SSD , NVMe PCIe 3.0 and NVMe PCIe 4.0 which one is better??? Mind you think before u post!!

Depends on the Motherboard and your Workload. How many Input Output Operations you need. If you are someone running 6 to help 7 virtual machines at the same time and working on a CRM along with a DB. NVMe will be crunching numbers like anything.
 
Depends on the Motherboard and your Workload. How many Input Output Operations you need. If you are someone running 6 to help 7 virtual machines at the same time and working on a CRM along with a DB. NVMe will be crunching numbers like anything.
So , the speed is about just loading an app faster, apart from that , as u have referred a WORK STATION with atleast 6-7 virtual machines, is fine, but still traditional SSDs are fine with the kinda heavy load u talk about, it is only in our imagination, IMHO, after watching this link.
 
^ Exactly that. Try extracting a Windows ISO on a sata ssd vs a nvme ssd. There's a very noticeable difference, when you're not cpu bound.

If your primary use case for a computer is for content consumption like videos and games, then there's no practical difference aside from slightly slower boot times with a sata ssd.

But if you're doing anything at all with files, like importing photos for editing, or even extracting rar's downloaded off the internet, then an nvme drive shows it's worth.

Actually traditional SSDs are faster than NVMe

If I remember correctly, that LTT video didn't test or allow for file copy tests, which would have shown the clear difference between the three technologies. It's also unclear if they did anything to prevent from Windows caching apps into memory, which would invalidate any subjective experience.

Another point to note is that they used a pretty substandard gen3 drive (an older entry level Corsair MP300, rated at 1600MB/s). Also back then, gen4 nvme had bios issues with X570. Other factors like SSD age/health and trim weren't disclosed. So it's a flawed test, or at the very least, incomplete.

If you use them long enough on a daily basis, you can even tell the difference between various gen3 drives.
 
So , the speed is about just loading an app faster, apart from that , as u have referred a WORK STATION with atleast 6-7 virtual machines, is fine, but still traditional SSDs are fine with the kinda heavy load u talk about, it is only in our imagination, IMHO, after watching this link.
They are not same in performance, I know it because I run 6 to 7 VM and work on a DB along with a CRM. Plus there would be other tabs running too at the same time. So for Heavy work load NVMe is the way to go but for a simple user, normal Sata based SSD would do.

Sometimes I have even powered up almost 15 Virtual Machines at the same time with no sweat.
 
If you're looking at drives like the MX500 or the 860, it's better to get a NVMe drive like the SN550 since the price difference isn't much and the perf gains are significant.
 
^ Exactly that. Try extracting a Windows ISO on a sata ssd vs a nvme ssd. There's a very noticeable difference, when you're not cpu bound.

If your primary use case for a computer is for content consumption like videos and games, then there's no practical difference aside from slightly slower boot times with a sata ssd.

But if you're doing anything at all with files, like importing photos for editing, or even extracting rar's downloaded off the internet, then an nvme drive shows it's worth.



If I remember correctly, that LTT video didn't test or allow for file copy tests, which would have shown the clear difference between the three technologies. It's also unclear if they did anything to prevent from Windows caching apps into memory, which would invalidate any subjective experience.

Another point to note is that they used a pretty substandard gen3 drive (an older entry level Corsair MP300, rated at 1600MB/s). Also back then, gen4 nvme had bios issues with X570. Other factors like SSD age/health and trim weren't disclosed. So it's a flawed test, or at the very least, incomplete.

If you use them long enough on a daily basis, you can even tell the difference between various gen3 drives.
tks brothers! @rsaeon, @Fenix , @psyph3r , tks for the inputs, i take your words of experience, i am able to comprehend the difference between 500MB/sec and 3500MB/sec, as u say the performance will be different with a different kinda loads, how come it is different with Games? or other Multimedia applications for that matter? coz the difference is minimal comparatively as u all know, we can't write off faster SSDs, but are we falling for marketing Hypes???? 3500MB/sec in a NVMe is it stable??? or consistent??? lets not hold flags for a marketing hype!! coz new technologies are a amazing thing, but for sustained innovations, hypes are necessary for the flow of funds, IMHO, i will be more than happy if u can post some believable links of the performance comparisons. thanks a lot.
 
