Article Thumbnail

SATA 3Gb/s vs. 6Gb/s Cable Performance (Revisited)

Written on February 11, 2013 by Matt Bach


When SATA 6Gb/s drives were first launched, there was a lot of confusion over whether or not you needed a special cable in order to get the full speed benefit of SATA 6Gb/s. While the official SATA-IO documentation states that there is no difference between SATA 3Gb/s and SATA 6Gb/s cables, there were still some people that insisted that you needed a SATA 6Gb/s cable to fully utilize SATA 6Gb/s drives.

In order to debunk this myth, we examined a number of supposed SATA 3Gb/s cables and compared them to a SATA 6Gb/s cable in our SATA cables: Is there a difference? article. In that article, we found no significant physical or performance difference between any of the cables we tested. Due to how long ago that testing was performed, however, we have recently had some readers that have questioned the relevancy of that article. SATA drives are faster now than they were when SATA 6Gb/s was new and advances in SATA controller technology might change our results.

In order to determine if there is still no difference between SATA cables, we decided to revisit this topic. Luckily, relevant hardware and cables are readily available for us as a custom computer company, so this type of testing is very quick and easy for us to do. We will be using the latest Intel Z77 chipset, the fastest mainstream Intel SSD available, and cables that are currently being bundled with both Asus motherboards and Intel SSDs.

Test Cables

Some of the cables we used in our original SATA cable article are no longer being manufactured, so we decided to update our test cables to reflect those that you would find included with modern hardware. Many of these cables are not actually labeled as either SATA 3Gb/s or SATA 6Gb/s, but by researching the model number of the cable itself we were able to find the rated SATA revision for these cables.

Asus SATA 6Gb/s
P/N: 14G000130223
Asus SATA 3Gb/s
P/N: 14G00012707H
Intel SATA 3Gb/s
P/N: E156437
Intel SATA 3Gb/s
P/N: E92245-001

The Asus cables are what come bundled with almost every modern Asus motherboard and are labeled in the manual as being compatible with either SATA 3Gb/s or SATA 6Gb/s. The two Intel cables are bundled with the retail version of Intel's SSD drives and required some research to find the compatible SATA revision. The black cable (E156437) is included with the Intel 320 series of SSDs, while the red cable (E92245-001) is included with the Intel 520 series of SSDs. Interestingly, looking up the specs for the actual cable reveals that both of these cables are only rated for SATA 3Gb/s even though the Intel 520 series of SSDs are rated for SATA 6Gb/s speeds. So clearly, Intel is not worrying about which SATA revision their cables are rated for.

Since our original SATA cable article, a few things have changed in regards to the physical cables. First, most of the cables we are testing have four ground wires (the bare wires) rather than three that was the norm in our last article. Second, one of the cables (The black Intel cable) actually has a slightly smaller gauge of wire. By our measurements, this wire (not counting the shielding) is .25mm in diameter (30 gauge), while the other cables are .4mm in diameter (26 gauge). We will find out if this has any impact on performance when we run our performance testing later in this article.

Test Setup

To compare the performance of each cable, we configured our test SSD as a secondary storage drive and ran CrystalDiskMark with each cable. We will be testing using both compressible (All 0x00) and incompressible (Random) data. We recently benchmarked the performance difference between these two types of data in our SSDs: Advertised vs. Actual Performance article, and the difference is large enough that we feel it necessary to use both types of data in our testing.

Our test system consists of the following hardware:

Testing Hardware
Motherboard: Asus P8Z77-V Pro
CPU: Intel Core i7 3770K 3.5GHz
CPU Cooler: Gelid Tranquillo Rev2
RAM: 2x Kingston DDR3-1600 4GB
Video Card: Intel HD 4000
Hard Drive: Intel 520 480GB SATA 6Gb/s
Chassis: Open Air Test Bench
Testing Station

The Intel 520 480GB is one of the fastest mainstream drives currently available, with an advertised Sequential Read speed of 550 MB/s and a Sequential Write speed of 520 MB/s. This is much higher than the actual throughput of SATA 3Gb/s (roughly 300MB/s), so it should be very clear if the drive is being limited to SATA 3Gb/s speeds. Between each test we will perform a secure erase on the hard drive to ensure that the drive is operating at peak efficiency.

