Solid state will match hard-drive price eventually

iTS new step for consumer desktop. SiS had firsat SSD controllers Out last month & turned Out NOT that HOT, yet do support 4 DDRx lanes or 8 memory slots & few where designed, then OPPS.BAD TESTS. SSD Went bonkers & speed cut back by viral like problem, where SSD keeps on Pumping even if line cann't handle pace. So step Up & Be Warned, Its Completely Pristine SnowJob & ANYTHING May Come To Bear As Routes Are mapped into Unkown Field. Electricity takes Shortest Route(Least Resistance) To GROUND. So making 1,2,4 or More Gb/s transfers, Well NASA Microwave Towers run at 4 Ghz/s, thats covers every satelite & ground station, & nasaTv. whole show. So speed coming into some quality, when its controlled right inside desktop, Both ends of net will be running in harmonious sync. TS Drashek

I think you almost had something tinfoil hattish going on there, but instead rambled into, I just don't know man. One day you shall find a translator (or spouse to spellcheck) worthy of your genius!

Drashek said:

SiS made a chipset specked out for SSD usage, and it panicked and was slower than the standard SATA chipsets. And even though they ran slower than the SiS, they maintained a faster overall throughput. (Then something more about the above with the 'routes' thing about there will be more "SiS" failures to come about)

Then the last part, something about 4GHz through communication/computing etc (could be wrong there lol)

This seems like an obvious prediction ... until you think about it.

Yes, chips have been following Moore's law, with an exponential increase in density over time. But hard drive density has too, with a larger exponential term.

Solid state drives do catch up to the lower end of the hard drive business due to the large fixed cost of storing the first byte, but at the high end, it isn't obvious that flash will catch up to rotating magnetic media.

In brilliant fashion, Drashek brings us the concept of compute-acceleration! How could we have for so long ignored the GHz/s???

When 500GB SSD's are less than $100 (as SATA disks are now) won't a 500GB SATA disk be about $10? I don't think they'll ever be on par.

Actually Jim B, if you do the math it is quite obvious that it will catch and surpass magnetic media. The way platters are constructed, the density of bits in-track is more of a limiting factor than the density from track to track, because they have to be readable at speed even on the outermost part of the drive. But where reading in-track is mostly a solid-state operation limited by the media and head properties, it's becoming impractical to design a motor that can position the head accurately enough to achieve the same density on the other axis. So even if overall magnetic density increases exponentially, storage density increases at a linear-plus density, and speed increases at a sqrt(density) factor. Flash, following Moore, will increase in both storage and speed at exponential rates. HDs have really slowed down since about the 100GB mark, and flash has been accelerating since about the time 4GB thumb drives came out. If you plot the recent graphs it looks like density is doubling roughly every 10 months for flash vs roughly every 20 months for HDs. That's going to mean quick catch up time, and if we figure that in terms of volume a 3.5" desktop HD is 4x the size of a 2.5" HD that flash (and the market) seem to prefer, Flash is less than one iteration away from density parity (though far from cost parity). Given that higher "first bit cost" from the mechanics, I can't see how magnetic discs can hold a cost advantage for very long. Even if they had another breakthrough it would probably barely delay things since before long the performance gap will be so great disks will go the way tape did with its gigantic storage density advantage and huge latency disadvantage.

Seems Someone is Jealous: way got first Commentos In Here. Thats Double Points. When said above SSD SiS has 4 lanes, really Sis SSD has 4 Channels, 8 memory slots & controller didn't work,End of Generation ONE. so Faraday will try again. Sis makes Televison chips, its part of S.S. of listening to your comments thru TelCo port while TV is Running its connection. field differnt than desktop computing, especially large SSD (unless someone puts DVR in & net hookup). While NetBook SSD might of been SiS first controllers main market, pressure was too much. HDD Will NEVER be $10 New, for C Note you might get 4 Tb when SSD is $100.YOU GOT TO GET UP PRETTY EARLY IN MOURNING TO BEAT STeWie.

Unless Moore's Law stops "between 2013 and 2018 with 16 nanometer manufacturing processes and 5 nanometer gates"

128Gb SSD 230$ = $1.80 per gig
640Gb HD 60$ = $0.10 per gig

double every 18 months so, 3 years from now
512Gb SSD 230$ = $0.45 per gig

3 more years
1024Gb SSD 230$ = $.022 per gig

3 more years, frak can't cut atoms in half. The whole time HDD's advancing as well thanks to HAMR etc. I'm not saying it could never happen, but it will take some voodoo, and good luck predicting that.

The end of Moore's law can be seen as the challenges of EUV processing to get below 20nm raise significant cost issues while Magnetic areal denisty still has a strong ramp.

However, I don't need parity in fact there is real value for an SSD over magnetic media such that if it costs me $200 for a 500GB SSD and only $10 for a 500GB magnetic disk - I will buy both. When the 256GB SSD's hit the $100-$200 price range - I will put them in everything and leave the rotating media to backup.

And yes you can cut atoms in half (waht do you think quantum computing is lol).
Atom isnt the smallest particle and 16 nm isnt the limit for transistors or wharev they invent/select to replace today's tech.

Gldm -- I tried finding some credible source for density trends in both flash and hard drive technology, but couldn't. I found this article:

http://en.wikipedia.org/wiki/Flash_memory

which says it is a jump ball whether flash will ever catch up.

As for track positioning -- hard drives don't base the track position on a linear stepper motor like a floppy. It is a servo system, meaning a linear actuator tries to position the head, then it reads track data to figure out how far off it was, then it make a correction and repeats until it gets it right.

Both technologies are on exponential density improvement curves. The question is whether the slope of flash is steeper than the slope of HD areal density, and whether it can be maintained long enough to reach the crossover point.

Another thing in HD's favor is that doubling the amount of storage in flash doubles the cost; for HDD, adding a second platter doubles the storage at far less than double the cost.

Flash will certainly win in the low end, and may win in the high end eventually, but it isn't obvious.

Is all very nice, but how many read/write cycle(s) it can do within its lifetime?

Someone stating the obvious? Surely not!?
It's like saying we'll run out of Oil in the end! :O
Well DUH!

Brian Beard ... Blue Beard ... Sam Sung Blue ... everybody knows one ...

"When 500GB SSD's are less than $100 (as SATA disks are now) won't a 500GB SATA disk be about $10?"

No. The 500 GB drive will never go below ~$50; it will be replaced by a 750 GB drive for less than $100 dollars, then a 1 TB drive and so on.

There are big, fixed costs in hard-drives.

"There are big, fixed costs in hard-drives."

If you need 4GB of capacity, compare HDD and SSD prices.

The right question is not whether SSDs will become cheaper than HDDs. The right question is, at what capacity are SSDs cheaper than HDDs. Each year, that capacity number will increase, until at some point in the future all mainstream needs will be satisfied with SSDs.

Also, R&D dollars will be spent primarily on SSDs from now on. If R&D dollars were split 50/50 between SSDs and HDDs, we could expect HDDs to be more competitive with SSDs longer into the future. But almost everybody thinks that HDDs are a dead end within 10-20 years, so people don't want to sink lots of R&D money into them.