Question what are the best ssd benchmark / test sites most reliable and correct?

Grealish01

Great
BANNED
Jan 22, 2022
224
2
85
I have to get a m.2 nvme pcie4 4.0 SSD and I would like to refer to some testing before purchasing. But strangely I don't find any on the internet. Can you link me some of them please?
 
Define "best"
The best benchmark is YOUR workload.
Synthetic benchmarks are only good if they reflect your workload.

To answer your question:
You can use anandtech ssd bench to compare a fair number of ssd devices in all manner of tests:
https://www.anandtech.com/bench/SSD21/3005

That said, which ssd you buy will likely not make a noticeable difference in performance in real life.
These guys could not tell what ssd was in the box:
View: https://www.youtube.com/watch?v=4DKLA7w9eeA


You may also consider reliability.
Good info on that is hard to come by.
Puget systems has an article on that:
 
  • Like
Reactions: Grealish01

Karadjgne

Titan
Ambassador
You can't dedicate storage to read only or write only. Just using windows you both read and write, accessing anything stored on a different drive will both read and write, read when you open it, write when you save it.

The only dedication you can do is specify where programs will be saved/accessed.

What you should look for in an SSD is longetivity. Larger the drive, the higher the i/o, the longer and more reliable it is. SSDs have cells that store voltage. The bigger the drive, the more cells it contains. The more cells it has, the less often each individual cell is used for storage and TRIM, longer the drive lasts in a healthy state.

Performance is different, there you'd need to read reviews as to each drive. Some drives have a capped performance, the buffer cache fills up with large file use, and slows performance to a crawl. Some don't. Doesn't affect small file transfers like games and OS etc, but massive file transfers under sustained loads can be an issue. Some drives do not do well when filled beyond 50%ish, slowing down the read/writes, some don't. Benchmarks don't always show such. Reviews will.
 
  • Like
Reactions: Grealish01

Grealish01

Great
BANNED
Jan 22, 2022
224
2
85
You can't dedicate storage to read only or write only. Just using windows you both read and write, accessing anything stored on a different drive will both read and write, read when you open it, write when you save it.

The only dedication you can do is specify where programs will be saved/accessed.

What you should look for in an SSD is longetivity. Larger the drive, the higher the i/o, the longer and more reliable it is. SSDs have cells that store voltage. The bigger the drive, the more cells it contains. The more cells it has, the less often each individual cell is used for storage and TRIM, longer the drive lasts in a healthy state.

Performance is different, there you'd need to read reviews as to each drive. Some drives have a capped performance, the buffer cache fills up with large file use, and slows performance to a crawl. Some don't. Doesn't affect small file transfers like games and OS etc, but massive file transfers under sustained loads can be an issue. Some drives do not do well when filled beyond 50%ish, slowing down the read/writes, some don't. Benchmarks don't always show such. Reviews will.
where can i see the reviews of a model?
 

USAFRet

Titan
Moderator
just googling for ssd review shows.....

https://www.pcmag.com/categories/ssds
https://www.pcworld.com/article/407542/best-ssds.html
 
  • Like
Reactions: Grealish01

Karadjgne

Titan
Ambassador
For reviews, I started out with pcpartpicker.com. What were the popular ssds ppl were buying. Then simply Googled those "xxx review" and skipped over the trash reviews from ppl buying, looked for more professional reviews. Many professional sites will compare multiple sizes, so you'll see the different results for a 500,1 and 2Tb etc as far as performance, longetivity etc. They also usually have breakdowns as to the type of nand used, the controller used and what that means exactly for the ssd, whether it's good for short burst files, long continuous use and where it stacks up against other brands/models.

Take the results with a grain of salt. While one particular model might seem better on paper, realize that you are talking about nanosecond technology, so in real world applications for many results you'd be physically unable to see any difference, you can't physically split a second down far enough to tell. An app taking 1.0004 seconds to open and an app taking 1.04 seconds to open on a different SSD are exactly the same, regardless of benchmark results.

Even just Google hdd vs ssd vs nvme will show most times that an SSD and NVMe can make little to no discernable difference to games, unless stuck in side-by-side comparison.
 
  • Like
Reactions: Grealish01

Grealish01

Great
BANNED
Jan 22, 2022
224
2
85
Per le recensioni, ho iniziato con pcpartpicker.com. Quali erano i popolari SSD che stavano acquistando ppl. Quindi ho semplicemente cercato su Google quelle "recensioni xxx" e saltato le recensioni spazzatura dall'acquisto di ppl, cercando recensioni più professionali. Molti siti professionali confronteranno più dimensioni, quindi vedrai i diversi risultati per un 500,1 e 2Tb ecc. Per quanto riguarda prestazioni, longevità ecc. Di solito hanno anche guasti per quanto riguarda il tipo di nand utilizzato, il controller utilizzato e cosa ciò significa esattamente per l'ssd, se è buono per file a raffica breve, un uso prolungato e continuo e dove si accumula rispetto ad altre marche/modelli.

