• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

very slow nvme speed compared to my friend who has the same drive

Joined
Dec 12, 2012
Messages
711 (0.17/day)
Location
Poland
System Name THU
Processor Intel Core i5-13600KF
Motherboard ASUS PRIME Z790-P D4
Cooling SilentiumPC Fortis 3 v2 + Arctic Cooling MX-2
Memory Crucial Ballistix 2x16 GB DDR4-3600 CL16 (dual rank)
Video Card(s) MSI GeForce RTX 4070 Ventus 3X OC 12 GB GDDR6X (2610/21000 @ 0.91 V)
Storage Lexar NM790 2 TB + Corsair MP510 960 GB + PNY XLR8 CS3030 500 GB + Toshiba E300 3 TB
Display(s) LG OLED C8 55" + ASUS VP229Q
Case Fractal Design Define R6
Audio Device(s) Yamaha RX-V381 + Monitor Audio Bronze 6 + Bronze FX | FiiO E10K-TC + Sony MDR-7506
Power Supply Corsair RM650
Mouse Logitech M705 Marathon
Keyboard Corsair K55 RGB PRO
Software Windows 10 Home
Benchmark Scores Benchmarks in 2024?
But you find, open and close the files on an HDD too, right? If this is a file system problem, why are my HDDs faster than my SSD when copying this particular folder with thousands of small files?

What makes the SSDs worse than HDDs at handling small files?
 

Mussels

Freshwater Moderator
Staff member
Joined
Oct 6, 2004
Messages
58,413 (8.21/day)
Location
Oystralia
System Name Rainbow Sparkles (Power efficient, <350W gaming load)
Processor Ryzen R7 5800x3D (Undervolted, 4.45GHz all core)
Motherboard Asus x570-F (BIOS Modded)
Cooling Alphacool Apex UV - Alphacool Eisblock XPX Aurora + EK Quantum ARGB 3090 w/ active backplate
Memory 2x32GB DDR4 3600 Corsair Vengeance RGB @3866 C18-22-22-22-42 TRFC704 (1.4V Hynix MJR - SoC 1.15V)
Video Card(s) Galax RTX 3090 SG 24GB: Underclocked to 1700Mhz 0.750v (375W down to 250W))
Storage 2TB WD SN850 NVME + 1TB Sasmsung 970 Pro NVME + 1TB Intel 6000P NVME USB 3.2
Display(s) Phillips 32 32M1N5800A (4k144), LG 32" (4K60) | Gigabyte G32QC (2k165) | Phillips 328m6fjrmb (2K144)
Case Fractal Design R6
Audio Device(s) Logitech G560 | Corsair Void pro RGB |Blue Yeti mic
Power Supply Fractal Ion+ 2 860W (Platinum) (This thing is God-tier. Silent and TINY)
Mouse Logitech G Pro wireless + Steelseries Prisma XL
Keyboard Razer Huntsman TE ( Sexy white keycaps)
VR HMD Oculus Rift S + Quest 2
Software Windows 11 pro x64 (Yes, it's genuinely a good OS) OpenRGB - ditch the branded bloatware!
Benchmark Scores Nyooom.
But you find, open and close the files on an HDD too, right? If this is a file system problem, why are my HDDs faster than my SSD when copying this particular folder with thousands of small files?

What makes the SSDs worse than HDDs at handling small files?
Running out of cache. your SSD is not full performance across the entire SSD.

You'll have to research your SSD on its own, and find out if its SLC, TLC, MLC, QLC, if it has DRAM cache or not, how full it is, and so on

The screenshots you posted show the SSD at roughly twice the speed of the mech drive, btw
 
Joined
Dec 12, 2012
Messages
711 (0.17/day)
Location
Poland
System Name THU
Processor Intel Core i5-13600KF
Motherboard ASUS PRIME Z790-P D4
Cooling SilentiumPC Fortis 3 v2 + Arctic Cooling MX-2
Memory Crucial Ballistix 2x16 GB DDR4-3600 CL16 (dual rank)
Video Card(s) MSI GeForce RTX 4070 Ventus 3X OC 12 GB GDDR6X (2610/21000 @ 0.91 V)
Storage Lexar NM790 2 TB + Corsair MP510 960 GB + PNY XLR8 CS3030 500 GB + Toshiba E300 3 TB
Display(s) LG OLED C8 55" + ASUS VP229Q
Case Fractal Design Define R6
Audio Device(s) Yamaha RX-V381 + Monitor Audio Bronze 6 + Bronze FX | FiiO E10K-TC + Sony MDR-7506
Power Supply Corsair RM650
Mouse Logitech M705 Marathon
Keyboard Corsair K55 RGB PRO
Software Windows 10 Home
Benchmark Scores Benchmarks in 2024?
The screenshots you posted show the SSD at roughly twice the speed of the mech drive, btw

Either you are not seeing what I am posting, or you do not understand it.