With games it depends on how the engine loads its resources, it has to read the drive, decompress and load the textures all of which require time and depends on the other components of the system, not just the storage.
Since games until now have been optimized for HDD's, the engines aren't programmed to fetch resources from multiple locations because not everyone owns an SSD and if they go that route, the performance for the people with HDDs will be abysmal, hence you see minimal to no difference once you cross 500MB/s speeds.
This will change in a few years since next gen consoles come with SSDs, it's already evident with games like Ratchet & Clank: Rift Apart and Spiderman: Miles Morales.


 
3500MB/sec in a NVMe is it stable??? or consistent???

Well, there's a lot of factors to consider, as well as the fact that SSD testing can get incredibly complex. But yes, there are gen3 drives that can consistently benchmark at 3500MB/s sequential read speeds. And when you raid them together, you'll get linear increases in speeds: https://techenclave.com/threads/the-everyday-show-off-thread.57621/post-2274431

Write speeds are more bursty in nature - You'll get super high write speeds until the built-in cache is saturated, after which it drops down to the actual write speed of the memory chips. For example, the 970 EVO can read at 3500MB/s without any issues, and write at 2500MB/s until the cache is full, then it drops down to 1000MB/s. In day to day use, it's unlikely you'll fill the write cache, so you'll see the higher of the two speeds.

In that link above, I have three 1TB drives in a raid 0 type configuration, each drive has a 150GB write cache. So with three of them, I can write 450GB continuously, at full speed, without the drives dropping down to the actual write speed of their flash (around 400MB/s per drive for that particular ssd).

i will be more than happy if u can post some believable links of the performance comparisons.

It's difficult to convey how the numbers translate to real world experience, I find that it's something you need to experience firsthand. I don't fully understand SSD benchmarks (especially queue depths), all I know for sure is that I saw an improvement in my workflow (which is very heavy compared to a regular user) when I moved from a sata ssd to an older gen3 drive to a newer gen3 drive to a top end gen3 drive. Each upgrade brought with it a noticeable difference.

If and when I start gaming again, I'll probably get a larger sata ssd (or slower nvme ssd since they're about the same price-wise these days) for game storage and a smaller faster nvme for a system drive.

It's worth saying that on my main personal computer, I have a simple sata ssd because that's all it needs for me to write this post and check my email.
 
Last edited:
If and when I start gaming again, I'll probably get a larger sata ssd (or slower nvme ssd since they're about the same price-wise these days) for game storage and a smaller faster nvme for a system drive.
It is likely that coming games will benefit from having an nvme since there's also tech like directstorage coming up, which can take advantage of nvme SSDs (both gen3 and gen4) to actually become a significant differentiator for experience compared to sata SSDs.
 
tks brothers! @rsaeon, @Fenix , @psyph3r , tks for the inputs, i take your words of experience, i am able to comprehend the difference between 500MB/sec and 3500MB/sec, as u say the performance will be different with a different kinda loads, how come it is different with Games? or other Multimedia applications for that matter? coz the difference is minimal comparatively as u all know, we can't write off faster SSDs, but are we falling for marketing Hypes???? 3500MB/sec in a NVMe is it stable??? or consistent??? lets not hold flags for a marketing hype!! coz new technologies are a amazing thing, but for sustained innovations, hypes are necessary for the flow of funds, IMHO, i will be more than happy if u can post some believable links of the performance comparisons. thanks a lot.
Game loads consist of loading a very large number of small files due to which the peak sequential read/write benefit of NVMe drives does not kick in. In other cases, the loading gets bottlenecked by the CPU due to the sheer number of resources it has to process.

