Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
The more free space left the faster performance on MBP, True or False ?
- Thread starter bo-waleed
- Start date
- Sort by reaction score
In my experience 10-15% are necessary for the machine not to slow down, but anything past that won’t just magically speed up the CPU.
From my understanding with the 2019 16" and now the current late 2021 16" the ssd's that are 2tb seem to be faster then the 1tb's from my own use of Blackmagic and Aja to measure.
I only checked after reading about it in a few different articles which always seemed to show 2tb ssd from Apple faster than the 1tb offering.
I tried with both empty drives and also with the drives having approx 350 gbs of data on drives.
This is hardly any type of scientific test but was just for me to see any differences.
Using both Intel and Apple chipsets.
And was between 4 machines and only those SSD combinations but both with 64gb memory and either the 2.4 for the intel 2019 16" and the 10/32/64 for 2021 Max chipsets.
YMMV
I only checked after reading about it in a few different articles which always seemed to show 2tb ssd from Apple faster than the 1tb offering.
I tried with both empty drives and also with the drives having approx 350 gbs of data on drives.
This is hardly any type of scientific test but was just for me to see any differences.
Using both Intel and Apple chipsets.
And was between 4 machines and only those SSD combinations but both with 64gb memory and either the 2.4 for the intel 2019 16" and the 10/32/64 for 2021 Max chipsets.
YMMV
Last edited:
Absolutely false. Storage is just where your bits are kept. Free space has very little to do with the machine's speed, which is mainly determined by the CPU.
Actually, it’s well known that SSDs tend to slow down as they fill up. While the CPU may not be directly effected, the SSD can in fact impact performance. This will largely depend on the type of work you do on the machine of course. Premium NVMe drives will of course take less of a hit then their SATA counterparts.
Lots of people say that on the internet, but I'm not convinced it's accurate. Here is a guy who ran some actual tests on various SSDs and he didn't detect any meaningful slow down.
Is full SSD slower? Is it a MYTH or a FACT? • Epic Game Tech - PC Builds, Hardware unboxing and How-to Guides
Tested 3 different SATA SSDs (Samsung 860 EVO 2TB, Crucial MX500 1TB and Team Group L3 EVO 120GB) speed with CrystalDiskMark using 64 GB and 1GB test files
epicgametech.com
Oh yes, epicgametech. Great resource. Try this articles from anandtech.
"For drives on the market today that don't already prioritize consistent IO, it is possible to deliver significant improvements in IO consistency through an increase in spare area. OCZ's Vector and Samsung's SSD 840 Pro both deliver much better IO consistency if you simply set aside 25% of the total NAND capacity as spare area."
"For drives on the market today that don't already prioritize consistent IO, it is possible to deliver significant improvements in IO consistency through an increase in spare area. OCZ's Vector and Samsung's SSD 840 Pro both deliver much better IO consistency if you simply set aside 25% of the total NAND capacity as spare area."
Oh, it's certainly accurate. Far more so with Spinning Rust (HDDs) than SSDs of course. The problem is that all OS's swap out to (slower) disk as RAM resources diminish, the faster the drive, the more seamless the experience. YEARS of experience of PCs, Macs, and all sorts of devices that handle disk swaps regularly benefit from as fast a drive as possible. Note that SSDs, as they fill, CAN get slower, as the controller is doing a lot more work housekeeping than it has to do with an empty drive.Lots of people say that on the internet, but I'm not convinced it's accurate. Here is a guy who ran some actual tests on various SSDs and he didn't detect any meaningful slow down.
Is full SSD slower? Is it a MYTH or a FACT? • Epic Game Tech - PC Builds, Hardware unboxing and How-to Guides
Tested 3 different SATA SSDs (Samsung 860 EVO 2TB, Crucial MX500 1TB and Team Group L3 EVO 120GB) speed with CrystalDiskMark using 64 GB and 1GB test files
I found this to be a pretty good layman's explanation, hopefully this helps you understand. The problem with the tests you linked above is that it was using spare SSDs doing nothing but writing benchmarks. Try running those benchmarks when you're also running the OS on the same drive or playing a game and the results would be vastly different. Below is a snippet of an article written by Aaron Mickunas at Riot Games.
