Intel's New Desktop SSD Is an Overclocked Server Drive

← Back to Stories (view on slashdot.org)

Intel's New Desktop SSD Is an Overclocked Server Drive

Posted by samzenpus on Thursday February 27, 2014 @03:09PM from the new-kid-in-class dept.

crookedvulture writes "Most of Intel's recent desktop SSDs have followed a familiar formula. Combine off-the-shelf controller with next-gen NAND and firmware tweaks. Rinse. Repeat. The new 730 Series is different, though. It's based on Intel's latest datacenter SSD, which combines a proprietary controller with high-endurance NAND. In the 730 Series, these chips are clocked much higher than their usual speeds. The drive is fully validated to run at the boosted frequencies, and it's rated to endure at least 70GB of writes per day over five years. As one might expect, though, this hot-clocked server SSD is rather pricey for a desktop model. It's slated to sell for around $1/GB, which is close to double the cost of more affordable options. And the 730 Series isn't always faster than its cheaper competition. Although the drive boasts exceptional throughput with random I/O, its sequential transfer rates are nothing special."

111 comments

Min score:

Reason:

Sort:

sequential transfer by edmudama · 2014-02-27 15:17 · Score: 5, Insightful

Hard for any SATA drive to distinguish itself on sequential transfers, given that SATA is capped around 550MB/s

--
More data, damnit!
1. Re:sequential transfer by Luckyo · 2014-02-27 15:28 · Score: 5, Interesting
  
  Pretty much this. Vast majority of SSDs on the market today are very similar in terms of speed in normal usage, because the bottleneck is now in SATA. You can overclock it all you want, but you'll need to start pushing disks to PCI-E or similar bus for it to start to matter.
  And then there's the whole issue of "does it really matter when it's this fast on desktop?"
2. Re:sequential transfer by Joe_Dragon · 2014-02-27 15:41 · Score: 1
  
  To bad that DMI is only pci-e 2.0 X4 and all chipset is has to go over that. And Intel chips have limited PCI-E IO.
3. Re:sequential transfer by houstonbofh · 2014-02-27 16:37 · Score: 1
  
  Hard for any SATA drive to distinguish itself on sequential transfers, given that SATA is capped around 550MB/s
  Which is why every fast SSD has data rates for SATA2 and SATA3. SATA3 is a lot harder to cap. But even then, for the ultrafast are SSD cards, and no SATA involved.
4. Re:sequential transfer by Blaskowicz · 2014-02-27 16:52 · Score: 1
  
  You can cheat with bridges, that may be acceptable, or play with the PCIe - bringing the graphics card down to PCIe 8x 3.0 so you use the other 16x physical slot is a good deal.
  Now the trouble is the expected new connector, SATA Express (ought to get common and cheap, 2x PCIe 2.0) won't be present on yet-to-be-launched Z97 and down chipsets. Maybe because of that DMI bus limit or because they were lazy and risk adverse. It was formerly announced it'd be released with that chipset.
  So the arrival of SATA Express drives and expasion cards risk getting delayed. Bummer if you wanted it.
5. Re:sequential transfer by jones_supa · 2014-02-27 17:29 · Score: 4, Insightful
  
  Hard for any SATA drive to distinguish itself on sequential transfers, given that SATA is capped around 550MB/s
  Which is why every fast SSD has data rates for SATA2 and SATA3. SATA3 is a lot harder to cap. But even then, for the ultrafast are SSD cards, and no SATA involved.
  The 550MB/s is for SATA3 and has been capped for a good time already.
6. Re:sequential transfer by Anonymous Coward · 2014-02-27 21:07 · Score: 0
  
  Huh?
  The PCIe part of SATA express in the final SATA 3.2 spec is 2 lanes at 8GT/s each... that's PCIe 3.0.
7. Re:sequential transfer by Blaskowicz · 2014-02-27 21:44 · Score: 1
  
  I didn't know that but I'm pretty sure planned or existing designs are at PCIe 2.0, ditto the M2 format.
  See this Asus Z87 motherboard prototype
  http://www.tweaktown.com/artic...
  SATA express was for Z97, and it was canned.. well that doesn't prevent you from including it on a piece of hardware. I have no problems imagining the spec supports PCIe 3.0, it's an easy and logical upgrade. Simply what exists now runs at 2.0, and 2.0 is what comes out of the motherboard's chipset.
8. Re:sequential transfer by Anonymous Coward · 2014-02-27 23:00 · Score: 0
  
  Looked into it some more... I think I know where the confusion comes from.
  The SATA express draft from 2012 and early SATA 3.2 drafts only mention 2.5 and 5GT/s PCIe 2.0.
  While the final SATA 3.2 spec from mid 2013 has this to say:
  "SATA Express supports x2 PCIe, Gen2 and Gen3, signaling and references PCIe Base and Card Electromechanical (CEM) specifications Rev 3.0 for electrical requirements"
  In light of that it's not exactly surprising that current prototypes are PCIe 2.0 only.
9. Re:sequential transfer by Anonymous Coward · 2014-02-28 01:19 · Score: 0
  
  I wouldn't worry about it too much, while the drive is fast, it is not as fast as some current desktop drives.
  http://techreport.com/review/26086/intel-730-series-solid-state-drive-reviewed
10. Re:sequential transfer by jon3k · 2014-02-28 02:26 · Score: 1
  
  Well, SATA Express and NVMe will be here soon and we should see a pretty massive jump in sequential throughput.
11. Re:sequential transfer by Anonymous Coward · 2014-02-28 06:14 · Score: 0
  
  Isn't normal usage less about sequential transfers and more about IOPS and random read/write performance? In which case SATA isn't a bottleneck and the vast majority of SSDs on the market today aren't very similar.
12. Re:sequential transfer by houstonbofh · 2014-02-28 06:23 · Score: 0
  
  Hard for any SATA drive to distinguish itself on sequential transfers, given that SATA is capped around 550MB/s
  Which is why every fast SSD has data rates for SATA2 and SATA3. SATA3 is a lot harder to cap. But even then, for the ultrafast are SSD cards, and no SATA involved.
  The 550MB/s is for SATA3 and has been capped for a good time already.
  Did you miss a part of my post?
13. Re:sequential transfer by Anonymous Coward · 2014-02-28 06:31 · Score: 0
  
  True, but sequential performance beyond 550MB/s is pretty much irrelevant for most purposes. The really cool thing about SSDs is their awesome random access time. The remaining factor that interests me the most is reliability. Give me an SSD that can last a minimum of 5 years of database server use and I will be eternally grateful. Samsung 840 Pro is looking good so far.
14. Re:sequential transfer by Hal_Porter · 2014-02-28 06:59 · Score: 1
  
  Look at page 4
  http://techreport.com/review/2...
  The 730 drive is in the middle of the pack for sequential reads. None of the drives reach 550MB/s.
  