5-10 MB/s is not twice as fast as ~65 MB/s. And 20-50 MB/s is not twice as fast either. It is SLOWER.

These are the transfer speeds I was seeing for most of the time. The 104 MB/s number is not the average, it was the current speed at the time of the screenshot.
The copy process from the SSD to HDD took longer than from HDD to HDD.

The HDD NEVER dropped below 50 MB/s, yet the SSD did MOST of the time.


Running out of cache, drive being full, those are all general things that explain why SSDs are not always at their full speeds. But it does not explain why an SSD can be much slower than an HDD at a specific workload with lots of small files.
I want to learn why an SSD reads a tiny file at 5 MB/s, while an HDD reads the same file at 50 MB/s.
 
Joined
May 2, 2017
Messages
7,762 (3.08/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
Either you are not seeing what I am posting, or you do not understand it.

5-10 MB/s is not twice as fast as ~65 MB/s. And 20-50 MB/s is not twice as fast either. It is SLOWER.

These are the transfer speeds I was seeing for most of the time. The 104 MB/s number is not the average, it was the current speed at the time of the screenshot.
The copy process from the SSD to HDD took longer than from HDD to HDD.

The HDD NEVER dropped below 50 MB/s, yet the SSD did MOST of the time.


Running out of cache, drive being full, those are all general things that explain why SSDs are not always at their full speeds. But it does not explain why an SSD can be much slower than an HDD at a specific workload with lots of small files.
I want to learn why an SSD reads a tiny file at 5 MB/s, while an HDD reads the same file at 50 MB/s.
It might be as simple as Windows caching a lot of the copied data - that happens when you repeatedly copy or move the same data, windows' copier keeps the data cached for a short while which can artificially inflate performance - though that typically means higher performance than this, and in shrorter bursts.

It might also be a bad NVMe driver, iffy firmware on your drive (I see reviews of it point out "poor application performance", which indicates a firmware tuned for benchmarks and sequential transfers, not real-world use, though how big those detrimental effects are can vary a lot), or a bunch of other factors. It's still pretty weird though.
 
Joined
Dec 12, 2012
Messages
711 (0.17/day)
Location
Poland
System Name THU
Processor Intel Core i5-13600KF
Motherboard ASUS PRIME Z790-P D4
Cooling SilentiumPC Fortis 3 v2 + Arctic Cooling MX-2
Memory Crucial Ballistix 2x16 GB DDR4-3600 CL16 (dual rank)
Video Card(s) MSI GeForce RTX 4070 Ventus 3X OC 12 GB GDDR6X (2610/21000 @ 0.91 V)
Storage Lexar NM790 2 TB + Corsair MP510 960 GB + PNY XLR8 CS3030 500 GB + Toshiba E300 3 TB
Display(s) LG OLED C8 55" + ASUS VP229Q
Case Fractal Design Define R6
Audio Device(s) Yamaha RX-V381 + Monitor Audio Bronze 6 + Bronze FX | FiiO E10K-TC + Sony MDR-7506
Power Supply Corsair RM650
Mouse Logitech M705 Marathon
Keyboard Corsair K55 RGB PRO
Software Windows 10 Home
Benchmark Scores Benchmarks in 2024?
I guess it is a mystery, then.

I feel like random reads/writes should be the main focus of engineers, both hardware and software. SSD progress is basically dead when it comes to that aspect. What is the point of introducing more bandwidth (Gen4) at almost double the price, when those drives have the same garbage random access performance as slower and cheaper drives.


By the way, what does the 4K-64Thrd benchmark represent? Accessing 64 different random 4K blocks at the same time? So in theory software (including Windows) could be programmed to access data in that way? Is that what is needed to get the max out of SSDs?
 
Joined
May 2, 2017
Messages
7,762 (3.08/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
I guess it is a mystery, then.

