had a hard drive last me more then 5-7 years before they fail. So the deminish rate on a HDD seems to go from 100% to 0%. If they dont outright fail in 5-7 years you need to replace them anyways due to size and speed limitations.

I'm very sorry if that's been your experience with hard drives. I've never had a hard drive fail on me -- EVER -- going all the way back to my first 50MB hard drive.

I definitely agree with you on your second point, though. That's been my experience: size and speed always dictate my new hard drive purchases.


By animedude on 6/8/2006 6:20:35 AM , Rating: 2
With proper management, a flash drive can last 56 years. Yes, that is intensive writing, too.


By ET on 6/8/2006 6:58:52 AM , Rating: 2
You're confusing "write" with "write cycle". A write cycle only happens when you need to erase data of a specific block to write new data there. To get what you describe you'd have to write once per second to the exact same place on the disk . You can see why this is unlikely.

As has been mentioned in another post, even when you do try to write to the same place, a well managed flash will actually write to another place. That's not only because there's a limited number of cycles, but because erasing a block is a slow operation.


By lemonadesoda on 6/8/2006 2:45:43 PM , Rating: 1
Hello???? Read the post please. We are talking about the swap/paging file.

Please don't tell me you are recommending a "constantly crawling", mega-fragmented swap file!


By AndreasM on 6/8/2006 3:05:27 PM , Rating: 2
quote:
Hello???? Read the post please. We are talking about the swap/paging file.


Fragmentation doesn't matter with flash memory as much.

quote:
It depends very much on your machine purposing. If you are running office apps, then you will not thrash the swap file as hard.

If you are running P2P, database app, or you are a graphics designer, you may indeed be working that little apging/swap file even harder.


Ditch Azureus in favor of utorrent and your P2P won't swap and hog all your RAM. I think database appers and graphics designers would be better of buying enough RAM so they don't need to swap. 2 GB is cheap nowadays, and should ensure that one doesn't need to chug along with constant swapping. :P


By lemonadesoda on 6/8/2006 7:03:24 PM , Rating: 1
...thanks for the tips. Unfortunately, laptop limited to 512MB. Otherwise laptop is perfect and would otherwise not need upgrading.


By lemonadesoda on 6/8/2006 8:06:25 PM , Rating: 2
I politely advise all people who think that a windows system doesn't do much HDD writing to download the following:

http://www.sysinternals.com/Utilities/Diskmon.html

Please remember the following high-write activities:

1./ Registry updating
2./ NTFS last access
3./ IExplorer temp folder
4./ Paging/swap file

e.g. Loading PHOTOSHOP takes about 1500 reads and 300 writes.

(and thats before I've started working on the pictures!!!)


By saratoga on 6/8/2006 11:10:00 PM , Rating: 4
quote:
e.g. Loading PHOTOSHOP takes about 1500 reads and 300 writes.


Thats almost nothing. Even if the 1 million write thing was a hard limit (its generally much higher), you'd have on the order of 100 trillion writes before the thing failed. Realistically, the number is much higher then that.

Do the math here. That thing writes at maybe 20MB/s. Maybe. 1,000,000 writes * 64,000,000 blocks / 20,000 blocks per second == 104 years of continuous writing. Not to mention, most people will occasionally want to read values too.

Also, I doubt that app can tell the difference between a cache hit and a cache miss, so its probably massively overreporting the number of writes actually commited.


By mindless1 on 6/13/2006 9:00:42 AM , Rating: 2
untrue

1) It can only write to free space. That's already lower (than traditional magnetic HDD) because of the inherant cost per GB.

2) it's not a "thing failed" when no single block can be written, it's a thing failed when the reliability of successful data retention is below the user's needs. It's not quite enough to have a failure then a remap where data is lost.

3) Writes are not 20MB at a time, a few KB here or there will use more space and there is still a filesystem table of some sort to update.

This doesn't make flash unuseable by any stretch but IMO the tendency for apps and windows itself to do so much writing needs to be rethought, done more conservatively, which I think may happen as flash becomes more predominant but it seems Vista is too near for it to happen effectively in this generation.