  --
  echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
15. Re:sequential transfer by jones_supa · 2014-02-28 20:26 · Score: 1
  
  I didn't miss that. I was talking about only the SATA part.
16. Re:sequential transfer by Anonymous Coward · 2014-03-01 05:38 · Score: 0
  
  Protocol overhead is already taken into account in the 600 (or 300, or 150) number. 6 Gbps = 750 MBps 750 MBps * (8b/10b) = 600 MBps
17. Re:sequential transfer by List+Lurker · 2014-03-01 09:43 · Score: 1
  
  Pretty much this. Vast majority of SSDs on the market today are very similar in terms of speed in normal usage, because the bottleneck is now in SATA. You can overclock it all you want, but you'll need to start pushing disks to PCI-E or similar bus for it to start to matter.
  And then there's the whole issue of "does it really matter when it's this fast on desktop?"
  Although i'm running 840 Pros (b/c of cost), i did contemplate grabbing some hard-to-get Hitachi SAS (6Gbs) SSD's to run off my current LSI 9300-8i 12Gbs SAS HBA... don't have the link handy, but read one write-up where these Hitachi's were running near as fast as the PCI-e stuff Apple is running in their new Mac Pro ... i.e., completely smokes existing SATA SSD throughputs ... maybe these SAS variants will come down in price (i can hope can't i??)
Performance consistency versus peaks by Wrath0fb0b · 2014-02-27 15:29 · Score: 5, Interesting

For many (but certainly not all) applications, especially when it comes to UI, what matters is 95% worst performance, not peak throughput. From the Anandtech review, that's where this drive really shines.
Different tradeoffs have to be made for different workloads -- it can't be boiled down to a single (or even a set of) number(s). Some applications are far more tolerant of worst-case performance than others.
1. Re:Performance consistency versus peaks by Anonymous Coward · 2014-02-27 15:37 · Score: 2
  
  For many (but certainly not all) applications, especially when it comes to UI, what matters is 95% worst performance, not peak throughput.
  Applications of porn acting being the notable exception.
Overclocked? by mc6809e · 2014-02-27 15:38 · Score: 4, Insightful

Running something at the speed it was designed and verified to run at by the maker isn't overclocking.
1. Re:Overclocked? by Jeremi · 2014-02-27 16:25 · Score: 2
  
  Running something at the speed it was designed and verified to run at by the maker isn't over clocking.
  If Intel designs a component to run at speed X, then later finds out that it can run some of those components at speed 1.5X, and verifies and sells them at that higher-than-rated speed, I think it's fair to say Intel is over clocking. The only difference is that in this situation, the warranty will be honored if it stops working.
  
  --
  
  I don't care if it's 90,000 hectares. That lake was not my doing.
2. Re:Overclocked? by Anonymous Coward · 2014-02-27 16:42 · Score: 1
  
  If Intel designs a component to run at speed X, then later finds out that it can run some of those components at speed 1.5X, and verifies and sells them at that higher-than-rated speed, I think it's fair to say Intel is over clocking. The only difference is that in this situation, the warranty will be honored if it stops working.
  You mean like how Intel has been increasing processor frequencies for years by now?
3. Re:Overclocked? by Kjella · 2014-02-27 21:51 · Score: 1
  
  Chips have tolerances which means there's a spread on how fast they'll run. Binning is not overclocking, if Intel finds a i7-4770K that can run 100MHz above the rated speed they won't sell it now. They'll put it in storage and wait until they have enough of them then launch a new model i7-4790K (coming to you in Q2). People analogy, if you select the best people to go into elite forces and the rest in the regular army you've binned, but not overclocked. That'd be more like putting them on drugs to amplify their combat ability at the long time cost of their health.
  
  --
  Live today, because you never know what tomorrow brings
4. Re:Overclocked? by ericloewe · 2014-02-27 22:07 · Score: 1
  
  I seriously doubt they are binning these differently than the enterprise version.
5. Re:Overclocked? by AmiMoJo · 2014-02-28 00:24 · Score: 1
  
  It's not overclocking, it's thermal and power management. For example Intel sells the same CPUs for use in a variety of different devices, and allows the manufacturer to set thermal and power limits via their BIOS. An ultrabook laptop might set the limits much lower than a full size laptop with bigger battery and more cooling.
  In this case Intel set the server drives up for server workloads and realized there was no point in having the high clock rates, so chose instead to generate less heat and save power. Probably improves reliability too. For desktop workloads they decided to go with the a higher clock, but it is still within the design spec.
  