You will see better performance for NVMe drives for non-sequential, parallel workloads which mainly occur in various productivity applications. You will probably see the biggest difference between NVMe and SATA drives for operations with high I/O like video editing and definitely in database operations. However, this too will last till the DRAM or SLC cache is saturated which usually occurs after tens of GBs for the better drives and in a few GBs for cheaper drives without DRAM and limited SLC cache.

If you are looking for benchmarks, you can compare the 4KQ1T1 numbers since that captures how fast the drive randomly reads and writes small files. You will find that the most expensive drives are not much faster in that regard.
 
LTT is a retard when he makes videos about storage. my guess is that he doesn’t understand storage technology enough before making his conclusions. His storage server dying on him and losing all his videos because of the way he built it is the most cringe worthy video i’ve seen. /end rant

now coming to ssd vs nvme us pcie… comparing them is like apples and oranges.

if we draw a path from cpu to storage it would look like this.
Sata: cpu->pcie protocol->scsi protocol->sata protocol->ssd
nvme with pciegen3: cpu->pcie protocol->nvme protocol->ssd
nvme with pcie gen4: same as gen 3 but with gen4.

when i say protocol it is both combination of packet structure as well as physical components. they are closely coupled.

now in ssds there are 3d crosspoint, qld etc technologies that make them fast or slow. also depends on ssd controllers.

sata/scsi protocol adds a lot of software overhead when the drives themselves started getting faster. so nvme protocol came along. nvme added things like 64k queue depth etc because read writes are not sequential anymore.

i could go on but just wanted to say that video is dumb and designed to make viewers dumb and generate likes based on emotion.
 
LTT is a retard when he makes videos about storage. my guess is that he doesn’t understand storage technology enough before making his conclusions. His storage server dying on him and losing all his videos because of the way he built it is the most cringe worthy video i’ve seen. /end rant

now coming to ssd vs nvme us pcie… comparing them is like apples and oranges.

if we draw a path from cpu to storage it would look like this.
Sata: cpu->pcie protocol->scsi protocol->sata protocol->ssd
nvme with pciegen3: cpu->pcie protocol->nvme protocol->ssd
nvme with pcie gen4: same as gen 3 but with gen4.

when i say protocol it is both combination of packet structure as well as physical components. they are closely coupled.

now in ssds there are 3d crosspoint, qld etc technologies that make them fast or slow. also depends on ssd controllers.

sata/scsi protocol adds a lot of software overhead when the drives themselves started getting faster. so nvme protocol came along. nvme added things like 64k queue depth etc because read writes are not sequential anymore.

i could go on but just wanted to say that video is dumb and designed to make viewers dumb and generate likes based on emotion.
LTT is pretty clueless about a lot of stuff but at the same time the videos are geared towards those who take them at face value. They had completely got a video wrong on how a camera sensor captures light and when it was pointed out to them logically, they simply started ranting and doubling down on the wrong side of physics.
 
LTT is a retard when he makes videos about storage. my guess is that he doesn’t understand storage technology enough before making his conclusions. His storage server dying on him and losing all his videos because of the way he built it is the most cringe worthy video i’ve seen. /end rant

now coming to ssd vs nvme us pcie… comparing them is like apples and oranges.

if we draw a path from cpu to storage it would look like this.
Sata: cpu->pcie protocol->scsi protocol->sata protocol->ssd
nvme with pciegen3: cpu->pcie protocol->nvme protocol->ssd
nvme with pcie gen4: same as gen 3 but with gen4.

when i say protocol it is both combination of packet structure as well as physical components. they are closely coupled.

now in ssds there are 3d crosspoint, qld etc technologies that make them fast or slow. also depends on ssd controllers.

sata/scsi protocol adds a lot of software overhead when the drives themselves started getting faster. so nvme protocol came along. nvme added things like 64k queue depth etc because read writes are not sequential anymore.

i could go on but just wanted to say that video is dumb and designed to make viewers dumb and generate likes based on emotion.
i have one doubt, that is there are 3 flavours of MOBO in Ryzen 3 that is 520, 550 and 570, it is understandable that x570 is the advanced of the lot and it is perfectly PURE GEN 4 between CPU and CHIPSET and other connections within the board, 520 is purely GEN3, SOO, when u come to 550 it is a combo of gen 4 and GEN 3, the communication between CPU -> NVMe is gen 4 and CPU->PCIe 1st slot is GEN 4, is it true??? but the connection between CPU-> and other chipset features are Gen 3, I am unable to fathom how come the CPU will communicate in both GEN 4 and GEN 3, OR Is it a marketing hype ??? or will u suggest people to go for still B550 boards ???
 