I feel like random reads/writes should be the main focus of engineers, both hardware and software. SSD progress is basically dead when it comes to that aspect. What is the point of introducing more bandwidth (Gen4) at almost double the price, when those drives have the same garbage random access performance as slower and cheaper drives.
It should be, but it doesn't look good on marketing slides, so it rarely is. Also, increasing random performance is a lot harder than increasing sequential performance, both due to the aforementioned OS bottlenecks as well as how flash memory functions physically. That's also why good SSD testing relies on real-world applications rather than canned benchmarks, as those benchmarks have little to do with real-world use and drives are often tuned to perform well in them (their access patterns and performance requirements are well known and relatively simple).
By the way, what does the 4K-64Thrd benchmark represent? Accessing 64 different random 4K blocks at the same time? So in theory software (including Windows) could be programmed to access data in that way? Is that what is needed to get the max out of SSDs?
Both increasing thread counts and queue depths are ways of increasing the amount of work an SSD is asked to do simultaneously/in rapid succession, which increases performance in random workloads. After all, if a single thread is asking for a single 4k read at a time, especially with processing or other stuff in between but even just waiting for that read to finish and the data to arrive before asking for another, the SSD is sitting idle for the vast majority of the time. The problem is that you can't just change how software is programmed - while programming is full of hacks and poorly optimized code, at the end of the day any application only needs the data it needs, and only needs it when it's needed. Making your app capable of accessing data with tons of threads doesn't matter if none of those threads are actually needed for accessing data.

Of course, this quickly gets complicated: games, for example, have historically been developed with an expectation of HDD storage, as supposing an SSD (for example, near-instant seek times) can lead anyone using an HDD to experience severe performance issues - stuttering, data failing to load, major pop-in, etc. So a lot of things are developed for a lowest common denominator of performance. This also means stuff like how data is packaged together in files (things likely to be used together are put together to avoid random reads on HDDs; a lot is done to avoid fragmentation, etc.) is sub-optimal for SSD use. If the data structures, software behavior, and access patterns of these applications were tuned with the expectation of an SSD instead, this could lead to significant performance improvements. But AFAIK we're still not quite at the point where the industry has shifted over - SSD storage as a system requirement has started showing up, but it's not ubiquitous yet.
 
Joined
Feb 9, 2020
Messages
401 (0.27/day)
Location
Panama City Beach, Florida
System Name EventHorizon
Processor Intel® Core™ Processor i9-13900KF 8P/16 + 16E 3.00GHz [Turbo 5.7GHz] 36MB Cache LGA1700
Motherboard ASUS PRIME Z790-P
Cooling CyberpowerPC MasterLiquid Lite 240mm ARGB CPU Liquid Cooler
Memory 32GB (16GBx2) DDR5/6000MHz Dual Channel Memory (KINGSTON FURY BEAST RGB)
Video Card(s) GeForce RTX™ 4080 16GB
Storage 2TB WD BLACK SN850X (PCIe Gen4) NVMe M.2 SSD - Seq R/W: Up to 7300/6600 MB/s, Rnd R/W up to 1200/110
Display(s) LG 34''
Case CyberPowerPC HYTE Y60 Dual Chamber Mid-Tower Gaming Case w/ Panoramic View Tempered Glass + 2x120mm
Audio Device(s) Asus Strix w/Alan Finote mod for Windows 11
Power Supply High Power 1300W 80+ GOLD Full Modular w/ PCIE 12+4Pins Connector for PCIe 5.0 graphics cards
Mouse Steelseries Rival 600 wired
Keyboard Steelseries Apex 7 TKL red Switch
Software Win 11 Pro
Joined
Dec 12, 2012
Messages
711 (0.17/day)
Location
Poland
System Name THU
Processor Intel Core i5-13600KF
Motherboard ASUS PRIME Z790-P D4
Cooling SilentiumPC Fortis 3 v2 + Arctic Cooling MX-2
Memory Crucial Ballistix 2x16 GB DDR4-3600 CL16 (dual rank)
Video Card(s) MSI GeForce RTX 4070 Ventus 3X OC 12 GB GDDR6X (2610/21000 @ 0.91 V)
Storage Lexar NM790 2 TB + Corsair MP510 960 GB + PNY XLR8 CS3030 500 GB + Toshiba E300 3 TB
Display(s) LG OLED C8 55" + ASUS VP229Q
Case Fractal Design Define R6
Audio Device(s) Yamaha RX-V381 + Monitor Audio Bronze 6 + Bronze FX | FiiO E10K-TC + Sony MDR-7506
Power Supply Corsair RM650
Mouse Logitech M705 Marathon
Keyboard Corsair K55 RGB PRO
Software Windows 10 Home
Benchmark Scores Benchmarks in 2024?
Both increasing thread counts and queue depths are ways of increasing the amount of work an SSD is asked to do simultaneously/in rapid succession, which increases performance in random workloads. After all, if a single thread is asking for a single 4k read at a time, especially with processing or other stuff in between but even just waiting for that read to finish and the data to arrive before asking for another, the SSD is sitting idle for the vast majority of the time. The problem is that you can't just change how software is programmed - while programming is full of hacks and poorly optimized code, at the end of the day any application only needs the data it needs, and only needs it when it's needed. Making your app capable of accessing data with tons of threads doesn't matter if none of those threads are actually needed for accessing data.
But would it not be possible to program Windows to copy multiple files at a time when copying a folder? Would that not increase performance drastically?