By saratoga on 6/8/2006 10:58:16 PM , Rating: 2
quote:
Fragmentation doesn't matter with flash memory as much.


For page files it doesn't really matter even with a hard disk since the page size will always be smaller then the block size on your disk, and since access is generally poorly localized anyway.


By sonoran on 6/8/2006 3:47:40 PM , Rating: 2
>Please don't tell me you are recommending a "constantly crawling", mega-fragmented swap file!

If you can afford these babies, my recommendation would be no swap file at all. Just toss a few GB of real DRAM into the machine and forget about swap files (which are only needed when you don't have enough DRAM to hold what's "in memory" anyhow).


By ET on 6/9/2006 9:37:49 AM , Rating: 2
Do you know what will happen to a memory location that changes every second? It will never be swapped out! Unless you're getting swap file thrashing, in which case you're probably taking time to sleep while your computer is responding to you. The point of a swap file is to keep on disk the areas of memory that aren't active.


By AndreasM on 6/8/2006 7:53:47 AM , Rating: 2
quote:
Assuming 1,000,000 maximum write cycles, and one write to swap/paging file per second, the device would last for 1,000,000 / 8hrs per day / 60 minutes per hour / 60 seconds per minute = 1 month.


One write to swap per second? I recommend buying more RAM if your computer swaps that much. :P


By lemonadesoda on 6/8/2006 2:47:46 PM , Rating: 2
It depends very much on your machine purposing. If you are running office apps, then you will not thrash the swap file as hard.

If you are running P2P, database app, or you are a graphics designer, you may indeed be working that little apging/swap file even harder.


By GoatMonkey on 6/8/2006 10:24:20 AM , Rating: 2
How about a solid state drive. They use regular memory with a battery backup.



By Trisped on 6/8/2006 11:42:55 AM , Rating: 2
Power drain when off would not be good for a laptop. The price would only be 3x as much as a strait NAND. You could also set it up so the drive would have to be re-filled from a standard rotating drive everytime the laptop was turned on, but that would greatly increase the load time.

Another solution would be to add 1 Gig of DRAM for use as a cach. When you are already spending over $2000 the extra 50 dollars isn't going to add that much. The problem would be when people imporperly turn off their computers. A battery backup system (or one that maintains power from the laptop battery) would allow the mantanancy of this cach while writing completed. Then it would power down. I think that right there would be the best solution.


By Xenoterranos on 6/9/2006 12:55:15 AM , Rating: 2
Samsung....YO SAMSUNG. Over here. Yeah, I have an idea...DO WHAT THIS GUY SAYS. That is all.


By PandaBear on 6/8/2006 2:20:58 PM , Rating: 2
This is why buying a good brand like SanDisk helps. I have personally know some shady manufactures using reference firmware design that doesn't do wear leveling, bad block mapping, or allocate reserved blocks and sell their flash cards as is. While you may be lucky, they get corrupted in many access pattern (like using it as swap or log). I would bet many of the fake memory stick are just using reference firmware (just so the controller manufacture can prove that the product works, but you need to add additional feature you need) to make a quick profit.


For anyone that's interested...
By lemonadesoda on 6/8/2006 8:55:19 PM , Rating: 2
I politely advise all people who think that a windows system doesn't do much HDD writing to download the following:

http://www.sysinternals.com/Utilities/Diskmon.html

Please remember the following high-write activities:

1./ Registry updating
2./ NTFS last access
3./ IExplorer temp folder
4./ Paging/swap file

On a 1GB RAM P4, here is what I did:

1./ Loaded MS Word, Excel, Powerpoint, Access
2./ Loaded Adobe Acrobat, Photoshop, GoLive, Indesign, Stock Photos
3./ Load IExplorer

then I shut them all down again.

NOTE THAT I DID NOT OPEN ANY DATA FILES OR ACTUALLY USE THE APPLICATIONS. JUST OPEN ALL, THEN CLOSE ALL

Result (according to Diskmon), approx:

Total READS: 60,000
Total WRITES: 5,000

Activity took: 90 secs

******* (This one is just for fun ;-) *******

With IExplorer open. Reset the counter

http to a thumbnail pr0n site.

