Review Index:
Feedback

PS4 Pro SSD Upgrade: Does SATA III Make a Difference?

Author: Jim Tanous
Subject: Storage
Manufacturer: Sony
Tagged: ssd, ps4 pro, ps4, consoles

Testing Methodology & Results

Games & Testing Methodology

For this article, we tested three configurations:

  1. Original, non-Pro PS4 with Samsung 850 EVO SSD
  2. PS4 Pro with stock 1TB HDD (HGST 5400RPM)
  3. PS4 Pro with Samsung 850 EVO SSD

Each configuration featured a clean installation of the PS4 system software and each tested game. The games and the way we tested each are described below:

View Full Size

Fallout 4: load saved game in the heart of Diamond City

NHL 17: load single-player game

GTA V: load single player from launch and switching characters

Call of Duty Infinite Warfare: load ground mission (Operation Burn Water) and space combat mission (Operation Safe Harbor)

The Witcher 3: load saved games in the Nilfgaardian Garrison and in the center of Novigrad

The video for each game's test was directly captured to a PC and we used playback of these captured video files to determine the loading time. Some tests, specifically those for GTA V, were impossible to compare objectively between systems, as the game world is constantly in flux and each time a saved game is loaded, or the player switches characters, the position of the loaded characters and complexity of their environment is different. Therefore, for the GTA V tests, we loaded each scenario five times for each PS4 configuration and then averaged the results.

PS4 SSD Loading Times

First up is Fallout 4, where we took our character into Diamond City, one of the most densely-populated areas of the game.

View Full Size

Loading a saved game from the heart of Diamond City, we see that the PS4 Pro's SATA III interface does indeed seem to help. The non-Pro PS4 with the Samsung 850 EVO SSD reduced loading times by about 9 seconds, or 18.5 percent, while the same SSD in the PS4 Pro reduced loading times by a further 13 seconds, resulting in a roughly 46 percent improvement over the PS4 Pro's stock hard drive and 33 percent over the SATA II-limited original PS4.

Next is NHL 17, in which we tested loading a single player game featuring the Toronto Maple Leafs at the Buffalo Sabres.

View Full Size

We can see from the results that the storage interface doesn't appear to be the bottleneck for this title, with the SSD in the PS4 Pro offering only a 2-second loading time advantage over the stock HDD. The original PS4's SSD took even a second longer to load than the PS4 Pro's HDD, further suggesting that NHL 17's single player loading time is dictated more by the console's processing power rather than loading assets from the storage drive.

As mentioned earlier, the results presented for Grand Theft Auto V are averages of five tests each due to the game's dynamic nature, which makes repeatedly testing an identical situation impossible.

View Full Size

The averaged data still reveals some interesting results, however. The game's initial load time for the single player experience is heavily dependent on the storage drive, with the PS4 Pro's SSD loading the game at a relatively blistering 25 seconds compared to the 73 seconds required for the PS4 Pro's stock HDD. This is a significant 66 percent improvement over the stock HDD and 58 percent improvement over the non-Pro PS4's SSD configuration.

When it comes to switching characters, however, the advantages of the SSD and a SATA III interface are virtually nonexistent. The time it takes to switch characters can vary significantly depending on the location of both your current character and the character you're switching to. Still, our testing revealed that the SSD didn't help this process, with both SSD and HDD-based systems taking as much or as little time as the other over our five test runs, with an average for all of 13 to 14 seconds. So while an SSD will help you get into the single player game much faster, don't expect much when switching between characters during gameplay.

Turning to Call of Duty: Infinite Warfare, the game features many varied environments on both land and in space. We chose to test the first stage of the Operation Burn Water mission for ground combat, and Operation Safe Harbor for space combat.

View Full Size

The results show that the PS4 Pro's SSD provides a noticeable 4-second, or 28 percent, improvement in loading times for the ground combat, but reveals that space combat is fairly well optimized with practically identical load times between the Pro and non-Pro PS4. Even with the PS4 Pro's stock HDD, loading a space combat mission takes only about 9 seconds.

Finally, we looked at The Witcher 3, and wanted to test both moderately- and heavily-populated areas. We therefore tested loading saved games from within the Nilfgaardian Garrison and from within the heart of Novigrad.

