in the SSD department the best performance gains were made from SATA3 600MB to Nvme pcie gen3 X4 4000MB, and only if pc games had direct storage enable a pce gen4 X4 will be enough.
I was surprised by the results so I added Forza Motorsport to my tests to see if Direct Storage would have an impact, but since both drives can use it there was no difference. Unfortunately there is a perception out there that since the read speeds double the performance in games will improve but it’s simply not true (other than a small benefit to load times).
for gaming yes I don't think we see any performance improvements kinda like 10Gig port at at home you'll never need or use it. But what about productivity folks? say video/photo editing folks that transfer large data back and forth I can see this being a big improvement for these guys!
The problem is they target gamers with their advertising … I fell for it and purchased multiple drives so I was genuinely curious when I started testing them.
Interesting, I didnt know the difference between the SATA and NVME, thanks for the lesson! It is a shame that there was no real performance boost but at least now I am more aware that I should save money on my drives and put it towards something else. Normally I just buy as much storage as possible and don't really think about speed (hence why I am yet to upgrade to an SSD for my other non-C drives) 😅Was there any significant differences in the boot times? Great video btw!
Thanks and yes, there was a 25% reduction in load time for games (it's on the 3DMark Storage Benchmark chart labeled Access Time) ... I was genuinely surprised by the results and regret buying all of the Gen 5 M.2 SSDs that I have.
@blackbirdpctech oh wow, I'm sure most people will not want to spend double the price for even 25% less load times. At least you did the test now rather than later, now you know not to fall for the marketing scams and this shows us viewers to be weary of these marketing tricks used by these companies to make a quick bag!
That’s the point of the video, to show that Gen 5 SSD’s provide no performance benefit in games. Furthermore, I did test load times, but the difference between gen 4 and gen 5 is so small that the only way to accurately measure it is with benchmark software.
I did test load times ... the second chart shows "Access Time" results from the 3DMark Storage Benchmark which measures load time across multiple games such as Battlefield V, COD: Black Ops 4, Overwatch, Counter-Strike, etc. Games load so fast on Gen 4 and Gen 5 SSDs that physically measuring load times with a stop watch is extremely inaccurate, so using a benchmark that is repeatable is a much better way to do it.
From my understanding, right now it doesn't matter if you have a Gen 4 or Gen 5 if you have an Intel Motherboard, because "if you have your Gen 5 SSD installed in the first M.2 slot, the bandwidth of your graphics card slot will be halved and my RTX 4090 will operate at PCIe 4.0 x8 link speeds instead, thereby impacting your graphics card's performance." So basically until Intel boards come out with designated slots for the Gen , like the AMD boards already have, it is not worth it. Pretty annoying when trying to build right now but held up by a stupid slot in a board. It's like do you buy a Gen 5 and just put it in the 2nd slot etc this way when you go to upgrade later you have it already or just buy Gen 4 and buy Gen 5 or 6 by the time you want to upgrade your Mobo again.
That is only part of the story, which I covered in the video, but that is not why it doesn't matter. I conducted the testing for this video on an AMD X670e board with full bandwidth available to the SSD and GPU ... Gen 5 M.2 SSD's simply don't provide any performance benefit in games except for slightly lower load times. I think in this situation Intel made the right move ... even if you enable the Gen 5 SSD slot and halve your GPU bandwidth on an Intel Z790 board (which I wouldn't recommend) it will only make a very small difference to game performance.
why do they have to make pc terms so confusing and unintuitive? samsung makes a gen 5 SSD 2 tb for about 160 on amazon right now. it says its 5,000 mp/bs as opposed to crucial being 12,000 for over $200. i wonder if those advertised speeds are accurate. Also wonder if the samsung gen 5 SSD is truly gen 5 or amazon is mislabeling it.
To make it even more confusing the Samsung drive, which I think is the 990 EVO, is a hybrid Gen 4 / Gen 5 drive ... it only offers Gen 5 at half the lanes or bandwidth (i.e. Gen 5x2), which is why the speeds are about half that of Gen 5x4 drives. So yes, the advertised speeds would be accurate. It actually looks like a great drive to pickup.
I am curious if the Gen5 drive "feels" more snappy in day to day usage compared to the Gen4 drive. Some things can't be seen in a regular benchmark. For instance, Intel Optane drives would lose in all of those benchmarks, but they actually have the best latency input.