Performance Results

Starting with compressible data, we see virtually no performance difference with any of the cables. There are small variations, but all are well within the margin of error for a hard drive benchmark.

For incompressible data, we again see no performance difference outside of normal testing variations. Even the black Intel cable with a thinner gauge of wire shows no performance drop at any point in our testing.


Our results confirm that despite the faster hardware available today, there is still no performance difference between SATA 3Gb/s and SATA 6Gb/s cables. The SATA 3Gb/s revision only supports transfer speeds around 300MB/s, yet we saw transfer speeds up to 500 MB/s with each cable that we tested. This clearly shows that the SATA revision designator on cables is mostly just marketing and has no bearing on the actual performance the cable can provide.

This is not to say that all cables are created equally, but rather that you cannot base the quality of the cable from the SATA revision it is supposed to be compatible with. A wire is a wire, and as long as the gauge of the wire is adequate, the end connections are good, and the right metal is used, there is no performance difference between one cable and another. The place where some users get into trouble is when they are using a particularly cheap cable that has either bad connections or uses sub-par materials. Even in those instances, however, you would see problems with the drives dropping or not being detected long before you see any sort of decrease in performance.


And thats a wrap

Posted on 2013-02-15 20:14:51

Thank you for taking the time to actually do it. Very well written article and great info.

Posted on 2013-02-18 05:40:56
Diego Zenizo

" long as the gauge of the wire is adequate..." Here there may be a differences among cables IF the geometry of the cables themselves (impedance) is different, since impedance not only depends on the gauge but on the way cables are put together, like the differences between Cat 3, 4, 5, 5e, 6 etc in UTP cable for RJ-45 ethernet cabling.

In these tests, you're using same brand cables, what may suggest similar or even identical manufacturing facilities, thus similar or identical cable construction regardless of the intended speed.
It may be worthwhile to compare different brands which may have different construction characteristics.

Posted on 2013-02-27 18:27:25

In this test, there was actually at least three different actual brands of SATA cable. I don't know of any motherboard manufacturers that actually make their own cables (although Asus might since I couldn't find their model numbers referred to as anything but Asus) but rather just re-brand cables. For example, the two "Intel" cables are actually made by Golden Bridge and Bizlink.

The way the cables are put together shouldn't matter as far as SATA cables go. The reason for the differences in CAT 5,6,etc. is primarily due to how tightly twisted together the wire pairs are. The closer you get the wires to 90 degrees (which is achieved by twisting the wires together as many times as possible per inch), the lower the interference. has a decent explanation of why twisted pair wires reduce interference.

As far as what I remember from my Physics classes, since the wires in SATA cables are parallel to each other, the only thing that would affect the signal quality (assuming that the wire material, gauge, distance and voltage are the same) would be the distance between the wires. Since SATA cables vary very little in terms of spacing, theoretically there should be no difference in the signal quality (and thus SATA speed) between cables, which is exactly what our testing shows.

Posted on 2013-02-28 19:54:15

Sadly, because most SATA cables are not twisted pair (Akasa make some they claim are) they are vulnerable to interference despite having internal shielding. I don't know why the SATA group didn't choose cat. 6, the electrical specification is almost the same and it's got inherent noise rejection from the twists. 4-lane SATA in the spec would be able to utilize the extra pairs too. The SATA spec calls for 30-26ga wire which is much thinner than cat. 5 but plenty for lengths<1m.

On cable quality, there are cables that do not work reliably at the rated speed, I have 4 that are unusable with a new Toshiba HDD.

Posted on 2013-08-09 19:24:20

Did you check if the wires are pure copper or tinned copper?

Posted on 2013-05-20 18:26:23

Excellent evaluation and summary. Very useful for my current task at hand, replacing an SSD to a 6 Gb/s ASUS motherboard. Now I won't unnnecessarily go out looking for cables to replace the ones I have

Posted on 2013-07-04 16:57:48

Nice report, much appreciated. Thanks!