View Full Size

With a moderately populated area like the Nilgaardian Garrison, the PS4's bottleneck appears to be processing power rather than storage, as all three configurations loaded in approximately the same time, with only 2 seconds separating the SSD configurations from the PS4 Pro's stock HDD.

The much more densely-populated Novigrad was a different story, however, with the PS4 Pro's SSD cutting 17 seconds, or 22 percent, off the load time compared to the stock HDD, and 11 seconds (15 percent) compared to the non-Pro PS4's SSD.

Video News


May 22, 2017 | 02:33 PM - Posted by flippityfloppit...

What am I doing here at eurogamer, and where did pcper go?

May 22, 2017 | 09:14 PM - Posted by Jim Tanous

Touché. But the more markets that have an interest in PC components, like SSDs, the better. Right?

May 23, 2017 | 01:39 PM - Posted by flippityfloppit...

Hah, I'm glad you have a sense of humour about my dry "joke".

I dunno, I'm personally pretty much against console anything these days as it is always behind a walled ecosystem, and the consumers that follow along breathlessly throw so much money at bad products because they are told to, that we are left with companies not innovating or advancing. (DLC, micro-transactions, unfinished products, etc.)

I'd rather the way PC games are treated thanks to the consoles popularity not seep it's way into PC hardware.

But I'm probably just jaded.

March 9, 2018 | 06:12 PM - Posted by Richard Young (not verified)

Except what you say isnt completely true. as AAA titles on PC more often then not are console ports. as for closed system i would take that over open easily hacked environments of the PC world. unless your a lifer or a pro the ping times and slightly better graphics dont over shadow the hacking that goes on in pc gaming.

March 9, 2018 | 06:13 PM - Posted by Anonymous2018 (not verified)

Except what you say isnt completely true. as AAA titles on PC more often then not are console ports. as for closed system i would take that over open easily hacked environments of the PC world. unless your a lifer or a pro the ping times and slightly better graphics dont over shadow the hacking that goes on in pc gaming.

May 23, 2017 | 08:53 AM - Posted by Xebec

Well from a hardware perspective, the PS4 / Pro is basically a PC at least..

May 23, 2017 | 02:59 AM - Posted by psuedonymous

A more interesting (but more extensive, and difficult to graph as a quantity) test would be to look at how availability of a fast SSD affects the streaming of assets in-engine. Texture and geometry 'pop-in' are the most obvious symptoms of bandwidth starvation, but those can also be due to lack of available memory.

May 23, 2017 | 10:47 AM - Posted by Jim Tanous

That's an excellent point. I imagine we'd need to find a game that has an in-engine rendered cutscene that exhibits pop-in, as the tests would need to be virtually identical for meaningful comparison. I'm anxious about the growing library of PS4 games piled on my desk, but I'll keep an eye out for a potential test subject.

July 12, 2017 | 08:11 AM - Posted by chrisuworo (not verified)

This comparison is not correct. Actually PRO does NOT use SSD potential, as it has build in bottleneck. Probably in software. It is required a lot of horsepower even to simply backup data at , lets say 85 000 IOPS that - theoretically both: SSD and SATAIII should provide. Due to lacklustre performance of AMD chip and SATA Northbridge we also have meager performance of USB (near 7,5MBps, to 8MBps). This article is phrased to imply that PS4 Pro benefits from SSD, whereas it benefits most from additional 1GB RAM for OS.

November 10, 2017 | 03:45 AM - Posted by ChrisD (not verified)

The SSD improves the loading/copying, in general the PS4 performance. I got mine yesterday and it is worth. Haters can get a life as they break the record of how many times can someone have the period in one month.

November 20, 2017 | 01:54 PM - Posted by techinthewoods (not verified)

It's possible sony is still using a version of the Eboot format they used previously in the PSP. This caused terrible loading times due to compression. Once the PSP hacks occurred, and ISO formats could be placed on the Memory stick there were highly significant loading time differences. Untold legends for instance, would load via UMD around 2 minutes, PSN download around 1min 20, and as an uncompressed ISO 15-20 seconds. Unless Sony Uncompressed the format of gaming media on PS4 load times may not improve, but this is also used as a security feature.