Copying them one at a time seems like a leftover from the HDD era.
 
Joined
May 2, 2017
Messages
7,762 (3.08/day)
Location
Back in Norway
System Name Hotbox
Processor AMD Ryzen 7 5800X, 110/95/110, PBO +150Mhz, CO -7,-7,-20(x6),
Motherboard ASRock Phantom Gaming B550 ITX/ax
Cooling LOBO + Laing DDC 1T Plus PWM + Corsair XR5 280mm + 2x Arctic P14
Memory 32GB G.Skill FlareX 3200c14 @3800c15
Video Card(s) PowerColor Radeon 6900XT Liquid Devil Ultimate, UC@2250MHz max @~200W
Storage 2TB Adata SX8200 Pro
Display(s) Dell U2711 main, AOC 24P2C secondary
Case SSUPD Meshlicious
Audio Device(s) Optoma Nuforce μDAC 3
Power Supply Corsair SF750 Platinum
Mouse Logitech G603
Keyboard Keychron K3/Cooler Master MasterKeys Pro M w/DSA profile caps
Software Windows 10 Pro
But would it not be possible to program Windows to copy multiple files at a time when copying a folder? Would that not increase performance drastically?

Copying them one at a time seems like a leftover from the HDD era.
The windows file copy application is notoriously terrible in terms of performance - it's made to be simple to use and preserve file integrity, and performance is quite low on its list of priorities. Try something like TeraCopy, it can be quite a lot faster.

There's also something to be said for a copy operation not bogging down your system entirely with interrupts and CPU load from tons of concurrent operations. Settings some hard limits on resource usage can be a pretty good idea still.
 
Joined
Nov 25, 2021
Messages
34 (0.04/day)
System Name Poligon
Processor Intel Alder Lake i7 12700K
Motherboard Asus Maximus Hero Z690
Cooling Custom Water Loop Bykski CPU+GPU; 360 + 240 radiators
Memory 2X16 GB G.Skill DDR5 Z5 6000 CL36
Video Card(s) Nvidia GeForce RTX 3090 FE
Storage Western Digital NvMe SN850 black 2Tb; 2X2Tb SanDisk Ultra ssd
Display(s) Philips Momentum 436M6VBPAB 43" 4K HDR
Case Silverstone FT05
Audio Device(s) Presonus Eris e3.5
Power Supply Cooler Master V1000 Platinum
Mouse Logitech MX Master 3; Logitech G502 HERO
Keyboard Steelseries Apex Pro
Software Windows 11