  --
  const int one = 65536; (Silvermoon, Texture.cs)
  SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
an excellent point by raymorris · 2014-02-27 15:59 · Score: 1

That's an excellent point, and a metric I hadn't paid much attention to despite the fact that I run quite a few drives, including one storage pool of 28 drives and growing.
1. Re:an excellent point by Anonymous Coward · 2014-02-27 18:39 · Score: 1
  
  Dude, I just saw your soul - angry, dweeby nerd to the core.
The caps are electrolytic by williamyf · 2014-02-27 16:05 · Score: 0

Not supercaps, no, electrolytics.
What happens if your superduper SSD develops bad cap syndrome?
https://en.wikipedia.org/wiki/Capacitor_plague
I am stil finding equipment with those sorts of failures today...
Not recomending, even having two of them in parallel...
Nope, not for me, sorry

--
*** Suerte a todos y Feliz dia!
1. Re:The caps are electrolytic by Anonymous Coward · 2014-02-27 16:41 · Score: 0
  
  I love how these 47uF capacitors are described as "beefy."
2. Re:The caps are electrolytic by Anonymous Coward · 2014-02-27 17:01 · Score: 2, Funny
  
  Yeah, anything less than 1 Farad is just ridiculous. No way I would hook my Blaupunkt up to that thing...
3. Re:The caps are electrolytic by Anonymous Coward · 2014-02-27 17:12 · Score: 0
  
  Nippon Chemicon capacitors are pretty good actually.
4. Re:The caps are electrolytic by Anonymous Coward · 2014-02-27 17:34 · Score: 1
  
  His point was that the general capacitor quality these days is a bit dubious. Of course there are premium brands in all electronic components.
5. Re:The caps are electrolytic by arielCo · 2014-02-27 17:44 · Score: 4, Informative
  
  tl;dr: these are storage caps, which don't endure the ripple currents that kill filter caps.
  Electrolyte decomposition is usually caused by high ripple current, which is why caps pop mostly (only?) when used as filters, as in motherboard DC-DC converters and gadgets powered by wall-wart adapters. In this particular application, the PSU impedance is quite low and the caps are handled by on-board regulators (V=Q/C and all that), so there's no load ripple and the caps just have to sit pretty and charged with insignificant heat losses until the computer is shut down or outage occurs. Maybe that's why Intel didn't even bother to use the solid (polymer) kind.
  If these caps dry out due to age or bad quality they just won't hold as much charge for emergency sync'ing, which is still better than ordinary SSDs/HDDs with no caps.
  
  --
  This post contains no rudeness or derision of any kind. All arguments are friendly. Terms and exclusions may apply.
6. Re:The caps are electrolytic by arielCo · 2014-02-27 18:03 · Score: 4, Informative
  
  The caps only need to supply enough juice to sync the RAM buffers to flash to ensure consistency of its internal block-mapping metadata (the filesystem should handle the rest through journaling and whatnot). The caps are rated at 35v but let's assume that they're kept at 12v: E = (12 v)^2 * 47 uF / 2 = 3.4 mJoules. Even at full operating load that should last for half a millisecond counting losses, but when power goes out the drive is going to stop serving requests and all it has to do is write that 1 GB buffer to a few flash blocks. More than enough, methinks.
  
  --
  This post contains no rudeness or derision of any kind. All arguments are friendly. Terms and exclusions may apply.
7. Re:The caps are electrolytic by tlhIngan · 2014-02-27 18:07 · Score: 4, Informative
  
  What happens if your superduper SSD develops bad cap syndrome?
  https://en.wikipedia.org/wiki/...
  I am stil finding equipment with those sorts of failures today..
  Except those caps are Nippon Chemi-con. High end high quality capacitors made in Japan. And not the kind involved in the bad caps.
  Bad cap syndrome happens to the cheap caps - stuff like CapXon (aka CrapXon) and such.
  In fact, a lot of bad caps you're finding are the cheap crap ones by the crap manufacturers. You can easily buy them and they will fail.
  That's why you'll find people inspecting caps - and seeing if it's Nippon Chemi-con, Rubycon, Panasonic/Matsushita or other Japanese brand. (You can almost generalize it to those whose brands contain "con" in their name are higher quality - from when they used to be called condensers. The cheap brands all tend to have "cap" in their name).
  So no, I don't see the caps being the weak point because Intel went and spec'd top-quality caps.
8. Re:The caps are electrolytic by Phics · 2014-02-27 20:11 · Score: 1
  
  His point was that the general capacitor quality these days is a bit dubious. Of course there are premium brands in all electronic components.
  Well, I don't know.... his point sounded more like, "Oh, remember those awful Firestone tires from 2000? I'm never using tires again - they all just blow out. Tank treads all the way for me."
  Nothing wrong with solid caps, but the premise for the argument is a little weak.
  
  --
  There are two types of people in the world; those who believe there are two types of people, and those who don't.
9. Re:The caps are electrolytic by ericloewe · 2014-02-27 22:13 · Score: 3, Insightful
  
  A quality electrolytic capacitor will last a long time.
  The ones used here look like Nippon Chemi-Con, rated at 105 C. They'll most likely last forever.
10. Re:The caps are electrolytic by ericloewe · 2014-02-27 22:15 · Score: 2
  
  To be fair, an HDD can use its platters as a flywheel to quickly flush its (relatively tiny) buffer. I never did see proof that that was ever done, though.
11. Re:The caps are electrolytic by SuricouRaven · 2014-02-27 22:59 · Score: 1
  
  Plus they are one of the few components left that anyone with a soldering iron can diagnose and replace, even in this age of surface-mount stacked-chip fiddleyness.
12. Re: The caps are electrolytic by Anonymous Coward · 2014-02-28 00:14 · Score: 0
  
  Nothing is wrong with electrolytic capacitor technology. It's the cheap Taiwan and Chinese caps that are the problem, the chemistry they use is garbage... The reputable manufactures should last decades with no issues at all ( from my 30+ years of experience ). The problem is decent ones are expensive, like $4.00 vs $0.10 each for the fake ones... It's hard to tell as the fake ones have brand names on them... So if you buy somthing that has 10 good size caps for $20 their fake, if its $220 your probably ok... So don't ask for cheap -and- reliable, that's a contradiction! Hope that helps?
13. Re:The caps are electrolytic by Anonymous Coward · 2014-02-28 01:01 · Score: 0
  
  Only with a perfect controller. Just hooking up a capacitor to your powersupply isn't going to keep your device running that long after powerloss.
  Capacitors are not batteries. Their voltage drops fast when drained, causing the device to shut down. You need more capacitance to compensate or a controller that pulls enough of the remaining energy out of the capcitor to keep the device running as long as possible. More capacitance is cheaper so that's usually the chosen method.
14. Re: The caps are electrolytic by Anonymous Coward · 2014-02-28 03:19 · Score: 0
  
  no (normal) cap cost 4 bucks, not even the good ones break a dollar for single quantities
15. Re: The caps are electrolytic by arielCo · 2014-02-28 03:40 · Score: 1
  
  * This is effectively regenerative braking, which I'm not sure you can do with a stepper motor.
  * The arm servo needs extra energy; not sure the platter+rotor have enough.
  * What if it's stopped, heads unloaded, when the power fails?
  
