Samsung 950 pro vs 960 Evo

antok86

[H]F Junkie
Joined
Feb 26, 2006
Messages
8,252
other than warranty and some speeed differences any reason why to get a 960 EVO over the 950 PRO? The 250gb+ versions of eAch can be had for about the same price on ebay(used/new).
 

arestavo

[H]ard|Gawd
Joined
Mar 25, 2013
Messages
1,672
There isn't even a difference for game load times and Windows load times between the fastest NVME and a good quality SATA III SSD. Scroll to the bottom of the first post for professional reviews on the 950 PRO and 960 EVO (and for more confirmation of what I found).

https://hardforum.com/threads/nvme-...a-iii-ssd-game-load-time-comparisons.1911914/

Doing data base work and/or loading up ass tons of VMs? Then it might pay to get the faster NVME drive.
 

olavgg

Limp Gawd
Joined
Oct 27, 2010
Messages
232
Doing data base work and/or loading up ass tons of VMs? Then it might pay to get the faster NVME drive.

Nope,
My old Intel 320 beats the crap out of Samsung 950 Pro or 960 for database workloads.

You can easily test this with Ubuntu and PostgreSQL by running the following command:
Code:
/usr/lib/postgresql/9.5/bin/pg_test_fsync

Intel 320 gets between 2000-4000 IOPS.
Intel S3700 should get 3000-5000 IOPS.
 

arestavo

[H]ard|Gawd
Joined
Mar 25, 2013
Messages
1,672

olavgg

Limp Gawd
Joined
Oct 27, 2010
Messages
232
No I'm not trolling :)
That is why I told you how to test it.

You get similar results with fio.

Stop using broken benchmark tools.
 

arestavo

[H]ard|Gawd
Joined
Mar 25, 2013
Messages
1,672
Phoronix has tested Samsung 950 and 960.
SQLite performance is crap, because of horrible sync write performance. Just as I said.

http://www.phoronix.com/scan.php?page=article&item=samsung-960-evo&num=2

Maybe I'm reading it wrong - those results show MB/s, not IOPS as you previously listed.

EDIT: Page 4 has that same PostgreSQL pgbench (version 9.4.3), but shows TPS not IOPS. So I've no real way of converting that, do I?

Edit 2: Post up your results! From the articles end - "Those wishing to compare their own system's performance to the results in this article can simply install the Phoronix Test Suite and run phoronix-test-suite benchmark 1612151-TA-SAMSUNG9644 for a fully-automated, side-by-side comparison to the data in this article."

I don't run linux, nor do I do any database work - so I'll let others have fun with that.
 
Last edited:

KIAman

Limp Gawd
Joined
Jul 17, 2013
Messages
236
I have a 950 pro (NvMe) in my current build. My old build had an 850 evo (sata). And to be totally honest, I can't tell a damned difference in loading times for anything I usually do. SOOO... do I regret my 950 pro? no, but in hindsight, I might have gotten an bigger evo for the same price.
 

AlphaFox

Limp Gawd
Joined
Jul 26, 2016
Messages
297
I have a 950 pro (NvMe) in my current build. My old build had an 850 evo (sata). And to be totally honest, I can't tell a damned difference in loading times for anything I usually do. SOOO... do I regret my 950 pro? no, but in hindsight, I might have gotten an bigger evo for the same price.
Thats why I held off on the NVMe versions, the speed difference is not hugely noticeable in day to day operations.
 

arestavo

[H]ard|Gawd
Joined
Mar 25, 2013
Messages
1,672
I have a 950 pro (NvMe) in my current build. My old build had an 850 evo (sata). And to be totally honest, I can't tell a damned difference in loading times for anything I usually do. SOOO... do I regret my 950 pro? no, but in hindsight, I might have gotten an bigger evo for the same price.
Thats why I held off on the NVMe versions, the speed difference is not hugely noticeable in day to day operations.

Game load times and Windows boot times are identical to a good SATA III SSD. Found that out for myself after getting a 512GB 950 Pro and an expensive PCIE adapter with passive heatsink, and wished I had just stayed with my 850 EVO.
 

AlphaFox

Limp Gawd
Joined
Jul 26, 2016
Messages
297
Game load times and Windows boot times are identical to a good SATA III SSD. Found that out for myself after getting a 512GB 950 Pro and an expensive PCIE adapter with passive heatsink, and wished I had just stayed with my 850 EVO.

hehe, exactly. SATA III is good enough for just about everything and the drives are cheaper.
I run a hackintosh and SATA is supported native and NVme needs a patch so ill just stick with SATA, one less thing to worry about.
 

olavgg

Limp Gawd
Joined
Oct 27, 2010
Messages
232
Maybe I'm reading it wrong - those results show MB/s, not IOPS as you previously listed.

EDIT: Page 4 has that same PostgreSQL pgbench (version 9.4.3), but shows TPS not IOPS. So I've no real way of converting that, do I?

Edit 2: Post up your results! From the articles end - "Those wishing to compare their own system's performance to the results in this article can simply install the Phoronix Test Suite and run phoronix-test-suite benchmark 1612151-TA-SAMSUNG9644 for a fully-automated, side-by-side comparison to the data in this article."

I don't run linux, nor do I do any database work - so I'll let others have fun with that.

pgbench tests read performance. pg_test_fsync will test sync write performance
 

Eric1285

[H]ard|Gawd
Joined
Feb 10, 2003
Messages
1,516
I upgraded from an 840 EVO to a 950 Pro. I noticed a difference, especially booting. To be fair, I was running the 840 EVO on my board's crappy pseudo-SATA III controller. The 950 Pro is plugged into a PCIe slot.
 
Top