I didn't notice any real difference ... even with load times, the Gen 5 drive is faster but the Gen 4 drive is already so fast that it wasn't really meaningful. Perhaps that will change with newer motherboards and CPU architectures.
Thanks the explanation, I think I'll stick to my Crucial P3 Plus SSD choice for my new build for now! A bit of an off topic question, but given the Asus controversy, can you recommend a non Asus motherboard for a Ryzen 7 3800X3D?
No problem ... I really like the motherboard that I use in my test bench, it's the Gigabyte X670E Aorus Master (link is in the description to this video). MSI also has some great boards ... depends on how much you are looking to spend.
@@blackbirdpctech I should have specified, this will be a small form factor mitx build so my choices there would be the Gigabyte A620I AX, the Aorus Ultra or MSI's MPG B650I Edge WiFi from the looks of things. Sadly my budget is a bit too low for the Aorus Ultra at a max of £250 for the mobo.
They are backwards compatible, so if you put a gen 5 SSD in a gen 4 slot then it will be limited by gen 4 speeds. That said it will not have any impact on gaming performance. But I would highly recommend just getting a high speed Gen 4 SSD because the temps will be lower.
Yeah, I bought into the hype and purchased a bunch of Gen 5 drives before running the benchmarks … this explains why Samsung has not rushed out a Gen 5 drive.
Slower performance with Gen 5 could be heat related? I see the cpu ran much hotter when paired with the gen 5 drive, so maybe cpu thermals were causing the lower framerate on modern warfare.
I thought the exact same thing so I tested it … temps were well within spec and the performance in crystaldiskmark was the same after and before. It was super strange.
@@blackbirdpctech very strange indeed. Must just be game optimization related. At any rate, thanks for the video. I just got the 2TB T500 for my laptop on prime day sale this morning for $138.
Yes however the difference in load times was so small that I didn't really notice in the games I play. The second chart shows "Access Time" results from the 3DMark Storage Benchmark which measures load time across multiple games such as Battlefield V, COD: Black Ops 4, Overwatch, Counter-Strike, etc. Games load so fast on Gen 4 and Gen 5 SSDs that physically measuring load times with a stop watch is extremely inaccurate, so using a benchmark that is repeatable is a much better way to do it.
It can run Gen 5 PCIe and an M.2 SSD at the same time but I think what you mean is that it can’t run the GPU at full bandwidth while running a Gen 5 SSD … I cover this in the video.
Agreed, at the moment the performance impact is relatively small but that may change. Intel needs to get their new platform out beforehand to make sure they are positioned well to avoid bandwidth issues.
The problem with comparing load times is that I couldn't find an accurate way to do it. If you look at the access time that is measured in the 3DMark Storage Benchmark you can see that it's measured in micro seconds. A human can't measure micro seconds accurately, so when you see tech UA-camrs use a stop watch the data is complete nonsense. So I was more interested in seeing if there was any FPS difference ... data I couldn't find anywhere else online. At least now with this data I don't feel like I need to buy Gen 5 SSD's for gaming.
gen 5 drives are like what ssd originally where to the regular 7200rpm drives ahead of there time and it took developers a while to get full use of those drives 7000 read is plenty and write at 6000 is plenty. maybe when the ps6 comes out gen 5 drives will be worth it. once sony and microsoft get there heads out of there rear end and start to push there tech like what sony did for ratchet and clank thus far is a start but cant help feeling the ssd in consoles arent being fully tapped since pc version says on ssd is required which i say to sony and developers which one ?. so far newer games don't seem to be pushing that boundary as hard yet. the issue i have with tests for load times is great but texture pop in is less for nvme at least from what ive observed in hogwarts legacy. will have to test with 4.0 and see if loading into different zones is more seamless. ( castle seems to be split into 4 areas.
I didn’t expect to see a huge difference but I did expect to see something … I was a little surprised with the results … and I still can’t explain the COD MW3 result. I think you are correct wrt game developers not leveraging the technology … which is why I added Forza Motorsport because it uses DirectStorage … but it had no impact on the results because the Gen 4 drive also used it.
