TSSDR TEST BENCH AND PROTOCOL
SSD Testing at TSSDR differs slightly depending on whether we are looking at consumer or enterprise SSDs. For consumer SSDs, our goal is to test in a system that has been optimized with our SSD Optimization Guide, although CPU C States have may or may not have been optimized depending on the motherboard base configuration. Benchmarks for consumer testing are also benchmarks with a fresh drive so, not only can we verify that manufacturer specifications are in line but also, so the consumer can replicate our tests to confirm that they have an SSD that is top-notch. We even provide links to most of the benchmarks used in the report.
This Test Bench build was the result of some great relationships and purchase; our appreciation goes to the below mentioned manufacturers for their support in our project. Our choice of components is very narrow, in that, we choose only what we believe to be among the best available and links are provided to each that will assist in hardware pricing and availability, should the reader be interested in purchase.
|PC CHASSIS:||InWin D-Frame Open Air Chassis|
|MOTHERBOARD:||ASRock Z87 Extreme11/ac EATX MotherBoard|
|CPU:||Intel Core i7-4770K Haswell 3.5GHz Quad Core|
|CPU COOLER:||Corsair H100 High Performance Liquid|
|POWER SUPPLY:||be quiet Dark Power Pro 10 1000W PSU|
|SYSTEM COOLING:||be quiet Silent Wings 2 PC Fans|
|GRAPHICS CARD:||EVGA GTX 770 Superclocked with ACX Cooler|
|MEMORY:||Crucial Ballistix Sport DDR3-1600Mhz Memory|
|KEYBOARD:||Corsair Vengeance K95 Mechanical Gaming Keyboard|
|MOUSE:||Corsair Vengeance M95 MMO/RTS Laser Mouse|
|ROUTER:||NetGear R6300 AC1750 Dual Band Gigabit WiFi Router|
|HBA||HighPoint RocketU 1144C 4 x USB 3.0 20Gb/s HBA|
The software we will be using for today’s analysis is typical of many of our reviews and consists of ATTO Disk Benchmark, Crystal Disk Info, Crystal DiskMark, AS SSD, and PCMark Vantage. In consumer reports, we prefer to test with easily accessible software that the consumer can obtain, and in many cases, we even provide links. Our selection of software allows each to build on the last and to provide validation to results already obtained.
Crystal Disk Info provides some excellent information about the SSD itself to include its health, product information, ‘power on’ information as well as the characteristics of the SSD. We can see that the SSD is capable of TRIM as it is not grayed out as with AAM.
ATTO Disk Benchmark is perhaps one of the oldest benchmarks going and is definitely the main staple for manufacturer performance specifications. ATTO uses RAW or compressible data and, for our benchmarks, we use a set length of 256mb and test both the read and write performance of various transfer sizes ranging from 0.5 to 8192kb. Manufacturers prefer this method of testing as it deals with raw (compressible) data rather than random (includes incompressible data) which, although more realistic, results in lower performance results.
SINGLE 730 SERIES SSD
2 X 730 SERIES SSDS IN RAID 0
Both ATTO Disk Benchmark tests reflect performance above specifications listed by Intel. By looking at these results, we can tell that this SSD will be a strong performer as both read and write performance steadily improves during file size progression. Also, SSD performance hit SATA 3 speeds right at the 8K file size in both scenarios.
The first SSD which performs better random read in RAID than running alone I’ve ever seen! Good job Intel and thanks so much for the review! 😀
why are there no other SSDs in the comparison pool for any of the tests? other sites show this SSD as much slower than others, yet comparisons are avoided here.
The best thing about multiple site reviews is that we all have our own way of either comparing or not comparing. Comparisons are not avoided at all, but rather, we have provided report consistent to that we have provided for several years. Thanks for checking our report out!
I noticed that this SSD requires 12v and 5v inputs whereas other SSDs only require 3.3v. Is this just because of the controller being used?