Total READS: 20
Total WRITES: 1000

(no links were klicked. this was just the homepage)

Activity took: 20 seconds




RE: For anyone that's interested...
By saratoga on 6/8/2006 11:16:49 PM , Rating: 2
quote:
then I shut them all down again. NOTE THAT I DID NOT OPEN ANY DATA FILES OR ACTUALLY USE THE APPLICATIONS. JUST OPEN ALL, THEN CLOSE ALL Result (according to Diskmon), approx: Total READS: 60,000 Total WRITES: 5,000 Activity took: 90 secs


One word: disk cache.

Anyway, 5000 write / 90 second = 55 writes a second. 10^6*64*10^6/55 = 36898 years of continously open and closeing Microsoft office (provided you had no RAM I suppose other wise you'd cache many of those writes). How long were you planning on living?

I would like to advice all people who thought Windows doesn't do much HDD writing:

Congradulations on being correct.



RE: For anyone that's interested...
By lemonadesoda on 6/9/2006 6:20:16 AM , Rating: 2
saratoga... would you be so kind as to explain your math. Thanks


RE: For anyone that's interested...
By scruffypup on 6/9/2006 6:25:13 AM , Rating: 2
I second that!!!


RE: For anyone that's interested...
By lemonadesoda on 6/10/2006 6:49:47 PM , Rating: 2
Ok Saratoga, since you didnt answer my request for a math explanation... (and I think I can understand why you didn't), then I'll do the numbers for you:

64 GB = 64 x 10^9 bytes. It is NOT 64 x 10^12 as you show in your calculation. You were wrong by a factor of 1000x

But unfortunately for our longevity analysis, 64 GB = 125 million sectors (if one sector = 512 bytes), and when we "write" to Flash, whether just one byte, or a whole sector of data, we write the sector. Just like HDD.

So 125 million is the number we should use to base our calculations. And 125 million / 55 (writes per second) / 60 (secs per min) / 60 (mins per hour) / 24 (hrs per day) = 26 days

Yes. That was 26 days. Not 36898 years.

Whether a flash memory "block" is 128 bytes or 256 bytes (and not as large as a HDD 512 bytes sector) is moot, since the "writes" I had monitored were actually often 8 sectors at a time, ie, each "write" was in fact 4096 bytes and therefore up to 32 "blocks".

Now this survival period of 26 days assumes an even distribution of writes over the disk. However, in practice, data covers large portions of the disk, and system files, registry hives, paging files are going to have to be "dynamically remapped" within the existing unused portion of the flash drive. So the fuller the disk, the quicker it will die.

Moreover, because on a system like XP, 2003, or Vista we use NTFS (and not FAT like most USB drives) we also suffer from NTFS "transaction" logging and last access time. These will cause additional writes to "directory" sectors, MFT Zone, etc. etc.

Basically, for any demanding application running on Windows based boot drive, I cannot see how the flash disk will survive.

So therefore enthusiasts CANNOT use these drives for ultra-speed laptops. I can only see them being used as "damage proof" data drives, mobile devices, high G or "knock" environments, demanding low power or low noise situations, and other specialty circumstances.

****

I genuinely welcome a challenge, or correction, to my analysis and observations. But lets base them on facts and correct math! Thank you!


By lemonadesoda on 6/10/2006 8:25:50 PM , Rating: 2
For anyone interested in the ongoing discussion... I am of the impression that my numbers are way to low (in terms of lifespan) but the other number of 30000years+ is way to high.

Something is not right with the calcs. On both sides.

Now, somewhat curious about this effect, I'll go and monitor my HDD for a few day and do a more thorough "EXCEL" model of the numbers.

Will post results next week.


By ProviaFan on 6/12/2006 3:47:24 PM , Rating: 2
So you're saying that we write to every sector on the disk 55 times per second? Please clarify.


RE: For anyone that's interested...
By rainman1986 on 6/13/2006 2:49:47 PM , Rating: 2
Ok, I'll give it a shot.