Posted on 2013-07-29 10:09:37

Thanks, it was bugging me too, as I had unpacked and mixed up the cables before noticing a difference.

Posted on 2013-09-24 04:03:15

Thank you for this! Was very helpful, saved me some tears and €€.

Posted on 2013-10-16 16:38:13

Very clear and well done.

Posted on 2013-11-26 17:37:28
Ric Hales

Thanks for this confirmation, it's just saved me a few quid. I wasn't sure I'd be able to reuse my old SATA cables for my new P.C, but now I know that I can...Awesome!

Posted on 2013-12-11 16:08:50
Jessica Park

Nice post you have here. Here is Nutone replacement parts for you. Thank you so much for this post.

Posted on 2014-01-19 13:52:07
Jeremy Tyler

Is there a test that will show the data transfer rate as 6 gb/s or 3 gb/s or maybe a math to convert test #'s to reveal the actual #? I can't seem to find the answer to this. I recently installed an ssd that is 6 gb/s and the motherboard states that the port is 6 gb/s. I might have to concede to assumption that things are right but there's nothing like "proof!"

Posted on 2014-01-24 14:05:07
Cheska Gomez

This is a nice post. Like seriously, it might not be that easy to do something like this unless your a pro but your article explains it all.

Nutone parts

Posted on 2014-04-16 02:24:55

"A wire is a wire, and as long as the gauge of the wire is adequate, the end connections are good, and the right metal is used, there is no performance difference between one cable and another." That conclusion lacks any scientific foundation. How would you define "cheap" materials? Specifications for SATA cables include such characteristics such as: insulation resistance, dialectrical withstanding voltage, low level contact resistance, contact current rating, cable pull out, cable flexing, insertion force, removal force, durability, affect of vibration, humidify, temperature-life, thermal shock, mixed flowing gas, mated connector impedance, absolute impedance, cable pair matching, common mode impedance, insertion loss, crosstalk, rise time, etc. You have measured non of these characteristics. Grabbing some cables and failing to demonstrate that any impede the data transfer rate is not particularly useful information. We know nothing about the comparative length of the cables, and the environment under which they were "stressed". Performing the specification tests on a group of cables would be much more valuable information. It would give users a quantifying measure of "quality" to direct their decisions on which cable to purchase. Would your conclusion that "it makes no difference" hold up when using longer cables, placing bends in these cables, routing these cables closer to noise generating components? Your conclusions seem to add to the confusion rather than reduce it.

Posted on 2015-06-20 11:04:28
uumanebs .

So, we should test Usain Bolt's genetic material to determine if he's the fastest man in the world, rather than having him run competitively.

This isn't a "Who makes the best cable" thing. It's a "Do i really need to spend the extra cash for a new cable" thing. This just shows that if you're worried about needing to go buy the new models, you don't need to.

I get what you're asking, but it's like complaining that an article about the weather in China isn't informing you of the weather in Europe.

Posted on 2016-11-09 17:48:31
Jason Chester

The stanard and what the cable can physically perform at are not on in the same.

Posted on 2015-11-23 11:48:03
Todd Smeester

I just hooked up a samsung ssd, when using the samsung magician, it tells me I'm only connected to the sata 2.0, when clearly I can see that I'm connected to the sata 3.0/6GB port. Any ideas? I have the Asus CM6850 motherboard. I'm plugged into the blue port that's all by itself, and it's stamped SATA 3G_4 next to it.

Posted on 2016-07-10 14:46:58
Steve Wozniak

Thank you for the analysis! Appreciate the work.

So, what is a good cable to buy?

Posted on 2016-08-17 15:15:11

Thanks for your time to do this benchmark! I was under the assumption that was the cable the bottleneck on my system. While doing DBAN on two identical drives, one is 3 times slower. I was "hoping" it was the cable...

Posted on 2016-11-08 12:17:00

Just what I was looking for before using some old sata cables. thank you

Posted on 2017-01-28 23:33:49
Lou Vile

Thats fine and dandy but you did not answer my actual question or maybe you did, And that is, does the cable and or connector determine the speed of the entire system? The answer would seem to be yes.

Posted on 2017-03-11 13:03:34