Mussels

Freshwater Moderator
Staff member
Joined
Oct 6, 2004
Messages
58,413 (8.21/day)
Location
Oystralia
System Name Rainbow Sparkles (Power efficient, <350W gaming load)
Processor Ryzen R7 5800x3D (Undervolted, 4.45GHz all core)
Motherboard Asus x570-F (BIOS Modded)
Cooling Alphacool Apex UV - Alphacool Eisblock XPX Aurora + EK Quantum ARGB 3090 w/ active backplate
Memory 2x32GB DDR4 3600 Corsair Vengeance RGB @3866 C18-22-22-22-42 TRFC704 (1.4V Hynix MJR - SoC 1.15V)
Video Card(s) Galax RTX 3090 SG 24GB: Underclocked to 1700Mhz 0.750v (375W down to 250W))
Storage 2TB WD SN850 NVME + 1TB Sasmsung 970 Pro NVME + 1TB Intel 6000P NVME USB 3.2
Display(s) Phillips 32 32M1N5800A (4k144), LG 32" (4K60) | Gigabyte G32QC (2k165) | Phillips 328m6fjrmb (2K144)
Case Fractal Design R6
Audio Device(s) Logitech G560 | Corsair Void pro RGB |Blue Yeti mic
Power Supply Fractal Ion+ 2 860W (Platinum) (This thing is God-tier. Silent and TINY)
Mouse Logitech G Pro wireless + Steelseries Prisma XL
Keyboard Razer Huntsman TE ( Sexy white keycaps)
VR HMD Oculus Rift S + Quest 2
Software Windows 11 pro x64 (Yes, it's genuinely a good OS) OpenRGB - ditch the branded bloatware!
Benchmark Scores Nyooom.
Either you are not seeing what I am posting, or you do not understand it.

5-10 MB/s is not twice as fast as ~65 MB/s. And 20-50 MB/s is not twice as fast either. It is SLOWER.

These are the transfer speeds I was seeing for most of the time. The 104 MB/s number is not the average, it was the current speed at the time of the screenshot.
The copy process from the SSD to HDD took longer than from HDD to HDD.

The HDD NEVER dropped below 50 MB/s, yet the SSD did MOST of the time.


Running out of cache, drive being full, those are all general things that explain why SSDs are not always at their full speeds. But it does not explain why an SSD can be much slower than an HDD at a specific workload with lots of small files.
I want to learn why an SSD reads a tiny file at 5 MB/s, while an HDD reads the same file at 50 MB/s.
You do realise thats a meaningless way to look at the data, as the SSD can read/write faster so it's going to idle the rest of the time waiting on the other end to catch up?
"Microsoft time" has been around since... well, microsoft. Those are estimates and they're often wrong, you can pause a file transfer and hit resume and see zero activity for 5 minutes and then BRRRRRRRRRR it's done at impossible speeds.

You're using an unreliable source of information in an incorrect way, to agree with a pre-deteremined view.

I guess it is a mystery, then.

I feel like random reads/writes should be the main focus of engineers, both hardware and software. SSD progress is basically dead when it comes to that aspect. What is the point of introducing more bandwidth (Gen4) at almost double the price, when those drives have the same garbage random access performance as slower and cheaper drives.


By the way, what does the 4K-64Thrd benchmark represent? Accessing 64 different random 4K blocks at the same time? So in theory software (including Windows) could be programmed to access data in that way? Is that what is needed to get the max out of SSDs?
Yes, it's making 64 reqeusts at at time


The problem is you cant just tell an SSD to do 64 separate writes, unless you want it to write to 64 separate locations and wear out 64 times faster - the data gets sent, and the drive has to decide based on its internal firmware how to write them as fast as possible, without burning out the drive like early flash memory did
 
Joined
Dec 26, 2006
Messages
3,472 (0.55/day)
Location
Northern Ontario Canada
Processor Ryzen 5700x
Motherboard Gigabyte X570S Aero G R1.1 BiosF5g
Cooling Noctua NH-C12P SE14 w/ NF-A15 HS-PWM Fan 1500rpm
Memory Micron DDR4-3200 2x32GB D.S. D.R. (CT2K32G4DFD832A)
Video Card(s) AMD RX 6800 - Asus Tuf
Storage Kingston KC3000 1TB & 2TB & 4TB Corsair LPX
Display(s) LG 27UL550-W (27" 4k)
Case Be Quiet Pure Base 600 (no window)
Audio Device(s) Realtek ALC1220-VB
Power Supply SuperFlower Leadex V Gold Pro 850W ATX Ver2.52
Mouse Mionix Naos Pro
Keyboard Corsair Strafe with browns
Software W10 22H2 Pro x64
Interesting stuff. I don't think I have ever bothered transferring from drive to drive. Haven't had to.
NVMe - OS/programs
Sata SSD - Steam/games
hdd replaced to SSD - for storage - that transfer wasn't fast, but that was to be expected with all the different types and sizes of files and a regular HDD

ok, just copied windows iso from WD blue 2TB sata drive to my NVMe - held steady at 465MB/s
 
Top