When you save a file to your computer, your SSD has to do some (very) quick math to determine where on the drive the data should be saved. It figures out where the next empty space on the drive is, and it fills it with the new data. Simple enough, but it can get a little hairy.
When data on a block is changed, the SSD needs to recalculate over and over again. Meaning, although the data already exists in a block somewhere on the drive, in order to change the data, it needs to find another blank block and move it there. Why? Because it cannot write data to a block unless that block is empty. If the SSD tried to change data that already exists in a block, it would destroy the data.
Write amplification: a sliding block puzzle for your SSD.
Music Sorter at English Wikipedia, CC BY-SA 3.0
Writing new data entails copying the data from the existing block to a new, empty block, but with the changes included. The drive then marks the previous block as ready to be overwritten, so it can be used as an empty space for the next file that needs it.
For a single file, this process doesn’t really mean anything. But in a game, when things change on-screen very often, or in bulk when you load a saved game, it needs to happen on a much larger scale all at once. We’re talking hundreds or thousands of data rewrites. If the process is slowed down by even a fraction of a fraction, it adds up, even if the SSD can do all of this in a matter of seconds.
When you save a file to your computer, your SSD has to do some (very) quick math to determine where on the drive the data should be saved. It figures out where the next empty space on the drive is, and it fills it with the new data. Simple enough, but it can get a little hairy.
When data on a block is changed, the SSD needs to recalculate over and over again. Meaning, although the data already exists in a block somewhere on the drive, in order to change the data, it needs to find another blank block and move it there. Why? Because it cannot write data to a block unless that block is empty. If the SSD tried to change data that already exists in a block, it would destroy the data.
Write amplification: a sliding block puzzle for your SSD.
Music Sorter at English Wikipedia, CC BY-SA 3.0
Writing new data entails copying the data from the existing block to a new, empty block, but with the changes included. The drive then marks the previous block as ready to be overwritten, so it can be used as an empty space for the next file that needs it.
For a single file, this process doesn’t really mean anything. But in a game, when things change on-screen very often, or in bulk when you load a saved game, it needs to happen on a much larger scale all at once. We’re talking hundreds or thousands of data rewrites. If the process is slowed down by even a fraction of a fraction, it adds up, even if the SSD can do all of this in a matter of seconds.
Last edited:
Currently have 50% free. If I get down to 25% I will try to offload data to my thunderbolt NVMe SSDs and if that's not feasible, it'll be time for an upgrade (I do not see this happening any time soon).
Well, have you ever used a Mac with a full SSD? It can grind to a halt. I’m not sure if it’s just Swap that makes this happen, but it does.
Yes, it can impact performance due to the architecture of the drive. Look up “SSD Queue Depth” for some interesting reading.
Back when SSD's were a new item, we were told that it was pointless to "defrag" the data on an SSD.
Yet I'm wondering if an SSD with the data entirely "spaced out" all over-the-place can take longer to access than does one with the data grouped together in an "electrically logical" manner....?
(shows you how much I know...)
Yet I'm wondering if an SSD with the data entirely "spaced out" all over-the-place can take longer to access than does one with the data grouped together in an "electrically logical" manner....?
(shows you how much I know...)
The manner to retrieve data was different,Back when SSD's were a new item, we were told that it was pointless to "defrag" the data on an SSD.
Yet I'm wondering if an SSD with the data entirely "spaced out" all over-the-place can take longer to access than does one with the data grouped together in an "electrically logical" manner....?
(shows you how much I know...)
The older drives used a sensor that searched for data as a record album
while the ssd drives are micro chips.
i think they store info in specific areas of the flash drive
so no defragmenting is needed
(i think, and shows ya how much i know as well)