64gb. Say you have 30gb free. So, 30x10^9 bytes. Say you are saturating a SATA1 cable (re-writing the same data over and over, but such that it will actually write to the disk, and not fill it), so 150mb/sec or 150x10^6 bytes written/sec. It will take 200 sec to write over every empty byte on the drive (30gb).
Not that long, but here's the trick. Even if the flash memory is only rated for 300k writes, it will take 60,000,000 sec or 694.4 days to use up the rated writes writes on the drive. This assumes a good leveling algorithm, and it doesn't relocate any of the static data.

At 1x10^6 write operations, now you have 2314.8 days to burn out the drive. I don't know about you, but I find it unlikely I'm going to write flood a SATA drive for over 6 years continuously...


RE: For anyone that's interested...
By lemonadesoda on 6/13/2006 6:31:39 PM , Rating: 2
Hi rainman. Thanks for the back-of-the-envelope. A different (and simple = elegant) approach.

And based on the way you set up the calculation, the results seem correct.

Problem is that we need to consider NOT linear writes, one byte at a time, systematically going through the disk, efficiently, sector by sector... but the following issue:

1./ Every "write" to the flash disk will reduce the lifecycle of THE WHOLE "BLOCK" by -1
2./ A block on a flash disk is 128, 256, or 512 bytes depending on how it was built.
3./ An NTFS sector is therefore (typically) 4, 2 or 1 "block"
4./ When reading/writing NTFS under typical Windows activities, usually, 1, 2, 4, 8, or 16 sectors are read/written at once. Other permutations are possible.

If data is nicely cached before being written (as suggested in earlier posts) e.g. the comment "stick in more memory to your laptop"... then as long as WRITE instructions are being cached, then this will help improve the lifetime.

HOWEVER, in data critical situations, write caching is not recommended. Caching writes is not what this type of "industrial hard core" memory is all about. It's about creating fail-safe, knock proof, power outage protected, enviroments. etc. etc.

AND WORSE. Since this is to replace an OS HDD, we are taking NTFS, and not FAT. That means every read is followed at some point by a write, due to the file stamping "last access time date".

So, if in your example, we were taking about read/write random access database, with, for whatever reasons, limited write-cache, then you need to divide your numbers by at least the block/sector size which is usually 128, 256 or 512 bytes.

This makes your life cycle 18, 9 or 4 days.


By ProviaFan on 6/14/2006 2:38:41 PM , Rating: 2
lemonadesoda, your calculations make much less sense than rainman's. You failed to address the the idea that _each and every_ block on the disk is somehow touched by a write operation _55 times per second_. That is simply untrue, not to mention impossible (given that these things write at perhaps 20MB/sec, significantly less than the 150MB/sec of SATA 1).

This discourse would be useful if you seemed interested in finding the truth (which is as you said probably somewhere between the two extremes), but you seem much too inclined to reject anything that doesn't give a short-life estimate. Reasonable estimates suggest that there will be absolutely no problems within 10 or 15 years under normal usage patterns... I would like to know if any enthusiast has kept an OS and applications drive for that long (I've had a 36GB Raptor for not quite two years, and I'm already itching for these things to come out so I can replace it).


By ToeCutter on 6/9/2006 5:16:49 PM , Rating: 2
Talk about much adieu about nothing.

Why not just turn off virtual memory and see for yourself? I ALWAYS limit my swap size to less the 768MB (with 1GB RAM) and my desktop responds significantly faster. And yes, even with Photoshop.

I also have a ThinkPad X41 Tablet with 1GB RAM and a pathetic 1.8 inch Toshiba HDD spinning at 4200 RPM. I disabled virtual memory and now it screams. I've been using it that way for months, averaging 44 processes at any given moment.

Monitor your swap usage. You'll discover that it rarely exceeds physical memory capacity, but Windows insists on continuing to use slow swap space.

I think what we have here is some creative marketing. Besides lowering power consumption, I see no tangible benefit to these flash drives.