  --
  This post contains no rudeness or derision of any kind. All arguments are friendly. Terms and exclusions may apply.
16. Re:The caps are electrolytic by arielCo · 2014-02-28 04:11 · Score: 1
  
  The regulators are there - you can see the 'big' coils in TFA. The capacitance for a simple "drain till you drop" scheme would have to be a lot here - very roughly 2*Energy/efficiency/(Vddmax^2 - Vddmin^2). So, step the voltage up optionally, keep the caps charged as high as practical, squeeze them dry when needed through a step-down converter.
  TFA also says that the drive periodically monitors the "status" of the caps; I'm not sure if that means charge level or charge-holding capacity but it could test-discharge one cap at a time.
  
  --
  This post contains no rudeness or derision of any kind. All arguments are friendly. Terms and exclusions may apply.
17. Re: The caps are electrolytic by ericloewe · 2014-02-28 04:22 · Score: 1
  
  If the heads are unloaded, there shouldn't be any operations going on, so no harm done (unless some genius decided that it wasn't worth it to immediately get the heads in place the moment data comes in). Note that few hard drives actually spin down, but again, if it's not spinning, there's no data flowing (unless you're really unlucky and it just started).
18. Re:The caps are electrolytic by tlhIngan · 2014-02-28 05:22 · Score: 3, Informative
  
  To be fair, an HDD can use its platters as a flywheel to quickly flush its (relatively tiny) buffer. I never did see proof that that was ever done, though.
  None used it to flush the cache because it is too risky - the platters are not maintaining a fixed speed (they're slowing down to generate electricity) so writes to platters become tricky as the timing is off which means you can overwrite more than you expect.
  Far better to just dump the buffers.
  In fact, the electricity generated by the spinning platters slowing down is used to park the heads - it's called an emergency head park because it basically dumps the electricity into the voice coil that flings the heads to the mechanical stops in the park area. It's fairly violent and most hard drives have much less emergency head park life than standard power down (where the drive moves the heads to the parking area in a controlled fashion) life - a drive may have 50,000+ head load/unload cycles, but under 10,000 emergency park cycles.
  You can tell because a soft-park makes only the smallest of clicking sounds on a drive when it spins down. But emergency park it and it's a much louder clunk.
19. Re:The caps are electrolytic by Anonymous Coward · 2014-02-28 06:03 · Score: 0
  
  Mostly right.
  Any data in write cache is discarded (which is fine, if the OS wanted it written that badly it would have issued a flush/fua/... and that won't get acknowledged as done until the data is actually on platter).
  A currently active sector write *is* finished using mechanical energy of the platter (and on newer drives also with the help of electrical energy from capacitors ... that was one of the challenges in moving to 4k sectors).
  Think about it, if this wasn't done a mid-write power loss would cause a unreadable sector, as you'd have a mess of part new and old data and ecc on platter.
20. Re: The caps are electrolytic by karnal · 2014-02-28 06:21 · Score: 1
  
  On a side note, it has been YEARS since I've witnessed a laptop that actually had the chance to spin down it's drives. Probably since win3.x days. Software being what it is today (McAfee at my place of employment) seems to have this bug of reading/writing to disk every few seconds, defeating any power management setups.
  
  --
  Karnal
21. Re: The caps are electrolytic by ericloewe · 2014-02-28 09:34 · Score: 1
  
  On a laptop, it may make sense to enable Windows' write cache, since there's a battery backup available. That may help you save some power.
22. Re: The caps are electrolytic by Zan+Lynx · 2014-02-28 10:39 · Score: 1
  
  Eh. Except that a blue screen or if you have to do a forced power-off will lose data and require a chkdsk run. I ran my desktop that way for a little while (with a UPS) but it had problems.
23. Re:The caps are electrolytic by Bengie · 2014-03-01 06:23 · Score: 1
  
  Mostly right. Any data in write cache is discarded (which is fine, if the OS wanted it written that badly it would have issued a flush/fua/... and that won't get acknowledged as done until the data is actually on platter). A currently active sector write *is* finished using mechanical energy of the platter (and on newer drives also with the help of electrical energy from capacitors ... that was one of the challenges in moving to 4k sectors). Think about it, if this wasn't done a mid-write power loss would cause a unreadable sector, as you'd have a mess of part new and old data and ecc on platter.
  Many non-sas drives will falsely return that the flush is complete. Unless you look at the firmware on your drive, or do some statistical analysis of your write latencies, you can't really know.
I have news for you, Slashdot. by Anonymous Coward · 2014-02-27 16:21 · Score: 0

At least typically, it ain't overclocked if it came that way from the store.
To little, too late. by Guspaz · 2014-02-27 16:51 · Score: 3, Interesting

Don't get me wrong, I own five discrete SSDs (all currently in active use), and they're all Intel (one G1, two G2s, and two 330s). However, I've been disappointed with Intel of late. It used to be that they came with a premium price, but also dramatically lower failure rates than the competition, and you could usually find them cheaper than the competition if you waited for the right sale.
These days, however, Samsung's failure rates are lower than Intel's, and their price premium is so large that no sale is going to get their larger SSDs anywhere near as cheap as Samsung's. I was hoping that they might make a comeback with a new consumer model, but the 730 is a disappointment in terms of its extremely poor performance-per-dollar and capacity-per-dollar.
I've bought nothing but Intel in the past, because they were the safe bet, but at this point it looks like my next SSD will be from Samsung.
1. Re:To little, too late. by Zontar+The+Mindless · 2014-02-27 19:09 · Score: 1
  