Prendi i risultati con le pinze. Mentre un modello particolare potrebbe sembrare migliore sulla carta, renditi conto che stai parlando di tecnologia dei nanosecondi, quindi nelle applicazioni del mondo reale per molti risultati non saresti fisicamente in grado di vedere alcuna differenza, non puoi dividere fisicamente un secondo abbastanza da raccontare. Un'app che impiega 1,0004 secondi per aprirsi e un'app che impiega 1,04 secondi per aprirsi su un altro SSD sono esattamente le stesse, indipendentemente dai risultati del benchmark.

Anche solo Google hdd vs ssd vs nvme mostrerà la maggior parte delle volte che un SSD e NVMe possono fare poca o nessuna differenza distinguibile per i giochi, a meno che non siano bloccati in un confronto fianco a fianco.
[/CITAZIONE]
grazie il discorso nvme è chiaro. Personalmente ho notato una differenza nella programmazione quando le nand sono calde, l'ho sperimentato su un SSD m.2, ma anche le nand SATA se si scaldano aumentano le prestazioni di programmazione?
For reviews, I started out with pcpartpicker.com. What were the popular ssds ppl were buying. Then simply Googled those "xxx review" and skipped over the trash reviews from ppl buying, looked for more professional reviews. Many professional sites will compare multiple sizes, so you'll see the different results for a 500,1 and 2Tb etc as far as performance, longetivity etc. They also usually have breakdowns as to the type of nand used, the controller used and what that means exactly for the ssd, whether it's good for short burst files, long continuous use and where it stacks up against other brands/models.

Take the results with a grain of salt. While one particular model might seem better on paper, realize that you are talking about nanosecond technology, so in real world applications for many results you'd be physically unable to see any difference, you can't physically split a second down far enough to tell. An app taking 1.0004 seconds to open and an app taking 1.04 seconds to open on a different SSD are exactly the same, regardless of benchmark results.

Even just Google hdd vs ssd vs nvme will show most times that an SSD and NVMe can make little to no discernable difference to games, unless stuck in side-by-side comparison.
thanks the nvme speech is clear. Personally I have noticed a difference in programming when the nands are hot, I have experienced this on an m.2 SSD, but also the SATA nands if they heat up increase programming performance?
 

Karadjgne

Titan
Ambassador
It's a balance. Warm ssds supposedly work better, but hot do not. Samsung recommends for best performance and longetivity that ssds remain in the 30-50°C range under load. While that's often hard to gage, most software not having the resources or options to read ssd temps, it is important to realize differences. M.2 don't come with heatsinks for pcie3.0, just using a copper backed sticker as a heatsoak, while a standard 2.5" Sata ssd can simply use the entire casing as a heatsoak. There's also board placement, the m.2 slot directly next to pcie x16 slot will take more abuse in a liquid cooled cpu build than an air-cooled cpu build. Radiated gpu heat and little to no airflow. For heavy sustained usage it'd be better to use a different m.2 slot if possible or use motherboard heatsink/covers.
 
  • Like
Reactions: Grealish01

Grealish01

Great
BANNED
Jan 22, 2022
224
2
85
È un equilibrio. Gli ssd caldi presumibilmente funzionano meglio, ma quelli caldi no. Samsung consiglia per le migliori prestazioni e longevità che gli SSD rimangano nell'intervallo 30-50°C sotto carico. Sebbene sia spesso difficile da valutare, poiché la maggior parte dei software non dispone delle risorse o delle opzioni per leggere i temp ssd, è importante rendersi conto delle differenze. M.2 non viene fornito con dissipatori di calore per pcie3.0, ma utilizza solo un adesivo con retro in rame come dissipatore di calore, mentre un SSD Sata standard da 2,5" può semplicemente utilizzare l'intero involucro come dissipatore di calore. C'è anche il posizionamento della scheda, l'm.2 lo slot direttamente accanto allo slot pcie x16 subirà più abusi in una build CPU raffreddata a liquido rispetto a una build CPU raffreddata ad aria. Calore della GPU irradiato e flusso d'aria minimo o nullo. Per un uso intenso e prolungato sarebbe meglio utilizzare un m.2 diverso slot se possibile o utilizzare il dissipatore di calore/coperture della scheda madre.
so can i get an nvme? To me it would be useful to have the unit hot but my fear I repeat is that the nvme ssd would have too drastic drops in performance at temperatures that a non nvme can easily handle.
 

Grealish01

Great
BANNED
Jan 22, 2022
224
2
85
Define "best"
The best benchmark is YOUR workload.
Synthetic benchmarks are only good if they reflect your workload.

To answer your question:
You can use anandtech ssd bench to compare a fair number of ssd devices in all manner of tests:
https://www.anandtech.com/bench/SSD21/3005

That said, which ssd you buy will likely not make a noticeable difference in performance in real life.
These guys could not tell what ssd was in the box:
View: https://www.youtube.com/watch?v=4DKLA7w9eeA


You may also consider reliability.
Good info on that is hard to come by.
Puget systems has an article on that:
@geofeltro hello, sorry if I answer a little late. But Puget systems in the classification of power supplies brings only PSU EVGA and SuoerFlow, but this is because it uses only those or already from the choice of these 2 before a selection was made on their part? in the sense that for them those 2 are already the best brands