Expensive
By Trisped on 6/8/2006 11:44:45 AM , Rating: 2
Is anyone as concerned as I am? No hard drive spin means we won’t know when the drive is being accessed. People can now install viruses without our ever knowing. Just kidding. I still am concerned though, as hard drive lights don’t do a very good job of indicating hard drive access now (some don’t even have hard drive lights). As long as they include a working hard drive access light (or better yet include a hard drive access meter like the task manager has for RAM) I will be happy. It is nice to know when the computer is waiting on the hard drive as opposed to frozen or waiting on the LAN connection.

If I was to guess, I would say this will be $2000+ since 4GB of flash is currently $100 and there is 20 times that much in this drive. Hopefully they will have 200-300MBs data transfers so we can get better use out of our SATAII interfaces.




RE: Expensive
By ProviaFan on 6/8/2006 12:51:40 PM , Rating: 2
Nope, I'm not particularly concerned about that. My 36GB Raptor makes a noticeable whine and seeks noisily; the 15kRPM SCSI and SAS drives that I lusted after before I realized how loud they were are even louder. A perfectly silent drive would be a welcome replacement - then just move everything to 120mm fans or water cooling for perfect silence.

Anyway, since the HDD light is controlled by the SATA or PATA controller and not the hard disk itself, I don't think you'll have to worry about losing it.


RE: Expensive
By Bladen on 6/8/2006 12:55:02 PM , Rating: 2
More like 20-30.

Unless they are internally multi channeled or RAIDed, so to speak.


RE: Expensive
By highlandsun on 6/8/2006 11:06:18 PM , Rating: 2
They just need a decent DRAM buffer, same as regular disk drives. Add in about 128MB of DRAM cache and it'll be great. Then it could actually sustain the ATA100 100MB/sec speed for over one second...


RE: Expensive
By lemonadesoda on 6/8/2006 2:51:58 PM , Rating: 2
There are taskbar apps that flash a box green/red depending on whether a HDD is being read/written. Just use on of those.

If you are blind, then you could install a beeper. LOL. j/k


Prizeless!!
By Clauzii on 6/10/2006 7:30:27 PM , Rating: 2
When this hits market, for me as a musician it will be a new dawn. Storage for Samplers, Synthesizers etc. will be rugged enough that even a drop on the stage does nothing to the storage. And for the Studio PC, total silence is a step closer, leaving the CPU to a slow fan. Passive PSU, HDs, GPUs - and I'll guess the CPU can be passive too, given the right design...

Almost at any price this will be a must for me.

Even @ 4-500 each.




RE: Prizeless!!
By Clauzii on 6/10/2006 7:38:32 PM , Rating: 2
Btw.:

Only thing missing is the 3D-PolyAcrylic Storage Lazer R/W Data Block instead of ALL noisy optical discs!

The only thing left ticking in the PC will be the ultrafast moving mirrors for focusing/positioning the lazer.


RE: Prizeless!!
By mindless1 on 6/13/2006 9:12:32 AM , Rating: 2
Actually if you are still waiting you probably won't jump on this either. TODAY you can in fact use Compact Flash with CF-IDE adapters and have 4-8GB per card (in the more economical sizes or more if you want to pay a premium), as well as LAN (NAS) storage for anything more on a studio PC. Is 4GB enough for most users? No if you're running semi-modern apps, but it's not like you have to have it all on one card either, just as with mechanical HDD you would have higher performance through subdivision of cards to different I/O streams.

Total silence was already obtainable if it were only the flash HDD holding you back.


RE: Prizeless!!
By Clauzii on 6/14/2006 9:27:57 PM , Rating: 2
My demand in storage on one device is like >40GB..

As for the total silence - it should be possible using a passive PSU, and some custom CPU-cooler then - given some cabinet construction with some kind of metal mesh on the bottom and top areas iow. convection cooling :)


This isn't new
By shady28 on 6/11/2006 5:20:07 PM , Rating: 4

I'm not sure what's new about this. Solid state disks have been around a while :

http://www.bitmicro.com/products_edisk_25_ide.php

Some of the performance comments here are way off base. Transfer rate is only one measure, and for most people it is the least important.