  Just bought a new system that came with the first SSD I've ever owned.
  Been wondering whether or not Samsung was a good choice by the folks who built it for me.
  I'm guessing this means that it was. Thanks for that. :)
  
  --
  Il n'y a pas de Planet B.
2. Re:To little, too late. by ericloewe · 2014-02-27 22:17 · Score: 1
  
  I'm satisfied with my two Samsung 830s, but given the tendency of most Samsung products I own to let me down, I'm not too willing to buy anything else from Samsung.
3. Re:To little, too late. by Spoke · 2014-02-28 08:22 · Score: 1
  
  But the main reason for the Intel 730 is to get power loss protection so your data doesn't get scrambled if your computer suddenly loses power.
  The popular Samsung 840 series don't include that.
  Where are you getting your failure rates from?
4. Re:To little, too late. by WuphonsReach · 2014-02-28 08:40 · Score: 1
  
  the first SSD I've ever owned.
  
  You'll never go back to a 7200 RPM drive for your operating system / primary storage again. The performance when multi-tasking is just too good.
  
  As for the warm and fuzzy feeling side. Get a good backup program like Acronis True Image Home or learn to use rdiff-backup or whatever and write your backups to two different physical drives.
  
  --
  Wolde you bothe eate your cake, and have your cake?
5. Re:To little, too late. by Guspaz · 2014-02-28 09:19 · Score: 1
  
  French retail return rates:
  http://www.hardware.fr/article...
  That's a bit out of date (May 2013), but it includes the previous figures (the "contre" is from November 2012). I doubt they've changed too dramatically since then. I think it's fair to say that Intel and Samsung have similar rates, if nothing else, so Intel's huge price premium is hard to justify.
  Power failure protection is nice, but I don't have any computers that don't have some form of battery backup, be it a UPS or a built-in battery, and some of the SSDs aren't even in use cases where that would matter (who cares if a ZFS cache drive gets scrambled on power failure, it gets reinitialized every boot). So considering that I'd need to suffer a UPS or PSU failure to have the SSD suffer from a power cut, and that the chances of an SSD scrambling its own management data on a power failure is fairly small (user data isn't relevant because the same thing happens to HDDs, and filesystems expect this scenario), it's not really a scenario that I worry about.
6. Re:To little, too late. by Spoke · 2014-02-28 11:40 · Score: 1
  
  Even with built-in battery or UPS, while that reduces the risk of unexpected power loss, in my experience it still happens.
  As far as comparing reliability of SSDs to HDDs, an actual study found that SSDs were much more likely to lose lots of data, sometimes bricking the entire drive.
  http://www.cse.ohio-state.edu/...
  Enterprise HDDs were the most reliable, even the best SSD they tested was not as good (though similar to consumer grade HDD).
  Unfortunately, the study does not reveal which drives were tested.
7. Re:To little, too late. by Guspaz · 2014-02-28 13:23 · Score: 1
  
  The study you've linked to does not support your claims.
  Only two out of the 15 SSD in their test suffered from serious issues. One unit suffered from an SSD metadata issue that effectively prevented access to about 30% of the data on it, and another failed entirely, after having been subjected to 136 power failures in rapid succession.
  Only two HDDs were included in the test, and were subject to a much smaller of power failures. HDD #1 was subjected to only a single power failure (after which it corrupted a write), and HDD #2 was subject to only 24 power failures. How anyone could claim that this is evidence that HDDs are more reliable than SSDs is beyond me. Even the authors of the study don't claim that their results show HDDs to be more reliable than SSDs, they only claim that enterprise HDDs are more reliable than budget HDDs.
  Furthermore, none of the drives were recent, with most being from 2011, and one being as old as from 2009. It's difficult to draw any conclusions on the current SSDs on the market based on testing results on drives from three to five years ago. There has been enormous technological progress in the SSD market over that timespan.
  It's also worth pointing out that newer filesystems are more resilient to many of the faults noted in that paper, especially when redundancy is included. Linux has ZFS, and Windows has ReFS. Both support copy-on-write and block checksums, which go a long way to surviving the sort of issues that both SSDs and HDDs suffer from on power failure.
  Using a system with a UPS or built-in battery does not completely protect you from unexpected power loss, but I don't think that it's happened to me more than once in the past few years. It certainly has never happened to me 136 times in a row at 8-second intervals.
8. Re:To little, too late. by Zontar+The+Mindless · 2014-03-01 06:47 · Score: 1
  
  I keep /home on a separate physical drive (HDD for now). Makes an upgrade or moving to a new machine much more pleasant. I use an SVN repo for my backups
  The boot time is incredible. The new box also builds 8 different versions of $popular_foss_db from source, one after another, in about 25 minutes. Takes about 2 hours on my laptop, using exactly the same OS, build script, compiler options, gcc version, etc.
  (Now if I can just figure out why nouveau does not like my graphics card, and why X doesn't like the Nvidia driver, I'll be a really happy guy.)
  
  --
  Il n'y a pas de Planet B.
9. Re:To little, too late. by Wolfrider · 2014-03-02 18:34 · Score: 1
  
  > Now if I can just figure out why nouveau does not like my graphics card, and why X doesn't like the Nvidia driver
  --Try searching/posting on the support forum for your distro; you may get results.
  