Game loads consist of loading a very large number of small files due to which the peak sequential read/write benefit of NVMe drives does not kick in. In other cases, the loading gets bottlenecked by the CPU due to the sheer number of resources it has to process.

You will see better performance for NVMe drives for non-sequential, parallel workloads which mainly occur in various productivity applications. You will probably see the biggest difference between NVMe and SATA drives for operations with high I/O like video editing and definitely in database operations. However, this too will last till the DRAM or SLC cache is saturated which usually occurs after tens of GBs for the better drives and in a few GBs for cheaper drives without DRAM and limited SLC cache.

If you are looking for benchmarks, you can compare the 4KQ1T1 numbers since that captures how fast the drive randomly reads and writes small files. You will find that the most expensive drives are not much faster in that regard.
Samsung 980 is comparatively value for PRICE than samsung 970 evo plus, But the caveat in 980 is , itsa DRAM LESS NVMe, will u suggest to go for such SSDs, coz 980's SLC Cache is 160GB for a 1TB drive. Now is that good for Productivity applications running multiple VMs like say 5, or will u suggest to go for 970 evo plus with DRAM but SLC Cache is only 45 GB max. Please comprehend the difference and which one will u suggest.

Mind you there is another GEN 4 SSD in the ciruculation that is from CORSAIR MP600 CORE. so your opinion please
 
i have one doubt, that is there are 3 flavours of MOBO in Ryzen 3 that is 520, 550 and 570, it is understandable that x570 is the advanced of the lot and it is perfectly PURE GEN 4 between CPU and CHIPSET and other connections within the board, 520 is purely GEN3, SOO, when u come to 550 it is a combo of gen 4 and GEN 3, the communication between CPU -> NVMe is gen 4 and CPU->PCIe 1st slot is GEN 4, is it true??? but the connection between CPU-> and other chipset features are Gen 3, I am unable to fathom how come the CPU will communicate in both GEN 4 and GEN 3, OR Is it a marketing hype ??? or will u suggest people to go for still B550 boards ???
amd page says b550 has 20x pcie gen 4 and 10x gen 3 lanes. this means a motherboard can implement pcie16x for graphics card and 4x for nvme in pcie gen4. you have check that with exact motherboard model.
 
amd page says b550 has 20x pcie gen 4 and 10x gen 3 lanes. this means a motherboard can implement pcie16x for graphics card and 4x for nvme in pcie gen4. you have check that with exact motherboard model.
tks for your input! My question is even 550 offers primary gpu port and a nvme for pcie 4, if the cpu operates in gen 3 with other parts, is that still any good, it has a bottleneck in connecting to other peripherals. right!!!
To make it clear, is it possible if i use gen 4 compliant GPU and a NVME in the primary slots will a B550 imitate x570 when i dont have any other peripherals..
 
I am unable to fathom how come the CPU will communicate in both GEN 4 and GEN 3,

It's designed that way. PCIe is designed to be backwards compatible/switchable. You could even have your gen4 slots operate at gen1 speeds if you have older hardware, this is selectable in the bios.

The B550 chipset doesn't offer any gen4 lanes to any of the devices connected to it so there's no need for a gen4 downlink (unlike X570), so it's a gen3 link instead. The processor itself has a gen4 link meant for the chipset, and it is used as a gen4 link in X570, but for B550 that link is configured as a gen3 link:

EXaq9HTXYAA1moJ.jpg


Also as a point of interest, the Aorus B550 Master is the only AM4 motherboard that has three m.2 slots (4x gen4 each) connected to the processor, and in this mode the 1st pcie slot operates at 8x gen4. It's perfect for high speed storage oriented builds.
 