I had checked the datasheet last week and its either 12 V or 5V not both, using 12V is marginal gains. Check page 9 for more details from the whitepaper and much more: https://www.intel.com/content/dam/www/public/us/en/documents/product-specifications/ssd-730-series-spec.pdf
Les, I always wonder when I see RAID0 results worse than single-SSD. My suggestion for you : try to compare SW RAID0 with HW RAID0 using proper controller, like the new LSI or Adaptec7/8series you liked so much. They have definitely shown some phenomenal muscles.
I guess several readers would be interested in software VS hardware raid0 performance numbers…
Thanks for the advise.
is the corsair neutron gtx not faster than this ssd
The Neutron GTX has a rocket controller that has never got the true visibility it should. The same controller is used in the Seagate. Looking at numbers alone, you are absolutey correct but the 730 offers something just a bit more impressive that might lure in so many walking on the fence, proven reliability which has seen enterprise, as well as endurance.
I’m waiting for the X99 series to come out with 10 native Sata 3 ports and then I’ll probably throw 5 of them in raid 0. My 2×520’s could do for some replacing and the raid Caviar Blacks sitting next to them are starting to feel obsolete.
It’s me again. Decided to install 2 of these bad boys in Raid 0 to replace my 520 array. My numbers are substantially better than SSD Review’s: https://postimg.org/image/jr1f3hvot/
Compared to my 520’s these things are pretty much 2x as fast and chew through incompressible like nothing (here’s the benchmarks I took on the 520’s right before installing the 730’s): https://postimg.org/image/uib8tnjsf/
My setup is a 3770k @ 4.9 Ghz, ASRock Z77 OC Formula, 32GB DDR3 @ 2133 Mhz.
Nice results although I don’t see substantial difference and yours is a bit lower in the RAID CDM I think.
ATTO was pretty big difference in the smaller chunk sizes where your results don’t seem to ever catch up to mine until 32KB where the drive is pretty much at its max (300 vs 50k @ .5 for example).
Also with AS SSD where my overall score is 25% higher than yours, with a pretty big 50% difference in 4K-64 results. I saw results similar to yours (50% decrease) when I didn’t have Write Back something or other Enabled in Intel RST.
CMD I chalked up to being “close enough” with the variation maybe being explained by me not using totally empty partitions / a clean environment.
The only eyebrow raiser was my access times for read were .06msec vs. your .03msec and my write was .01 vs your .03.
Ran some benchmarks with Write Back Cache disabled in Intel RST and my 4K + ATTO results start to look closer to yours (and even my Write Acc. Time matches yours now instead of being 3x faster):
Hello — Intel gamed you, us. The final product in the retail market as opposed to the Engineering Sample does not contain the 2 capacitors and power loss protection circuitry, and this is confirmed by Intel. Can you please revise the review, and possibly compare against the actual retail product that consumers will actually use? I bought it based on the power loss protection, but it doesn’t exist.
Confirmation by Intel: “We have confirmed the SSD 730 series does not
come with the 2 capacitors needed to support the power loss data
Is this for real ?
Damn, another bait and switch. Although according to this, it has them
Someone should research whether or not the review sample provided has features that are different from the retail version.
apperently it is
Thanks for the link. It adds an additional confusion factor, though, rather than clear it up.
Anybody who reviewed one of the “engineering samples” should contact Intel and clear this up.
Frankly, it doesn’t matter. Intel never advertised those features (namely power loss protection), so it doesnt matter if drive has powerloss hardware or not. if reviewers marked this as a feature (even though intel nevel publicised or advertised it) its not intels fault, if drive ends up not having it.
Besides, present hardware doesnt mean present features. Powerloss protection could be easily disabled whithin firmware (for segmentation reasons or something else, doesnt matter) and its nothing you can do, since it was not an indended feature to begin with.
It was the same deal with mx100, which ended not having full power protection.