  --
  .
  == WolfriderV6 == I'm willing to admit that *I just might* be wrong... Are you??
Summary missed an important point by timeOday · 2014-02-27 17:21 · Score: 5, Insightful

Not only is this "rated to endure at least 70GB of writes per day over five years," it also comes with a 5 year warranty. Given there's still skepticism about SSD reliability from some quarters, a 5 year warranty is unbeatable.
I only wish Intel was offering this in a smaller size, say 100 GB. I think a SSD system drive + slow "green" HDD is a great combo in a desktop, and the price premium on this quality of SSD would be easier to swallow if the drive were $110 instead of $250 even though that would be the same $/GB.
1. Re:Summary missed an important point by Anonymous Coward · 2014-02-27 17:48 · Score: 2
  
  I have no idea where this fascination of making windows boot 5 seconds faster and load up paint lighting fast comes into play, its often weeks if not months between times I reboot, and its all my space hungry big ass applications that are slow, not calculator
2. Re:Summary missed an important point by timeOday · 2014-02-27 18:53 · Score: 2
  
  For me 80 GB is sufficient to store all the applications and data for my family except a few select "big" things that go on the HDD - my DVR recordings, my wife's flatbed scans of her illustrations, my son's screen recordings of minecraft... when the disk starts to get full, I find the offending directory, move it to the HDD and make a symlink. This is Linux; I find Windows to be a terrible drive space waster, and it just grows forever as you apply patches and service packs.
  Laptops are obviously more difficult, since no HDD. I found a 250 GB SSD to be somewhat tight on my OSX laptop for work - much happier now with 500 GB.
3. Re:Summary missed an important point by Anonymous Coward · 2014-02-27 19:31 · Score: 0
  
  I have no idea where this fascination of making windows boot 5 seconds faster and load up paint lighting fast comes into play, its often weeks if not months between times I reboot, and its all my space hungry big ass applications that are slow, not calculator
  Aren't you aware that this drive come pre-loaded with Gentoo?
4. Re:Summary missed an important point by Blaskowicz · 2014-02-27 21:30 · Score: 0
  
  It's needed in Windows because updates (most too often) need the computer to reboot. So there's a dance of install update -> initiate reboot -> update installation finalizes at either shutdown or boot up -> finish boot and log in -> reopen your apps and get back to work.
  You might also do a very long virus scan or spyware scan (esp. the first time you install either utility).
  Thus Windows users "need" an SSD an then, not waiting for an Explorer window etc. is a bonus. Well in linux using some heavyweight stuff (gnome 3/cinnamon etc.) may make you wait for a file manager to open too so using an SSD depends on your software bloat and expectations (and other reasons like running some I/O intensive stuff)
5. Re:Summary missed an important point by ericloewe · 2014-02-27 22:19 · Score: 1
  
  I'm not familiar with symlinks besides NTFS', but if NTFS is any indication, I'd avoid them wherever possible.
6. Re:Summary missed an important point by SuricouRaven · 2014-02-27 23:01 · Score: 1
  
  NTFS isn't. Symlinks work fine and are very useful on an OS made to handle them. On Windows they are bodged in awkwardly and clumsily. The only reason NTFS supports them at all is so it can claim POSIX compatibility.
7. Re:Summary missed an important point by Anonymous Coward · 2014-02-27 23:08 · Score: 0
  
  I install all my applications onto the SSD along with the operating system. Every program opens almost instantly. If I have a resource heavy game which slows down due to texture loads, I will move it to the SSD and the problem disappears.
  When I'm editing / stitching photos I use the SSD as the cache / scratch space, keeping the photos themselves on a regular platter disk. Editing a thousand+ RAW images of 25mb each used to be annoying because eventually the hard drive would just churn and thrash due to it not being able to meet the demand. Not with the SSD.
  Everything blazes on it, not just boot up. But now that you mention it, I like switching my computer on and having it ready to use in 6 - 8 seconds from a cold start. Oh, it's also completely silent.
  I could never go back to a non-SSD system disk. Never. Of all the desktop improvements which have been realized in the last 5 years, the SSD is by leaps and bounds the single best performance investment one can make. There is nothing close.
8. Re:Summary missed an important point by Blaskowicz · 2014-02-27 23:14 · Score: 1
  
  Symlinks were only introduced in Vista, so I think you're quite wrong.
  The crappy NTFS feature that predates it is called a junction point.
9. Re:Summary missed an important point by AmiMoJo · 2014-02-28 00:30 · Score: 1
  
  Does the warranty still apply if you write more than the rated 70GB/day?
  
  --
  const int one = 65536; (Silvermoon, Texture.cs)
  SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
10. Re:Summary missed an important point by Anonymous Coward · 2014-02-28 00:30 · Score: 0
  
  I bought a Intel 335 180GB SSD and it came with a 5 year warranty with no extra price attached.
11. Re:Summary missed an important point by AmiMoJo · 2014-02-28 00:31 · Score: 1
  
  I guess you are not a gamer. Games are ridiculously large these days.
  
  --
  const int one = 65536; (Silvermoon, Texture.cs)
  SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
12. Re:Summary missed an important point by SuricouRaven · 2014-02-28 00:51 · Score: 1
  
  Vista introduced the utility to make symbolic links. The filesystem supported them before that, there was just no support for actually using them. You can get them working in XP, it just takes some hackery.
13. Re:Summary missed an important point by timeOday · 2014-02-28 03:57 · Score: 1
  
  Even though I don't reboot my main linux system often, I find it benefits from a fast system drive because I use the computer for recording TV shows, which basically destroys the filesystem cache. (I.e. it dumps all the little files I'll actually be accessing again to store blocks of huge video files I really don't want cached). At least I think that's what happens.
14. Re:Summary missed an important point by Tharkkun · 2014-02-28 07:33 · Score: 1
  
  I have no idea where this fascination of making windows boot 5 seconds faster and load up paint lighting fast comes into play, its often weeks if not months between times I reboot, and its all my space hungry big ass applications that are slow, not calculator
  When you're doing presentations for customers you want your laptop to bootup fast. You want your reads in your applications to zoom because you're attempting to replicate a server class application on a desktop/laptop pc. There's all kinds of other examples but if you reboot every 3 weeks you could probably use a 4 year old machine. You're not utilizing it's potential.
dedupe firmware goes mainstream by Anonymous Coward · 2014-02-27 17:39 · Score: 0

Its a great drive that is very inexpensive to produce.
random vs sequential by roc97007 · 2014-02-27 18:38 · Score: 4, Insightful

> Although the drive boasts exceptional throughput with random I/O, its sequential transfer rates are nothing special."
But good random access will give you better overall performance in most cases. You rarely need to deathmarch through the drive.

