Article: Intel’s Plans for 3DXP DIMMs Emerge
By: Howard Chu (hyc.delete@this.symas.com), December 1, 2018 12:52 pm
Room: Moderated Discussions
anon (spam.delete.delete@this.this.spam.com) on December 1, 2018 8:21 am wrote:
> Michael S (already5chosen.delete@this.yahoo.com) on December 1, 2018 7:56 am wrote:
> > Howard Chu (hyc.delete@this.symas.com) on December 1, 2018 5:23 am wrote:
> > > Personally I don't believe Optane DIMMs are going to replace DRAM, but tech
> > > like STT-MRAM has a good shot, if they ever reach parity to DRAM density.
> >
> > I don't quite get it. Why parity is enough? Doesn't new tech has to be at least
> > 1.5x to 2x denser in order to get a shot at replacement of old tech? Especially
> > something like MRAM that seems to get an inherent disadvantage in write energy.
> >
> > Or do you consider non-volatility an advantage?
> > I used to think about non-volatility as minor disadvantage, but in lights of all those new
> > security threats around I am starting to think that may be disadvantage is not so minor.
What security threats are you thinking of? I believe with non-volatility and
non-destructive reads, MRAM is immune to rowhammer-style attacks. (But I don't
have any references that claim to have tested or verified this.)
Greater density would be a nice bonus, but yes, I believe reaching parity
is enough, because non-volatility is a significant advantage in terms of
system reliability, performance, and power savings. Eliminating the DRAM refresh
cycle and its periodic bottleneck is a significant win.
It's nice that Optane DIMMs are denser than DRAM, but they're still far behind
in latency. Nobody is going to use a machine whose system RAM is 100% Optane,
it would be much slower than a DRAM-based machine.
> For a database it's an advantage because then you don't need to guarantee
> that the RAM can't lose power before all data has been written to disk.
> For everything else you're probably right.
For any application that ever needs to store persistent state - which is
pretty much all of them. Would let you avoid lengthy fsck procedures on
most filesystems.
> Michael S (already5chosen.delete@this.yahoo.com) on December 1, 2018 7:56 am wrote:
> > Howard Chu (hyc.delete@this.symas.com) on December 1, 2018 5:23 am wrote:
> > > Personally I don't believe Optane DIMMs are going to replace DRAM, but tech
> > > like STT-MRAM has a good shot, if they ever reach parity to DRAM density.
> >
> > I don't quite get it. Why parity is enough? Doesn't new tech has to be at least
> > 1.5x to 2x denser in order to get a shot at replacement of old tech? Especially
> > something like MRAM that seems to get an inherent disadvantage in write energy.
> >
> > Or do you consider non-volatility an advantage?
> > I used to think about non-volatility as minor disadvantage, but in lights of all those new
> > security threats around I am starting to think that may be disadvantage is not so minor.
What security threats are you thinking of? I believe with non-volatility and
non-destructive reads, MRAM is immune to rowhammer-style attacks. (But I don't
have any references that claim to have tested or verified this.)
Greater density would be a nice bonus, but yes, I believe reaching parity
is enough, because non-volatility is a significant advantage in terms of
system reliability, performance, and power savings. Eliminating the DRAM refresh
cycle and its periodic bottleneck is a significant win.
It's nice that Optane DIMMs are denser than DRAM, but they're still far behind
in latency. Nobody is going to use a machine whose system RAM is 100% Optane,
it would be much slower than a DRAM-based machine.
> For a database it's an advantage because then you don't need to guarantee
> that the RAM can't lose power before all data has been written to disk.
> For everything else you're probably right.
For any application that ever needs to store persistent state - which is
pretty much all of them. Would let you avoid lengthy fsck procedures on
most filesystems.