December 7, 2017 | 06:18 AM - Posted by B (not verified)

Right now a high quality 1tb SSD is around $600-700 and a 2tb is more than $1000. Very few people are willing to pay this amount just to have an additional 1tb of internal memory and/or a SATA 3 capable internal drive.

However there is the option of getting a SATA 3 SSHD internal drive of up to 2tb that will fit in the PS4 Pro.
Even though the PS4 Pro can handle up to 8tb of storage there are no internal drives greater than 2tb that will fit into the PS4 Pro.

For those who are not familiar with SSHD it is a drive that is mainly a HDD drive but utilizes a small SSD as its cache in order to increase its performance. It strikes a balance between cost and performance.

There is only one option I know of and it is made by Seagate. These currently cost less than $100. This is obviously an affordable choice for most people.

But seeing as the PS4 Pro can now use an external drive as extended memory why would you ever want to spend the money on replacing the internal drive?
For around $100 right now you can get an external 4tb HDD drive and just leave your internal HDD drive alone.

External drives are only available as HDD or SSD with the external SSD drives being just as prohibitively expensive as the internal ones.

There are only a few reasons then that might justify replacing your internal SATA 2 1tb HDD drive with an internal SATA 3 2tb SSHD drive.
Again if you just want more storage you can purchase an external HDD drive and there is no compelling reason to only install a 1tb internal SSHD drive seeing as a 2tb one is only slightly more expensive.

The most obvious reasons to install an internal SATA 3 2tb SSHD internal drive would be to take advantage of the PS4 Pro SATA 3 capability and to also take advantage of the moderate speed increases that an SSHD drive offers over an internal HDD drive.

How much would this speed up the load times and improve the game performance of the PS4 Pro?
I am going to guess here that the performance would be improved over that of the standard SATA 2 HDD internal drive that comes with the PS4 Pro but wouldn't be quite as fast an internal SATA 3 SSD drive.

Ultimately of course I just don't know but I believe that if you really want to be helpful to most PS4 Pro owners this is what you should be testing.
Few of us are willing to spend 600-1000+ dollars on SSD drives but spending less than $100 on an internal 2tb SSHD is something many of use would be up for if there was a compelling reason.

A lesser reason to install an internal 2tb SSHD drive would be to expand your storage but since you could get a 4tb external HDD at around the same price this alone will motivate few people.

Should I pay $85 for a 2tb internal SSHD drive in hopes of improved performance or should I just add an external HDD drive of whatever size I need?
This is my primary question and the one I would like to see your testing answer.

Obviously if the internal SATA 3 2tb SSHD drive offered significant enough improvements in performance that you were persuaded into getting it but you also needed more than 2tb of storage you could end up doing both.
In that case though you would have to keep in mind that the games stored on your internal SATA 3 SSHD drive might perform slightly better than the ones stored on your external SATA 3 HDD drive.
How much faster would the games on the internal SSHD drive perform than the ones on the external HDD?
That would be another good thing to test.

Of course in that scenario even the games on your external SATA 3 HDD drive might perform better than they did on the stock internal SATA 2 HDD drive that came with your PS4 Pro since both the system software on your internal drive and the game on the external drive would be on SATA 3 drives.
But maybe not?
That is something else that would be interesting to test.

December 7, 2017 | 03:43 PM - Posted by B (not verified)

I would like to add a few notes to my previous comment.

FIrst of all I have to make the correction that it is not a 1-2 tb SSD that costs from 600 -1000+ dollars it is a 2-4 tb SSD that cost that much.
A 1 tb SSD is about $300, which for most people is still a cost that would not be justified by the improvement in performances that an SSD gives.

You disagreed with the conclusions that the Digitalfoundry authors drew based mostly on the fact that they used a SSD that was towards the low end in performance. I have a few observations on your assertion.

By the way you can see a write up of their test by googling an article titled "Is it worth upgrading PS4 Pro with an SSD?". It is on the Eurogamer website as well as elsewhere.

They chose to test the PS4 Pro with a OCZ Trion 100 SSD because they had already tested a PS4 with the same SSD in the past and they wanted to be able to compare the two PS4 models using the same SSD.
I think this was smart and informative.
It would have been even better if they had also tested the PS4 Pro with a top quality SSD but they didn't.