--
Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
1. Re:random vs sequential by Anonymous Coward · 2014-02-28 07:15 · Score: 0
  
  See the thing is that with SSD random vs sequential is a non-issue.
  With HDDs you get good sequential performance because the data is written to the disk in a way that maximizes this. Randoms are harder as you may have just missed your window to access the data you want and will need to wait on the disk to spin around again.
  SSDs don't have a disk, so accessing the 'next' block is just as hard as accessing a random block.
2. Re:random vs sequential by roc97007 · 2014-02-28 07:32 · Score: 1
  
  I agree, which makes me a little puzzled as to why random vs sequential is even a measurement for silicon drives. It's a completely different storage paradigm.
  
  --
  Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
No kidding by Sycraft-fu · 2014-02-27 18:44 · Score: 5, Insightful

What you discover with SSDs is that for desktop usage pretty much any drive is "fast enough" and that faster doesn't much matter. I went from a SATA-2 SSD that was fairly slow even for that generation (WD Siliconedge) to a SATA-3 SSD that is fairly fast for this generation (Samsung 840 Pro) and I don't notice any difference. I can benchmark a difference, but I don't see any difference in load times and so on. SSDs are fast enough that they are making themselves not the bottleneck.
That's also why there isn't a ton of interest in the PCIe SSDs. You can get way more performance, but it is a somewhat limited set of scenarios (on the desktop at least) where that would matter.
1. Re:No kidding by Luckyo · 2014-02-27 19:46 · Score: 4, Informative
  
  The much bigger reason for lack of interest in PCI-E SSDs is inability of that interface to pass on TRIM commands in the current implementations. In home use that is of far greater importance to speed over drive's life time than theoretical read and write times.
2. Re:No kidding by nateman1352 · 2014-02-27 21:54 · Score: 5, Informative
  
  Citation Please.
  In truth current gen PCIe SSDs appear to the OS as a PCIe bus connected AHCI controller with a single disk that supports TRIM. There makes it completely transparent... it works exactly the same as a SATA SSD from a software perspective.
  Pretty soon we will start seeing next gen PCIe SSDs that expose themselves as an NVMe controller instead of an AHCI controller. Those SSDs will be backwards incompatible with AHCI but the command protocol and DMA interface enables extreme parallism so we will see pretty incredible performance for those SSDs. From a software stack perspective they use a new NVMe host controller and a new command set (ATA commands are completely gone!) So you need new drivers for it. They have OSS Win7/8/8.1 drivers available for NVMe but due to kernel limitations only the Win8/8.1 version of the driver is capable of supporting TRIM (Maybe that is where you got confused.) Win8.1 also have a NVMe driver in-box from Microsoft.
  Don't worry though, AHCI PCIe/SATA Express SSDs will be with use for a very long time esp. since Win7 is rapidly turning in to the next WinXP (the version that everyone likes and uses despite Microsoft's best efforts.)
3. Re:No kidding by ericloewe · 2014-02-27 22:05 · Score: 1
  
  That's crazy. All you'd need is a controller that pretends to be SCSI. I'm fairly certain SCSI supports TRIM. Same goes for AHCI (SATA).
  Besides, many "PCI-e" cards are SATA controllers attached to a PCI-e - SATA bridge (often in a RAID-like format, sometimes all exposed to the OS).
4. Re:No kidding by dinfinity · 2014-02-27 22:44 · Score: 1
  
  Yeah, right. Because home users regularly say: "Oh, wow, that is a sweet PCI-E SSD; too bad it doesn't support TRIM or I would have bought the shit out of it!"
  Yes, that is much more probable than "home users can't tell the difference between different SSD experiences."
5. Re:No kidding by AmiMoJo · 2014-02-28 00:19 · Score: 1
  
  The main improvement is not the extra bandwidth provided by SATA3 but the improved caching using on-board DRAM and improved handling of background processes to shuffle data around. These only affect write speeds, read speeds are mostly unchanged.
  At the moment PCI-E SSDs are fairly pointless because the performance bottleneck is the write speed of the SSD. In benchmarks on an empty, virgin drive write speeds of 550MB/sec are not uncommon, but once the drive starts to get full up and blocks need shuffling or partially re-writing the performance drops to less than half that on most drives.
  
  --
  const int one = 65536; (Silvermoon, Texture.cs)
  SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
6. Re:No kidding by Anonymous Coward · 2014-02-28 01:14 · Score: 0
  
  That would be SAS, no?
7. Re:No kidding by jedidiah · 2014-02-28 03:29 · Score: 1
  
  > Yeah, right. Because home users regularly say: "Oh, wow, that is a sweet PCI-E SSD; too bad it doesn't support TRIM or I would have bought the shit out of it!"
  Your average home user probably still isn't even aware that they can gain any advantage out of an SSD. This is despite constant ongoing persistent propaganda about how just adding/replacing an SSD will change everything and even improve the weather.
  "home users" simply aren't relevant here.
  
  --
  A Pirate and a Puritan look the same on a balance sheet.
8. Re:No kidding by Luckyo · 2014-02-28 06:28 · Score: 1
  
  I've been looking at current gen PCI-E based SSDs and I haven't found a single one that would support TRIM yet. Maybe some server grade ultra expensive stuff would. But current generation of consumer products does not.
  This could be because most of these go for maximum speed, and shove a RAID0 controller on the SSD to improve speeds. This would be another factor that strips TRIM commands.
9. Re:No kidding by Anonymous Coward · 2014-02-28 06:39 · Score: 1
  
  Hi, I'm the average home user and I don't use an SSD because:
  1.) too expensive
  2.) no need
10. Re:No kidding by Bengie · 2014-02-28 17:42 · Score: 2
  
  The main improvement is not the extra bandwidth provided by SATA3 but the improved caching using on-board DRAM and improved handling of background processes to shuffle data around. These only affect write speeds, read speeds are mostly unchanged.
  At the moment PCI-E SSDs are fairly pointless because the performance bottleneck is the write speed of the SSD. In benchmarks on an empty, virgin drive write speeds of 550MB/sec are not uncommon, but once the drive starts to get full up and blocks need shuffling or partially re-writing the performance drops to less than half that on most drives.
  Most benchmarks that I've seen of Samsung, Intel, and other top end drives, which also tend to be some of the cheapest, is they have almost no discernible difference in performance in synthetic benchmarks, even up to 95%+ capacity. They tend to have enough spare space set aside to handle shuffling around data.
Me either by Sycraft-fu · 2014-02-27 18:47 · Score: 1

I think maybe because it is something that can be easily shown off, or because it can be done cheaper, or because they have a misguided belief that it makes everything fast.
Personally if I can't afford an SSD big enough to stick all the apps I normally want on there, I don't bother with an SSD in a system.
New standard: SATA express by IYagami · 2014-02-27 20:45 · Score: 2, Informative

There is a new standard which will increase SATA speed ( http://en.wikipedia.org/wiki/S... )
Currently, Apple computers use PCIe SSD disks, which increases their performance:
http://www.anandtech.com/show/...
"I'm very pleased with Apple's PCIe SSD, at least based on Samsung's new PCIe controller. Sequential performance is up considerably over last year's 6Gbps SATA drive. Go back any further and the difference will be like night and day, especially if you were one of the unfortunate few with an older Toshiba drive. Internal transfers are quicker, but to actually use the new SSD to its potential you'll really need a very fast external Thunderbolt array - even USB 3.0 can't completely tax it. There's still a lot more investigating that I want to do on Samsung's new controller, but my early results look very promising. It's sort of crazy that Apple now ships a mainstream consumer notebook with a PCIe SSD capable of almost 800MB/s. Now that Apple is off SATA, scaling storage performance should be much easier to do going forward. "
1. Re:New standard: SATA express by gnasher719 · 2014-02-27 22:55 · Score: 1
  
  In reality, the big advantage of SSD is zero access time for reading gazillions of 4 KB blocks. And there the transfer speed doesn't really make much difference. Even just 100 MB/second is 25,000 random four KB blocks.
2. Re:New standard: SATA express by SuricouRaven · 2014-02-27 22:56 · Score: 1
  
  If it's a PCI-e port on a cable, does that mean you can plug non-storage devices in too? I can see applications for things like video walls or GPGPU number-crunchers, where very 'sata' port is potentially a way to cram another video card in.
3. Re:New standard: SATA express by Anonymous Coward · 2014-02-27 23:18 · Score: 0
  
  As long as whatever you plug in doesn't need the PCIe REFCLK ... yeah, should work.
4. Re:New standard: SATA express by jedidiah · 2014-02-28 03:36 · Score: 1
  
  Yes. Quite.
  SSD tech is still quite expensive. Drives are still tiny. This doesn't leave a lot room for massive bulk storage. So the use cases for improved sequential IO access are limited.
  If you run an artificial benchmark you've got to be careful. Blink and you will miss the results.
  
  --
  A Pirate and a Puritan look the same on a balance sheet.
5. Re:New standard: SATA express by tlhIngan · 2014-02-28 04:48 · Score: 2
  
  If it's a PCI-e port on a cable, does that mean you can plug non-storage devices in too? I can see applications for things like video walls or GPGPU number-crunchers, where very 'sata' port is potentially a way to cram another video card in.
  And yet, that exists today, it's called Thunderbolt. Which is effectively a PCIe x2 over a cable. Thunderbolt drive arrays exist for performance gains that go beyond what SATA has and all that.
6. Re:New standard: SATA express by SuricouRaven · 2014-02-28 05:17 · Score: 1
  
  There are issues with thunderbolt being ridiculously expensive.
7. Re:New standard: SATA express by Anonymous Coward · 2014-02-28 07:40 · Score: 0
  
  Magma external PCIe is even more crazy expensive than Thunderbolt AFAIK.
still rocking some x25-e's by kcmastrpc · 2014-02-27 23:35 · Score: 1

At 90% health too. intel ssd drives are worth the premium
1. Re:still rocking some x25-e's by Anonymous Coward · 2014-02-28 04:33 · Score: 0
  
  I still own a couple of the much-villified OCZ Vertex2s, which are still working fine on main desktop use for the past 3 years (now as a part of a FusionDrive pair). I honestly feel (based on my experience with 5 total aftermarkets) that if SSDs don't fail within the first few months, they usually last for the long haul.
get help by raymorris · 2014-02-28 00:15 · Score: 1

Seek help for that.
Anything but OCZ by Anonymous Coward · 2014-02-28 02:21 · Score: 0

It'll be a nice, durable, SSD. I'll stick with anything but OCZ and buy what;s on sale.
Not really by mindcandy · 2014-02-28 02:26 · Score: 1

Intel (and everybody else) does this for good reason .. high endurance components (Milspec, server, whatever) are usually designed for tolerances far beyond the actual spec, because manufacturing issues can cause the tolerances of the finished product to deviate somewhat.

If they design a [gizmo] to operate at 1.5ghz and sell it as a 1ghz chip knowing full well there is plenty of overhead but chances of failure running it at 65% of design are pretty much nil, yay for them for meeting the rejection rate.

Then along comes marketing and says "hey, we can sell the rest of them at 1.5ghz as consumer units" .. and the failure rate there doesn't really matter as much because you just print a disclaimer about "your data may go poof" and RMA the broken ones. As long as the defect rate is low enough to remain profitable, yay again.
I find this ... by Anonymous Coward · 2014-02-28 04:38 · Score: 0

lol ... again limited by the bus speed. if they'd have tested it on xeon 2011 with more then 5 GT/sec bus bandwidth
then one could call it a test ... 550MB/sec read? really?