@@blackbirdpctech as long as there developing for ps4 and older consoles that used standard drives. transition is going to be slow its improving with adopting it granted developers may leverage it more in long run once they have tapped everything else out i believe it could be that the game engine doesn't have the support built in for anything faster then the ps5 and xbox one ssd. i think what you may see more is less texture pop in that just isn't shown in the data. This is my own personal view of where i think developers are focusing on is in the mid market gpus like a 3060 12gb. A thing to try would be using a older 3500/3000 drive on cod mw3 vs m.2 4. and see if it preforms more poorly. if you look at iw game engine it may give you some insight. wikipedia has a good list of its features. i get the feeling the cod mw3 was in development before the ps5 and xbox one which may also explain its behaviour kind of like they hobbled it in at the end lol. other thing to consider both games where probly developed around 2017. which would be nvme 1.3 speed.
I was genuinely surprised by the results too which is why testing is valuable, it provides evidence to make a sound assessment versus using emotion. Just because you think something should be a particular way doesn't mean it is. My data is accurate, so please feel free to provide data to support your thoughts that you expressed so eloquently 😉
@blackbirdpctech mw3 has recently forced a setting called on demand texture streaming, it requires 64gb free of ssd space, could this be the cause for difference in fps, there is 2 options for this setting, is it possible to retest this game, I haven't seen anyone test the performance on this based on ssd speed and could become more common with future games, thanks
Very informative video. Thanks a lot.
Glad you liked it, appreciate the feedback.
in the SSD department the best performance gains were made from SATA3 600MB to Nvme pcie gen3 X4 4000MB, and only if pc games had direct storage enable a pce gen4 X4 will be enough.
I was surprised by the results so I added Forza Motorsport to my tests to see if Direct Storage would have an impact, but since both drives can use it there was no difference. Unfortunately there is a perception out there that since the read speeds double the performance in games will improve but it’s simply not true (other than a small benefit to load times).
for gaming yes I don't think we see any performance improvements kinda like 10Gig port at at home you'll never need or use it. But what about productivity folks? say video/photo editing folks that transfer large data back and forth I can see this being a big improvement for these guys!
The problem is they target gamers with their advertising … I fell for it and purchased multiple drives so I was genuinely curious when I started testing them.
@@Widowmaker28A 🫵all facts 👍
Thank you!
You are very welcome!
Such an insightful video. Thank you for all the value you provide that makes sense even to newbies like me.
Thanks for watching and the feedback!
Interesting, I didnt know the difference between the SATA and NVME, thanks for the lesson! It is a shame that there was no real performance boost but at least now I am more aware that I should save money on my drives and put it towards something else. Normally I just buy as much storage as possible and don't really think about speed (hence why I am yet to upgrade to an SSD for my other non-C drives) 😅Was there any significant differences in the boot times? Great video btw!
Thanks and yes, there was a 25% reduction in load time for games (it's on the 3DMark Storage Benchmark chart labeled Access Time) ... I was genuinely surprised by the results and regret buying all of the Gen 5 M.2 SSDs that I have.
@blackbirdpctech oh wow, I'm sure most people will not want to spend double the price for even 25% less load times. At least you did the test now rather than later, now you know not to fall for the marketing scams and this shows us viewers to be weary of these marketing tricks used by these companies to make a quick bag!
I already purchased a bunch of Gen 5 SSD’s, so it’s too late for me but at least I can help the broader community save some money 😉
@@blackbirdpctech our saviour 🙌🏼
You said FOMO in a tech review. Legend!
That was the only reason I could think of to buy a gen 5 ssd given these results … and I purchased a bunch of them 😉
@@blackbirdpctech LOL!!!
Yep ... wish I had done this testing before buying.
@@blackbirdpctech Well, it could have been worse. You could have said FML instead of FOMO.
Ha ha … that would have been more appropriate with the last video 😉
Just bumped into this. In what world does the speed of your SSD have to do with frame rate. Better off testing load times
That’s the point of the video, to show that Gen 5 SSD’s provide no performance benefit in games. Furthermore, I did test load times, but the difference between gen 4 and gen 5 is so small that the only way to accurately measure it is with benchmark software.
Yeah, ok, you tested FPS for games, but how about *loading times* when you're testing a storage device??
I did test load times ... the second chart shows "Access Time" results from the 3DMark Storage Benchmark which measures load time across multiple games such as Battlefield V, COD: Black Ops 4, Overwatch, Counter-Strike, etc. Games load so fast on Gen 4 and Gen 5 SSDs that physically measuring load times with a stop watch is extremely inaccurate, so using a benchmark that is repeatable is a much better way to do it.