Of course it matters. The prospective purchaser should be able to identify features before purchase. Reviews should be accurate and published specifications should be accurate. If either one is not accurate, it should be corrected. The basic credibility of the reviewer and the manufacturer is involved.
> Reviews should be accurate and published specifications should be accurate.
Published specs WERE/ARE accurate. Its the reviewers that are at fault here, for assuming there is powerloss protection.
I am not ready to blame the reviewers, or anyone else for that matter, at this point. Many questions still remain.
Hey Guys — I got in touch with another review site who’s looking into this, but the last official word from Intel was this: “My apologies for the misunderstanding.
What I was trying to let
you know is that the link we found for the review, the third party link,
where it shows the drive with the capacitors is fake.
confirmed it is fake as my engineering department double check about it
and the SSD 730 was never built with the capacitor for the power loss
This means, the SSD does not have the capacitors at all, therefore the Intel’s website has the correct information of the drive.
Let me know if you need anything else.”
So, either the person in support at Intel doesn’t know what they are talking about, or all the review sites I sent him, including this one, has been faked. I saw that linked Ben posted before, which confirms that the retail version has caps, but as suggested here, and in the link and elsewhere, it doesn’t mean that the physical hardware that implements power loss protection has been implemented. I do disagree with Ben about segmentation and that it’s not important. It is. I bought this instead of an EVO because of this feature based on all the reviews before I made my purchase of two of these drives in desktops where I don’t have a UPS, so power failure protection is important, as they consume a lot of power, where it’s really not meant to be used in a laptop (but you can of course) but if that was the case, I would’ve bought the EVO 850, which is cheaper, faster and also has a 5 year warranty. Did the Intel rep who gave the sample for review mention it has this, or was this speculation based on what was visible on the PCB?
Very interesting, and all germaine, but it seems that the truth is not known at this point.
It is hard to believe that a review website, let alone multiple review websites, would fake it.
IMHO, I completely agree, and to be clear, sorry if i was ambiguous on that statement, as I didn’t mean to suggest that the review sites were faked, or that all the reviewers colluded together to fake it to get some free samples — that would be absurd, which is what I think of the reply that Intel sent me back. There’s no way all these review sites could come to the same conclusions about it have power loss protection, without an Intel rep suggesting that it does. If the final product released to retail has this disabled, then I would really suggest that review sites stop accepting engineering samples and only use final retail products to review, as the reviewers would have no idea whats been added or removed in the final product. I’m not blaming the reviewers, I’m just saying that in order for review sites to be beneficial to users who rely on their reviews for decision buying, they should be testing the same items people will buy off-the-shelf, not samples that are subject to change in the final product. Please, don’t interpret this as mean, snarky or anything like that, we all want learn about new products, what features they offer and compare them against others, and so we need review sites, and I really appreciate the service and utility they provide — the only point I’m stressing here since I discovered this is that I think unless any company that provides an “engineer sample” for review should state in writing that the features present on this sample sent to you for review will have the same components (not bait-and-switch cheaper brand caps, if you gave nippon chemi-con caps in the ES, then we should expect nippon chemi-con caps in retail) and features in retail — we are giving you as ES instead of retail due to whatever reasons (no warranty, no resales, etc….) Anyways, I’ll follow-up shortly.
You have been perfectly clear in your previous comments and in this one, and I have been in complete agreement with you throughout.
This isn’t so easy of a thought because of the inherent differences between enterprise and consumer SSDs, the most obvious being that enterprise SSDs meet industry standards approval and have a set BOM (bill of materials). Consumer SSDs dont have this and it is almost common for something to change for value sake, whether it be RAM, cache, or any other part.
I figured that, the feature set remaining the same, while the parts could change, as long as they still meet the specs of the features keeping product pricing relatively stable — with EE, the parts are explicit, so the pricing could vary depending on the costs of parts at production from the supplier. As an update, it looks like the Intel support rep was inaccurate and that the 730 series *may* support Flush-in-Flight power loss protection — I’ll reply back with the details once I get solid confirmation.