A Word on Reliability

Marc Prieur has been around writing about hardware for as long as I can remember (think back to the old school Tom days). Late last year he published some particularly controversial numbers on his website: failure rates of various PC components according to a French etailer. Among the components were SSDs and the numbers are below:

SSD Failure Rates - Hardware.fr
  Intel Corsair Crucial Kingston OCZ
Failure Percentage 0.59% 2.17% 2.25% 2.39% 2.93%

I should add that the numbers Marc published were accurate (confirmed by some of the manufacturers involved), although they don’t paint the complete picture of world wide failure rates - they are an important sample to look at.

Other than Intel, none of the companies listed in that article were particularly pleased with the numbers.

I mentioned earlier that the 510 would go through Intel’s extensive validation testing, just like any other Intel product. Presumably this means that the SSD 510 should have similarly low failure rates in the field (unless there’s something horribly wrong with the Marvell controller that is). Compatibility should also be a strong point of the SSD 510 due to Intel’s stringent internal testing.

Note that I am separating reliability and compatibility from drive longevity. There’s typically a good correlation between high random write performance and low write amplification. The Intel SSD 510 doesn’t have particularly high random write performance, and in turn should suffer from fairly high write amplification in highly random workloads.

I’ve already proved in the past that at 5,000 p/e cycles there’s no cause for worry for a normal desktop user. The likelihood that you’ll wear out all of your NAND within the next 5 years is very, very low. However I will say that when faced with enterprise workloads you’re going to have to pay much closer attention to write amplification and spare area than you would on say a SandForce drive.

The Test

CPU Intel Core i7 965 running at 3.2GHz (Turbo & EIST Disabled)
Intel Core i7 2600K running at 3.4GHz (Turbo & EIST Disabled) - for AT SB 2011, AS SSD & ATTO
Motherboard: Intel DX58SO (Intel X58)
Intel H67 Motherboard
Chipset: Intel X58 + Marvell SATA 6Gbps PCIe Intel H67
Chipset Drivers: Intel + Intel IMSM 8.9
Intel + Intel RST 10.2
Memory: Qimonda DDR3-1333 4 x 1GB (7-7-7-20)
Video Card: eVGA GeForce GTX 285
Video Drivers: NVIDIA ForceWare 190.38 64-bit
Desktop Resolution: 1920 x 1200
OS: Windows 7 x64
Intel’s SSD 510 Powered by Marvell Random Read/Write Speed


View All Comments

  • 7Enigma - Wednesday, March 02, 2011 - link

    Like Anand I'm shocked Intel went this route. After touting the random read/write so heavily (and for good reason) they pull this stunt and look horrible. I'm an owner of an 80gig G2 and love it but have been in the process of justifying an upgrade so I can put the 80gig in my laptop. I'm willing to sacrifice a bit of sequential performance (just like I did with the G2 over other competitors), but this is a complete step backwards.

    And remember the review unit was the larger drive. Put the 120gig version in those benchmarks and the critical numbers that are so low to start with are only going to become much worse. Just like all of the previous releases due to high cost the majority of people will be purchasing the smaller-sized drives (most people bought Intel 80gig when 160gig was out, etc.).

    When you are competing on firmware only, and that firmware is not the best to begin with (in comparison to Sandforce), you have to be perfect, and in this case Intel just didn't produce.

    Very disappointed indeed.
  • Concillian - Wednesday, March 02, 2011 - link

    "After touting the random read/write so heavily (and for good reason) they pull this stunt and look horrible."

    They don't look horrible, they are 2nd or 1st in most of the real world tests, Improving on the G2 performance.

    Random performance is lower, but the real world tests demonstrate that random performance has reached a point where it is "enough" for typical workloads.
  • semo - Wednesday, March 02, 2011 - link

    Whose real world are we talking about here? Average Joe "real world", mainstream "real world", enthusiast "real world"... you get the idea.

    As an average user (net browsing, office apps, flash gaming, etc...) this sort of performance is great. However, the same user would benefit just as much from a much cheaper 60GB SSD. The 510 is neither cheap overall or in $/GB sense. My "real world" is using VMs and lot's of tests involving high QDs and IOPS so this drive is not for me... At that price, capacity and performance, I don't know who this drive is for.
  • TEAMSWITCHER - Wednesday, March 02, 2011 - link

    It's a bit surprising that the Crucial C300 (on the 6 GB/s) interface did so well in the real world benchmarks. This drive is a year old and still competitive. Newegg had them on sale a couple weeks ago for only $480.00. Which is a bargain compared to the $615.00 for the Intel 510. This is an interesting technology to watch, with lots of plot twists and upsets. I can't wait to see the crucial M4. Reply
  • Lonesloane - Wednesday, March 02, 2011 - link

    I see the price point of the new Intel 510 drives extremely critical. Intel have always charged more than the competition and got away with it due to the fact that on the other hand there was the good quality, compatibility and reliablity of their SSD products.

    Now it seems like a clear win for Sandforce, especially if we take the superior write amplification of the sandforce controller, which is a big advantage regarding the reduced write cycles of the new 25nm NAND.
  • Shadowmaster625 - Wednesday, March 02, 2011 - link

    You have to tell them to stop sending you Ferraris because no one buys Ferraris. The 120GB versions of both the OCZ and Intel drives are much more relevant, and we dont have the foggiest clue how they line up. And even if we did, we would not know how the 40/60/80 GB versions stack up against each other. I am currently shopping for a 40-60GB SSD and I cannot even to this day get a clear answer on which product is better (performance per dollar). And all of these drives have been out for at least year.... Reply
  • 7Enigma - Wednesday, March 02, 2011 - link

    Exactly. I posted above about this and it will seriously hurt Intel as the poor random read/write will only become worse. Reply
  • Concillian - Wednesday, March 02, 2011 - link

    "I am currently shopping for a 40-60GB SSD and I cannot even to this day get a clear answer on which product is better (performance per dollar). And all of these drives have been out for at least year.... "

    This is a concern and all you really need to know is how they reduce capacity.

    They can do it by reducing the number of chips (usually will see a performance hit)
    They can do it by using the same number of chips, but each stores less (usually will perform about the same)

    When you look at the OCZ 25nm "debacle" it was this very decision that created the issue. The 25nm flash itself was not the problem (even though that got a fair bit of the blame)... it was how they implemented it. They used half the 25nm higher density chips in the drives, reducing performance. This also meant reserving one block for over-provision meant it was larger in size, so that ate into the usable capacity. The solution still uses 25nm chips, but they're the same number of chips as 34nm, so they perform at about the same level.

    There probably needs to be a chart at the beginning like the video card articles showing details for each drive. It would clear up any issues, as well as something that can be referred to in the future. I refer to the old video card reviews when looking for things like ROP / texture counts, memory bandwidth, speeds, etc... because those charts have really useful information in them.
  • jnmfox - Wednesday, March 02, 2011 - link

    We need a round-up of OS boot drives (40/60/80 GB). I love my 80GB G2 but I’m not interested in paying more than $200, hopefully less, for an SSD.
  • Stargrazer - Wednesday, March 02, 2011 - link

    I too would love to see the numbers for the 128(-ish) GB versions. Reply

Log in

Don't have an account? Sign up now