From my understanding, right now it doesn't matter if you have a Gen 4 or Gen 5 if you have an Intel Motherboard, because "if you have your Gen 5 SSD installed in the first M.2 slot, the bandwidth of your graphics card slot will be halved and my RTX 4090 will operate at PCIe 4.0 x8 link speeds instead, thereby impacting your graphics card's performance." So basically until Intel boards come out with designated slots for the Gen , like the AMD boards already have, it is not worth it. Pretty annoying when trying to build right now but held up by a stupid slot in a board. It's like do you buy a Gen 5 and just put it in the 2nd slot etc this way when you go to upgrade later you have it already or just buy Gen 4 and buy Gen 5 or 6 by the time you want to upgrade your Mobo again.
That is only part of the story, which I covered in the video, but that is not why it doesn't matter. I conducted the testing for this video on an AMD X670e board with full bandwidth available to the SSD and GPU ... Gen 5 M.2 SSD's simply don't provide any performance benefit in games except for slightly lower load times. I think in this situation Intel made the right move ... even if you enable the Gen 5 SSD slot and halve your GPU bandwidth on an Intel Z790 board (which I wouldn't recommend) it will only make a very small difference to game performance.
why do they have to make pc terms so confusing and unintuitive? samsung makes a gen 5 SSD 2 tb for about 160 on amazon right now.
it says its 5,000 mp/bs as opposed to crucial being 12,000 for over $200. i wonder if those advertised speeds are accurate. Also wonder if the samsung gen 5 SSD is truly gen 5 or amazon is mislabeling it.
To make it even more confusing the Samsung drive, which I think is the 990 EVO, is a hybrid Gen 4 / Gen 5 drive ... it only offers Gen 5 at half the lanes or bandwidth (i.e. Gen 5x2), which is why the speeds are about half that of Gen 5x4 drives. So yes, the advertised speeds would be accurate. It actually looks like a great drive to pickup.
I am curious if the Gen5 drive "feels" more snappy in day to day usage compared to the Gen4 drive.
Some things can't be seen in a regular benchmark.
For instance, Intel Optane drives would lose in all of those benchmarks, but they actually have the best latency input.
I didn't notice any real difference ... even with load times, the Gen 5 drive is faster but the Gen 4 drive is already so fast that it wasn't really meaningful. Perhaps that will change with newer motherboards and CPU architectures.
Thanks the explanation, I think I'll stick to my Crucial P3 Plus SSD choice for my new build for now! A bit of an off topic question, but given the Asus controversy, can you recommend a non Asus motherboard for a Ryzen 7 3800X3D?
Good choice … do you mean 5800X3D or 7800X3D?
@@blackbirdpctech my bad, missed the typo there, I mean the 7800 my apologies.
No problem ... I really like the motherboard that I use in my test bench, it's the Gigabyte X670E Aorus Master (link is in the description to this video). MSI also has some great boards ... depends on how much you are looking to spend.
@@blackbirdpctech I should have specified, this will be a small form factor mitx build so my choices there would be the Gigabyte A620I AX, the Aorus Ultra or MSI's MPG B650I Edge WiFi from the looks of things. Sadly my budget is a bit too low for the Aorus Ultra at a max of £250 for the mobo.
Of those options I would get the MSI board … they seem to be really good with bios updates lately and I generally like the look of their boards.
What if u put a gen5 ssd in a gen4 pc? What will be the max speed?
They are backwards compatible, so if you put a gen 5 SSD in a gen 4 slot then it will be limited by gen 4 speeds. That said it will not have any impact on gaming performance. But I would highly recommend just getting a high speed Gen 4 SSD because the temps will be lower.
I was just wondering about this, I have gen5 with a gen 4x4 drive...so, no real difference.
Yeah, I bought into the hype and purchased a bunch of Gen 5 drives before running the benchmarks … this explains why Samsung has not rushed out a Gen 5 drive.
Slower performance with Gen 5 could be heat related? I see the cpu ran much hotter when paired with the gen 5 drive, so maybe cpu thermals were causing the lower framerate on modern warfare.
I thought the exact same thing so I tested it … temps were well within spec and the performance in crystaldiskmark was the same after and before. It was super strange.
@@blackbirdpctech very strange indeed. Must just be game optimization related. At any rate, thanks for the video. I just got the 2TB T500 for my laptop on prime day sale this morning for $138.
Great deal for a great drive!
Did games load faster or was it even noticeable ?