When you 'hear' your hard disk, you are hearing the head move back and forth. Access time is the major killer, ie the time to move that disk head to a spot on the disk and read the data.

If you look at the specs on that disk linked above, the access time is measured in micro seconds. Regular disks the access time is in milliseconds. These solid state disks have access times that are 1/20th what a normal hard disk has.

The result - look at the IOPs readings for these things. 1500-16000 IOPs on a 2.5" disk. Folks, that is 3x to 30x more IOPs than the fastest of 15K RPM SCSI disks out there.

In a 2.5" drive.

With the exception of applications where you need to just stream data (like playing or serving up a video), these things are killer. For multitasking scenarios, databases, or just switching back and forth between a lot of apps they are far superior to conventional hard disk technology.




RE: This isn't new
By Clauzii on 6/12/2006 6:12:51 PM , Rating: 2
'Dat was a nice link - thanks. Lifesaver.... :)


mmm...
By jaybuffet on 6/8/2006 8:27:53 AM , Rating: 2
..4 gb ram.. disable pagefile... flash based hard drive.. can't wait




RE: mmm...
By ProviaFan on 6/8/2006 9:14:17 AM , Rating: 2
I've already got the 4GB of RAM, so all that remains is for the price to fall enough so that I can get two of these in RAID 0 for my OS and applications drive (bringing high STR with incredibly low latency for insane performance). :P


I want one
By tk109 on 6/8/2006 10:53:57 AM , Rating: 2
Nice. I cant wait until they have one for the desktop that's comparable to a raptor. Wonder how long that will be.




RE: I want one
By surt on 6/8/2006 1:46:07 PM , Rating: 2
Well ... you could just get one of these. Storage capacity is pretty similar to a 740GD. Write performance is only about half what the raptor can do, but seek time is going to be near 0, so it could be better, depending on your work load.

It seems likely they'll be able to increase the write performance pretty rapidly because they can parallelize as they increase capacity. So i'd bet on the next generation (say in a year and a half) outperforming traditional hard drives in all performance paramaters (though they'll still be behind in both price and capacity).


How fast is it?
By Eug on 6/8/2006 11:21:15 AM , Rating: 2
What's the speed of the drive?




RE: How fast is it?
By surt on 6/8/2006 1:38:32 PM , Rating: 2
Their advertised transfer speed is 25MB/sec.
http://digitimes.com/bits_chips/a20060605VL204.htm...

I certainly look forward to one or two generations hence when they'll get ahead of ordinary hard drives in transfer spees.


By PAPutzback on 6/8/2006 4:12:29 PM , Rating: 2
Little to no heat and no noise. And if you store movies permanantely this will probably be cheaper than hard drives. If the faile rate isn't bad you won't need a backup strategy. I'd like to see a NAS server with few dozen slots for flash cards and just add them as you need them and as cost drops. It won't be long before outputs of the media center are the biggest part of the box.




By Quasmo on 6/8/2006 8:30:34 PM , Rating: 2
are you crazy? when you're able to get a Terrabyte hard drive for $500 (near future) why in the world would you purchase one of these for such a premium price? With the money you save you can buy that completely silent case from Zalman, and load it up with 4TB of hard drives, let alone one measley 64GB drive.


pricey??? .... of course..
By nangryo on 6/8/2006 5:41:14 AM , Rating: 2
If u expect to buy this baby now,
Expect to shell out as the same bucks or more than your notebook/laptop price.

sigh..

I guess we'll just to to wait... and wait... and wait... adn wait




Stop whining about longevity
By zsdersw on 6/8/06, Rating: 0
pricey??? .... of course..
By nangryo on 6/8/06, Rating: -1
"Nowadays, security guys break the Mac every single day. Every single day, they come out with a total exploit, your machine can be taken over totally. I dare anybody to do that once a month on the Windows machine." -- Bill Gates

DailyTech Poll
Which web browser do you use on your primary personal machine? 






44 Comments









botimage
Copyright 2009 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki