OC your SSD

Intel Demonstrates Overclocking SSDs At PAX Prime 2013 But Is It Worth It? - HotHardware

we should go and tell intel we DESPERATELY WANT THIS, NAOUUW!!

http://youtu.be/K24wo7S2tPs

com'on people.. get busy *cracks whip* discuss !

XD

 

Not worth the risk, in my opinion, unless you tweak before installing anything else.

Besides, Samsung 840 EVOs are still faster than those OCed Intels.

 

ahh but if this technology persists, we would get to overclock the NAND on the SSD's ... you know, just like OCing your ram, and the FSB...(being the controller, right?)

I see it as a LOOOONG TIME in the coming.

 

There is a lot of risk but if you use the SSD as a caching drive, then if it breaks, your data is not lost. However, like Hybrys said, Samsung EVOs are super quick at the moment and beats out any OC'd Intel drive.

 

what risk ?? our volatile memory is made of the same stuff on these disks, yet is waaaaay higher clocked and provides bandwidths of 20-30,000mb/s instead of our 83mhz ssd/nand drives (technically, NAND gates ARE arranged as RAM..same man)

theres little to no risk. when they make it overclockable, they will finally start using better NAND... anyways this stuff is incredibly hardy. You overclock your processor. You overclock your memory, you FSB's and other controllers.. why not your most redundant of memory ?

 

Because it's the most redundant piece. It stores information long term, so if it fails, you're most likely never getting that 60-500gb back. RAM has nothing on it that, if I was to burn out, would be irreplaceable.

Along with that, the risk is involved in the quality of the onboard chipset, controller, and CPU, and overclocking them. I'm not worried about pushing the RAM faster, but the components handling and organizing the RAM.

 

I do not believe you have a clue what your talking about. If something goes wrong, and say a single bit somewhere is off.. well, we already use a 2 bit correction system even on ssd... theres multiple layers of shadow volume protection, plus theres actual physical backups... nevermind electrostatically, we can fix 8:10 of all errors physically and electronically out of every block on any physical memory.

losing 60-500GB of data, your a moron and should stop pretending to know about things you do not, and at least admit it.

Your talking as if chucking the thing into a fire. these ssd's are using cheap low budget nand , something around the bluesets of 2004-2003-ish, about the speeds and capabilities of the first RDRAM or DDR yet working at 1:3 -1:6th the clocks. but still already using the same voltage/current. Our RAM back then, in the day, even without ECC didn't just FAIL running 5x the speed of today's SSD's.. on stock cooling... having a 1000 posts doesn't mean you know what your talking about. When memory fails, it doesn't whipe out a whole disk.. things go slowly, small sectors of what was accessed, in the easy case, a redundant switching of clocks could be had to prevent loss pass or within acceptable limits, however, once this tech, if it starts rolling, will be the future.

IOFusion and subpartners nowadays, adecade ago started using RAm as their harddrives but at 15-50x the speed we use our ssd's today. They had little risk in overclocking the volatile memory and backing it up to be redundant and safe and usable. Course they did this on pci and eventually the next gens buses, which we are moving sata 4 to (if its to be called that) almost 12 years later. The whole concept of OCing our ssd's is years behind. The knowledge and possibility ahs been there, and been used.

whatever propertiery rights associated with the concept, if I had to guess, is running out of some patents IMO, from when it was first conceived and now going to be marketed. OR it could have always been done, but alebiet high costs, and things like earthquakes in japan when it was first being done, in 1999 slowed the process...because back then, I was using a RAM drive. paid 450$ for the thing... with 512meg of memory, then sold it weeks later for 800 due to insane price increases... I was able to clock it back then, why not today ? (thought bus limitations were a , thankfully due to owning an alliances systems u5 with a decent backplane and scsi adapters, with a native built in u320 from adaptec, I got 300+mb's, but it was still bottle necked)

And that was 14 years ago. we're barely touching that today - derp. the components you say you ARE concerned about, are pathetic, these little cheap microcontrollers we use for our bus's - theres 10x more powerful available for less then 50$ worth of tech.

 

Tell that to all of the Sandforce random brickings. Information totally lost because of a controller failure.

SandForce finally patches elusive 2200 series SSD controller bug. OCZ issues firmware, others soon to follow. | PC Perspective

Nothing is error-proof. Overclocking has a higher theoretical chance to produce errors, because you're using the device beyond the factory specs. Thus, there is a higher risk.

 

Debate and disagreement is fine, but please refrain from flaming and name-calling. This is not appropriate or necessary, and it is a violation of forum rules.

 

My bad foxy.

Sorry hybris.

But, yeah, OCing is an issue when things weren't designed to be rugged, but the controllers we're using are really small and archaic, if they would use something more powerful, heck slap on a first gen ARM chip or something - about the same size but 10x more powerful.... we could get much more out of it.

I believe, its the future for us enthiusists...

 

S'all good, I see that it's just the passion that you have, that's why I didn't take offense to it.

I totally understand where you're coming from, but that doesn't change the fact that I'd be worried about OCing anything that I couldn't replace in an hour's trip to the store.

Now, if I was overclocking an SSD before putting the system on it, or simply using it as a cache or scratch disk? Sure! Makes perfect sense. If it has my OS and Games on it, there's not a chance I'd touch it.

It's simply my opinion anyway. Maybe it would be safe as all get out. Or maybe, just maybe, SSD companies could pull their heads out of the sand, lock down the OCing, tier/bin their SSDs better, and sell higher clocked SSDs at a premium. OCZ did it well with the MAX IOPS editions of the Vertex 3, so why can't other companies?

 

oddly enough, its actually quite pathetic the number of IOPs we're getting.

It means they're producing these drives for something like a few bucks each and making a KILLING on them. They haven't upped the bar, on their performance really, what they've done is limit their potential, so no one is forced to go and become competitive and spent tons of dollars in research trying to win the harddrive race, and instead, everyone can have a share of the market and mainly marketing determines their quality instead.

Theres been very little innovation when it comes to harddrives, its been this way for many many years. We're not much faster today then what was possible 10 years ago, before SSDs even.

However, I will admit, the price for the size has gone down.

 

The money is all about R&D, and about what they can do for us, that's fast enough for us not to complain.

That being said, there are innovations coming down the pipeline. Those PCIe/M.2 SSDs are supposed to be stupid fast. If I had a gaming desktop, I'd get one immediately, and I hope the next gen of AW/Sager has the capability to have one in it. MacBook Airs are showing 800+mb/s with theirs.

Speed is all relative anyway. I'm comfortable even being gated by SATAII, because I don't do a lot of file transfers, and I don't mind if my computer takes an extra 2 seconds to start up. My music still plays at ~8ms, and my games still load in reasonable enough times. I'm content.

 

I feel, memory and processors have begun to stale, in terms of progressing performance. I had a cpu that I could hit 70,000GBs back in 2003 10 years later Im using a IVRY bridge that scores the same... with the 3840QM..

in 2001 I was getting around 500-600MB's on my drive setup, though a lot less space then I have today, I had 4x 60GB SCASCSI drives at 15k rpm, BUT 13 years later, Im getting that with 4 drives that are solid state - and they cost me more today then they did back then - however - I have 2x the amount of space, very very few RAID0 errors nowadays, and they're MUCH smaller and quitter and my room doesn't sound like an aircraft carrier...

course, I had a 4400 GeForce 4 gfx, with 64mb back then.. and tomorrow I'll be using SLI 780M with 4GB's My graphics cards have improved drastically. So has RAM performance, up until about 5-6 years ago - where it simple stopped making headroom and has been dead silent / stale in terms of innovation or improving bandwidth/speed/ whatever.

 

You were getting 500/600 in a straight line but your 4k would have been orders of magnitude slower.

 

4k back in the day was in the 20's, not unlike today.

 

I find that hard to believe since even a modern 10k drive stuggles to reach 500 4k IOPs and an SSD is in the 10s of thousands.

 

well they were Seagate SCA SCSI drives, 30Gb ea. 15k rpm, the adaptec was a 2150 if I remember correctly, had 128mb eec pc133 ram on it, the card was doing dual u320.