Yes however the difference in load times was so small that I didn't really notice in the games I play. The second chart shows "Access Time" results from the 3DMark Storage Benchmark which measures load time across multiple games such as Battlefield V, COD: Black Ops 4, Overwatch, Counter-Strike, etc. Games load so fast on Gen 4 and Gen 5 SSDs that physically measuring load times with a stop watch is extremely inaccurate, so using a benchmark that is repeatable is a much better way to do it.
Not with Intel's current offerings itx Intel 14th gen can't run Gen 5 PCIe and M.2 at same time.
It can run Gen 5 PCIe and an M.2 SSD at the same time but I think what you mean is that it can’t run the GPU at full bandwidth while running a Gen 5 SSD … I cover this in the video.
@@blackbirdpctech yeah I'm wondering when Gen 5 GPUs drop how's it'd react if it's drop to PCIe Gen 4 x8 then Gen 5 x4 it'll be interesting to see
Agreed, at the moment the performance impact is relatively small but that may change. Intel needs to get their new platform out beforehand to make sure they are positioned well to avoid bandwidth issues.
The focus of the video should have been loading times in my eyes. I would not expect a difference in FPS between gen4 and gen5. Interesting though.
The problem with comparing load times is that I couldn't find an accurate way to do it. If you look at the access time that is measured in the 3DMark Storage Benchmark you can see that it's measured in micro seconds. A human can't measure micro seconds accurately, so when you see tech UA-camrs use a stop watch the data is complete nonsense. So I was more interested in seeing if there was any FPS difference ... data I couldn't find anywhere else online. At least now with this data I don't feel like I need to buy Gen 5 SSD's for gaming.
@@blackbirdpctech appreciate your work
gen 5 drives are like what ssd originally where to the regular 7200rpm drives ahead of there time and it took developers a while to get full use of those drives 7000 read is plenty and write at 6000 is plenty. maybe when the ps6 comes out gen 5 drives will be worth it. once sony and microsoft get there heads out of there rear end and start to push there tech like what sony did for ratchet and clank thus far is a start but cant help feeling the ssd in consoles arent being fully tapped since pc version says on ssd is required which i say to sony and developers which one ?. so far newer games don't seem to be pushing that boundary as hard yet. the issue i have with tests for load times is great but texture pop in is less for nvme at least from what ive observed in hogwarts legacy. will have to test with 4.0 and see if loading into different zones is more seamless. ( castle seems to be split into 4 areas.
I didn’t expect to see a huge difference but I did expect to see something … I was a little surprised with the results … and I still can’t explain the COD MW3 result. I think you are correct wrt game developers not leveraging the technology … which is why I added Forza Motorsport because it uses DirectStorage … but it had no impact on the results because the Gen 4 drive also used it.
@@blackbirdpctech as long as there developing for ps4 and older consoles that used standard drives. transition is going to be slow its improving with adopting it granted developers may leverage it more in long run once they have tapped everything else out i believe it could be that the game engine doesn't have the support built in for anything faster then the ps5 and xbox one ssd. i think what you may see more is less texture pop in that just isn't shown in the data. This is my own personal view of where i think developers are focusing on is in the mid market gpus like a 3060 12gb. A thing to try would be using a older 3500/3000 drive on cod mw3 vs m.2 4. and see if it preforms more poorly. if you look at iw game engine it may give you some insight. wikipedia has a good list of its features. i get the feeling the cod mw3 was in development before the ps5 and xbox one which may also explain its behaviour kind of like they hobbled it in at the end lol. other thing to consider both games where probly developed around 2017. which would be nvme 1.3 speed.
I'd prefer you were a former Brain Surgeon. lol
Sorry to disappoint 😉
Yeah.
For gaming it makes no sense paying the extra money on a Gen5 drive.
I learned this the hard way after I had already purchased multiple Gen 5 drives.
🔥 🔥 🔥
🤣🤣
Absolute bullshit you get higher FPS with a. gen5 NVME.
Absolute bullshit🤣🤣🤣
I was genuinely surprised by the results too which is why testing is valuable, it provides evidence to make a sound assessment versus using emotion. Just because you think something should be a particular way doesn't mean it is. My data is accurate, so please feel free to provide data to support your thoughts that you expressed so eloquently 😉
@blackbirdpctech mw3 has recently forced a setting called on demand texture streaming, it requires 64gb free of ssd space, could this be the cause for difference in fps, there is 2 options for this setting, is it possible to retest this game, I haven't seen anyone test the performance on this based on ssd speed and could become more common with future games, thanks