It's designed that way. PCIe is designed to be backwards compatible/switchable. You could even have your gen4 slots operate at gen1 speeds if you have older hardware, this is selectable in the bios.

The B550 chipset doesn't offer any gen4 lanes to any of the devices connected to it so there's no need for a gen4 downlink (unlike X570), so it's a gen3 link instead. The processor itself has a gen4 link meant for the chipset, and it is used as a gen4 link in X570, but for B550 that link is configured as a gen3 link:

View attachment 109519

Also as a point of interest, the Aorus B550 Master is the only AM4 motherboard that has three m.2 slots (4x gen4 each) connected to the processor, and in this mode the 1st pcie slot operates at 8x gen4. It's perfect for high speed storage oriented builds.
ok my point is if all the communication should either be in Gen 3 or GEN 4, even though the CPU has the design , i find that regardless of having 1 pcie GPU and 1 Nvme operating in GEN 4 will not perform the expected GEN 4 speed as the CPU bottlenecks with GEN 3 with other board components, it is still GEN 3 only!!! what gain will u have only part of the board performing in GEN 4 speed when the CPU is connected with GEN 3 SPEED with everyother component on board.
correct me if i am wrong!!
 
So it appears that your understanding of interconnects is incomplete. Mine isn't so great either but I'll try to explain.

Think of a processor as a headmaster of a school, and the various links to it as meetings he or she has scheduled for the day. The most important meeting would be with the board of education, and then maybe some faculty members, then perhaps some custodial staff and at the very bottom of importance is you, being suspended because you didn't tuck in your shirt. The headmaster is able to meet with all of these different people, and give each person the importance that is needed. With important matters like with the board of education, the meeting is serious and prolonged. With delinquent students, it's a few seconds of yelling and then sending them home with a reprimanding letter.

So the interconnects with the processor operate in an asynchronous manner. The processor's operating speed doesn't slow down to communicate with a gen3 device, it continues to operate at it's full speed while the gen3 devices send their data, and then works with that data after it has been transferred in full. This handling of different data speeds, from the high internal frequency of the processor to the lower frequency of the interconnect, is by a part of the processor called a buffer.

In computers particularly, the analogy for 'bottleneck' is incorrectly applied to a lot of things. For a bottleneck to occur, there needs to be a constriction of flow, between two unconstricted points. There's nothing faster than gen3 at the other end of the chipset, there's just sata ports and usb. If you had a GPU attached to pcie lanes coming from the chipset, then maybe you could say the gen3 link is bottlenecking the performance of the gpu. Or in some cases, an nvme ssd that is attached to the chipset.

And that flow, must be critical to a particular operation for a bottleneck to occur. So there's no bottleneck because there's nothing that's connected to the chipset that's important enough to slow down the processor or keep it waiting.

what gain will u have only part of the board performing in GEN 4 speed when the CPU is connected with GEN 3 SPEED with everyother component on board.

What other components are there? If you have storage and graphics connected directly to the processor with a gen4 link, that's everything you need to have a fast and responsive system that benefits from the gen4 upgrade. You wouldn't even need the chipset if you have your peripherals attached directly to the CPU's usb ports.

And as I mentioned above, there's nothing important that is connected through the chipset that it would suffer from being a gen3 link. The gen3 link to the chipset is 32GT/s which can be translated to 32Gbps for the purposes of this explanation. What device can we have on the chipset that operates at that speed? USB gen3 2x2 tops out at 20Gbp/s. How likely is it that you'll have two of those usb controllers, operating simultaneously, at full speed?

Actual real-world performance gains is a completely different topic. It's generally accepted that there isn't much use case for gen4 today for the average user. This has been true for the early days of every generation of PCIe, except maybe gen1. But introducing gen4 to the mainstream market now gives the technology and time to mature until it actually is needed. For power users though, it's really easy to saturate a 16x gen4 link with a few ssds installed in a pcie adapter card.
 
Back
Top