Because they found that the differences in the load times between the two PS4 models using the same SATA 3 SSD was mild they assumed that the use of a SATA 3 drive in the SATA 3 capable PS4 Pro only provides moderate benefit.
They are likely correct in this assumption.

You also have to keep in mind that a percentage of the faster speeds seen in the PS4 Pro with the SATA 3 SSD were partially due to the increased processing power of the Pro, making the potential benefit of using a SATA 3 drive that much smaller.
How much of the increased performance was due to SATA 3 and how much was due to increased processing speeds is impossible to say from these tests.
It is safe to say they both probably played a role.

I think Digitalfoundry came to the correct conclusion that given the high cost of internal 1-2 tb SSD's at this time that the investment is not worth it for most people. This is true for having a SATA 2 or 3 SSD and having it in either the PS4 or the PS4 Pro.

They also observed that a SSD in a laptop/desktop PC greatly improves the performance of that PC much more than it does in the PS4 or the PS4 Pro.
They already knew this about the PS4 and conducted these tests in the hope that bottleneck in the PS4 was largely due to the SATA 2 architecture.
They hoped that the Pro would get a much bigger boost in performance from a SATA 3 SSD due to its SATA 3 capable architecture than the SATA 2 PS4 did and that this performance boost would justify the high cost of SSD's in a Pro.
It didn't get a much larger boost and the high costs were therefore not much more justified.

The biggest improvement in load times seen is obviously from simply having a SSD. Having a PS4 Pro with its increased processing improves load times as well, but to a lesser extent, and having a SATA 3 drive in your Pro probably also improves load times, but to an even lesser extent.
Keep in mind that we can't say for sure that a SATA 3 drive definitely improves anything from these tests, but it is likely that it does, at least a little.
This might give you the idea that maybe you could find an internal 1 or 2 tb SATA 2 SSD to use that you could get a great deal on since having a SSD drive is what improves load times the most.
I looked around and SATA 2 SSD drives are cheaper than SATA 3 SSD drives, but not by very much.

Besides load times having a SATA 3 SSD (or SSHD) in your Pro might also improve game performance for many games in ways that these tests could not reveal. That might be especially true for PSVR titles.
Or not.
All we know about SATA 3 in the Pro so far is that it probably mildly improves load times.
Regardless of SATA 3, having a Pro obviously improves graphics and game performance for Pro enhanced games as well as PSVR titles, as this is a large part of its sales pitch.

Going back to my previous comment the main question I had was if a SSHD in a Pro would improve performance enough to justify the $85 cost of a 2tb SATA 3 SSHD drive.
$85 is obviously a much lower bar to meet than the 300-700 dollars of 1-2 tb SSD's in terms of justifying the cost.
It is not worth it simply for the extra 1tb of storage as that can be done cheaper with an external HDD.

I think it is safe to say the potential SATA 3 improvement in load times by itself, being so small if existent at all, would not justify even the smaller $85 expenditure (at least for most people).
However the potential SATA 3 improvements plus the improved performance of the SSHD, due to its SSD components, together might justify the $85.
We know having a SSD is what most improves load times in a PS4 / PS4 Pro. Having a SATA 3 drive may or may not add a little to the performance as well.
How much of that improved performance will also be found with a SATA 3 SSHD? That is what I want to know.

Also wanted to say that I know that writing SSD drive or HDD drive is redundant. It was late when I made my first comment.

January 5, 2018 | 12:20 AM - Posted by Mriom (not verified)

To get Sata 3 speed you need a 7200RPM drive.. I'm not sure why DF only included the stock 5400rpm drive. A 7200rpm 1TB Disk is easy to find and cheap

May 15, 2018 | 11:05 AM - Posted by techinthewoods (not verified)

Digital Foundry also doesn't take into account a manufacturers compression method when doing drive tests. Ps4 titles are eboot.bin, the same compression they have used for years. Even in a digital download it will bottleneck a system with extraction time. The only way to test this properly is with an SSD connected, modified firmware to run uncompressed ISO files. It does not matter how expensive an SSD you purchase it does not control software compression algorithms.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

CAPTCHA
This question is for testing whether you are a human visitor and to prevent automated spam submissions.