IBM Develops Memory 100x Faster Than Flash

[Marcus Yam]

The memory of the future?

IBM Develops Memory 100x Faster Than Flash : Read more

 

[SpadeM]

With news like this you wonder why facebook is valued at 70 billion ...

 

[reprotected]

I went on Wikipedia to find out that this technology isn't as new as I thought it was.

 

[kriswitak]

I for one am glad that people are not just contempt with Flash Memory. As long as we keep pushing for higher standards, we can get close to the "future" we all keep imagining in our imaginations.

 

[michalmierzwa]

Its sad, but are we really going to see such vast hardware improvements in our lifespan? Its a pleasure to read about new discoveries, don't get me wrong, but its like waiting for Diablo 3 to come out, it takes such a long time that I am going to grow out of games form waiting... :-(

 

[master9716]

Wow . to bad its going to take years for this technology to reach our hands.

 

[klavis]

Hurray for IBM, they are continuing to show us why they are still relevant!

 

[hoofhearted]

There is alot of other "new tech" that they need to make money from before the newest tech makes it into our hands.

 

[winner4455]

[citation][nom]reprotected[/nom]I went on Wikipedia to find out that this technology isn't as new as I thought it was.[/citation]

I think this advancement is more about how they implemented this technology into an extremely fast and reliable memory rather than the technology itself. Think of it like a hard drive, we've had the technology for years but now we can have gigabytes of space instead of megabytes.

 

[enforcer22]

[citation][nom]razorburn[/nom]Soo, how many years till SkyNet goes online??[/citation]

Going to need a new terminator for the next date since it was suppose to have happen already

 

[razor512]

So what you're saying is IBM is lazy and wont be releasing the new memory any time soon.

 

[izmanq]

can't they just implement this sooner, why should we wait ? 😀

 

[razor512]

[citation][nom]izmanq[/nom]can't they just implement this sooner, why should we wait ?[/citation]

The people at IBM value the art of procrastination and spend a lot of time and money on procrastination.

 

[back_by_demand]

[citation][nom]razor512[/nom]The people at IBM value the art of procrastination and spend a lot of time and money on procrastination.[/citation]
The people at IBM are good at playing the long game, and seeing as they pump more into R&D than Intel makes in turnover I would defy you to do as good a job as they do.

Whilst other companies bang out a press release for a tiny incremental upgrade these guys are quietly building the future.

 

[alchemy69]

Now all they need is a bus that's 100x faster to go with it.

 

[michalmierzwa]

alchemy69 , agree with you 100%. When are there going to be authentic performance motherboards or hardware that actually keeps up to the speed? Why do we have theoretical top speed like in SATA HDD tech but real life test don't perform, anyone?

 

[azgard]

[citation][nom]michalmierzwa[/nom]alchemy69 , agree with you 100%. When are there going to be authentic performance motherboards or hardware that actually keeps up to the speed? Why do we have theoretical top speed like in SATA HDD tech but real life test don't perform, anyone?[/citation]

Latency.

 

[razor512]

Why not just move to a 500mm die and put 2TB L1 cache on the CPU.

Or just keep it the way it is but make it 2TB and allow the system to slowly cache the hard drive to it (many users rarely shutdown their systems so this can allow users to get the speed benefit of a storage like medium that can do over 200GB/s ?

Or take the standard ssd and move each memory chip to it's own channel, then make it larger and sell it for under $ 50

the manufacturing cost of a SSD is far less than that of a standard HDD.

IBM needs to stop being lazy and release a SSD with the new tech and allow users to max out their current bus and allow performance to be boosted when they get a new board with a faster bus.

 

[mauller07]

[citation][nom]michalmierzwa[/nom]alchemy69 , agree with you 100%. When are there going to be authentic performance motherboards or hardware that actually keeps up to the speed? Why do we have theoretical top speed like in SATA HDD tech but real life test don't perform, anyone?[/citation]

The tech never reaches its max due to overhead caused by error checking and send receive acknowledgments.

[citation][nom]razor512[/nom]Why not just move to a 500mm die and put 2TB L1 cache on the CPU. the manufacturing cost of a SSD is far less than that of a standard HDD.[/citation]

No other memory is as fast as sram used on a processor die so you would actually cause a performance hit to the processor, SSD's may be cheaper to manufacture but they are still a niece product and they recoup the low volume production and R&D through their high prices.

 

[techguy911]

Some scientist believe, that the advancement rate will simply begin to slow down because of increasing costs of developing new technology.
Notice that innovations are now years in the making 10+years due to costs so new ideas like this take close to 20 years to hit the shelf.

 

[alidan]

anyone read this as a thumb drive?

dont get me wrong, but i have a thumb drive slow as hell, but make it 100 times faster, and hell, better than the best ssds non raid, and possibly in raid too.

now, i want to know about this ssd are cheaper than hdd to make. can someone show me a link on this? and if its true, can someone show me a way to take my rage out...

 

[ik242]

IBM was messing with "successor of core memory" for quite some time, made plenty of promises with F-RAM announcements etc. But all they did is talk, there was never product on the market... I guess I'll be joining Missouri (the "show me" state)

 

[shin0bi272]

ha ha ha I posted this days ago. Way to be on the ball THG. I would suggest that the people who post news for THG get an RSS feed from engadget.com, dailytech.com and anandtech.com... might get some of these stories up faster.

 

[digiex]

The problem with this press releases is they never or it will take sooooo loooong to come out to the market.

 

[Guest]

I imagine this technology may be available in ten years or so. It seems like only yesterday (circa 1999) that a 4 GB SCSI SSD cost $2500.