Topic | Posted By | Date |
---|---|---|
New article on Intel's 3DXP | David Kanter | 2018/07/23 09:02 AM |
New article on Intel's 3DXP | Groo | 2018/07/23 12:53 PM |
New article on Intel's 3DXP | Michael S | 2018/07/23 01:47 PM |
New article on Intel's 3DXP | Teemo | 2018/07/23 04:38 PM |
New article on Intel's 3DXP | Wes Felterw | 2018/07/23 08:41 PM |
Flash DIMMs = bad idea | David Kanter | 2018/07/24 03:31 AM |
Flash DIMMs = bad idea | Emil Briggs | 2018/07/24 05:30 AM |
Flash DIMMs = bad idea | David Kanter | 2018/07/24 05:49 AM |
Flash DIMMs = bad idea | Michael S | 2018/07/24 05:59 AM |
Flash DIMMs = bad idea | Emil Briggs | 2018/07/24 07:29 AM |
Flash DIMMs = bad idea | Doug S | 2018/07/24 07:49 AM |
price | Michael S | 2018/07/24 02:16 PM |
price | Doug S | 2018/07/24 02:32 PM |
price | Michael S | 2018/07/24 02:49 PM |
Flash DIMMs = bad idea | blaine | 2018/12/03 03:40 PM |
Flash DIMMs = bad idea | Wes Felter | 2018/12/04 11:07 AM |
Flash DIMMs = bad idea | RichardC | 2018/12/04 03:09 PM |
Flash DIMMs = bad idea | Michael S | 2018/07/24 05:51 AM |
Flash DIMMs = bad idea | Adrian | 2018/07/24 06:35 AM |
Flash DIMMs = bad idea | Ricardo B | 2018/07/24 08:24 AM |
Flash DIMMs = bad idea | bakaneko | 2018/07/24 05:55 PM |
New article on Intel's 3DXP | Etienne | 2018/07/25 04:02 AM |
New article on Intel's 3DXP | Howard Chu | 2018/12/01 05:23 AM |
New article on Intel's 3DXP | Michael S | 2018/12/01 07:56 AM |
New article on Intel's 3DXP | anon | 2018/12/01 08:21 AM |
New article on Intel's 3DXP | Howard Chu | 2018/12/01 12:52 PM |
New article on Intel's 3DXP | Adrian` | 2018/12/01 02:43 PM |
New article on Intel's 3DXP | Adrian | 2018/12/01 10:05 PM |
New article on Intel's 3DXP | Howard Chu | 2018/12/11 04:17 AM |
New article on Intel's 3DXP | Adrian | 2018/12/11 04:42 AM |
New article on Intel's 3DXP | Maynard Handley | 2018/12/11 07:20 AM |
New article on Intel's 3DXP | wumpus | 2018/12/11 08:36 AM |
New article on Intel's 3DXP | Anon | 2018/12/11 04:21 PM |
New article on Intel's 3DXP | Maynard Handley | 2018/12/11 04:32 PM |
New article on Intel's 3DXP | Anon | 2018/12/11 11:29 PM |
New article on Intel's 3DXP | Maynard Handley | 2018/12/12 10:32 AM |
New article on Intel's 3DXP | wumpus | 2018/12/12 11:07 AM |
New article on Intel's 3DXP | Maynard Handley | 2018/12/12 11:41 AM |
New article on Intel's 3DXP | Anon | 2018/12/12 02:55 PM |
New article on Intel's 3DXP | Anon | 2018/12/12 02:49 PM |
New article on Intel's 3DXP | Anne O. Nymous | 2018/12/12 12:14 AM |
New article on Intel's 3DXP | anon | 2018/12/12 05:28 AM |
New article on Intel's 3DXP | Maynard Handley | 2018/12/12 10:26 AM |
New article on Intel's 3DXP | Anne O. Nymous | 2018/12/12 01:10 PM |
New article on Intel's 3DXP | innocent bystander | 2018/12/12 09:34 PM |
New article on Intel's 3DXP | anon | 2018/12/12 01:42 PM |
New article on Intel's 3DXP | Howard Chu | 2018/12/02 04:53 AM |
New article on Intel's 3DXP | Adrian | 2018/12/02 06:01 AM |
New article on Intel's 3DXP | Howard Chu | 2018/12/02 10:34 AM |
Intel's 3DXP availability | Etienne Lorrain | 2018/12/03 03:50 PM |