What do i need to speed up installing programs and WinRAR?

I'm guessing it's write speed for installing programs and both for WinRAR?

Here are the three i am looking at -

Samung 850 EVO M.2 - sequential read/write (540/500), random read/write (97,000/89,000)

Samsung PM951 (512gb) - sequential read/write (1050/550), random read/write (250,000/140,000)

Samsung SM951 - sequential read/write (2150/1200), random read/write (90,000/70,000)

So looking at the above, the PM951 has better random write but the SM951 has better sequential write.

Or would i not notice any speed difference and should i just stick with the cheaper 850 EVO?

 

CPU+RAM is what will get 'Work' done. Period.

The storage subsystems you have listed don't matter.

An i7 QC/8T (non 'U' models) processor and the most/fastest RAM you can afford.

If/when this is taken care of, then the biggest SSD you can install with adequate/proper cooling (especially for the M.2 models...) and 33% (or greater) OP'ing on Windows 10 x64 Pro with at least 100GB free space (or more) will give you a system that installs programs fast and runs WinRAR as you wish.

The storage subsystems you list may vary many percentage points between them. But the absolute time will be so small for the tasks you list (depending on the CPU+RAM you have) that we've already wasted too much time with you worrying/thinking about it and I've spent replying.

Hope this helps.

 

Any SSD will feel noticeably faster for installing programs compared to a hard drive, but you're unlikely to notice differences between SSDs as much. Also, as tiller noted, your processor and memory are going to affect your experience as well.

However, with that said, the fastest SSD for progdam installation of the 3 you mentioned would likely be the PM951 due to the faster random write speed.

 

Have been searching around threads and gathering information but couldn't find this specific answer, thanks for the replies.

I should have mentioned that i am looking to buy a new laptop (Alienware 15) so the specs will be quite good and having an quad i7 and plenty of memory.

I guess going from HDD to SSD is the big difference because SSD's are way much faster, (maybe going from one SSD to another is kinda like one HDD to another?).

Obviously in a laptop the temperature will be more of an issue and some people saying the SM951 does overheat and others saying it doesn't.

Will be looking at either 850 EVO at 250gb/500gb or PM951 at 512gb.

 

Yeah, the M.2 SSD's all overheat (and throttle themselves to below HDD levels and also throttle other platform components too - CPU+GPU and even RAM).

The reason why some seem to not be affected is the chassis and cooling design of the notebook in question. And of course, when an M.2 drive comes from the notebook manufacturer as a base design, I (hope) that they have accounted for the extra heat that must be dissipated.

I would suggest you forget about any SSD smaller than 480/500/512GB going forward. And if you were sticking to 2.5" SSD's for your O/S (my recommendation still...), the 1TB capacity or larger would be my minimum if you want the longest useful life with the least reinstalls and the maximum performance over the ownership of that platform/configuration.

Bottom line is M.2 drives, in real world workflows, are barely above a good 2.5" SSD. And they're smaller capacity, more expensive, throttle easily and the only benefits they offer now/still is the bragging rights for irrelevant 'scores' for synthetic benchmarks.

Keep in mind that if you want your new system to still feel fast a few days/weeks/months (depending on your workflows) after you receive it; OP the actual capacity shown in Windows by 33% or more. Yeah; that 1TB SSD now becomes a not so big ~624GB.

Just like Short Stroking HDD's made them an order of magnitude more responsive, OP'ing SSD's is 'required' if you want maximum performance over time, (almost) no matter how you use your system. Note too that that maximum performance is not what is written by the liars/Ad men... It is already much less than half on any decent SSD (and the not so decent SSD's? They're throw away...).

See:
http://www.hardocp.com/article/2013/06/05/sandisk_240_gb_extreme_ii_ssd_review/8#.VrJGpGb2Z9A

That is an old review, but just look at the 'fresh' graphs and the last graph which is a mixed R/W steady state plot. That performance is not what SSD's promised in 2009 (when I started to get interested in them) and they're still a long, long way from giving anything close to what is promised on the box.

The SSD to beat is (still) the 2.5" SanDisk Extreme Pro 960GB - no matter what format - when overall performance is the goal (for an O/S drive).

See:
http://www.tomshardware.com/reviews/ssd-recommendation-benchmark,3269.html

Good luck.

 

If you want your SSD to stay fast, then make sure you don't go past 50% of capacity.
I use a USB 1TB HDD to put files and back up my computer with...
My Samsung EVO 850 500GB SSD is blazing fast this way.

 

eh? Evidence of that?

 

Guess I should have explained better... I notice anytime my SSD's got about 3/4 full, the computer was slowing down... a lot. So when my SSD's get to 50%, I start moving files to my backup USB hard drive. I don't let it get anywhere near 3/4 full. I don't know if all SSD's act this way, but my last four did.

 

Yeah, all SSD's behave like that (just like HDD's do for different reasons).

But free space is the wrong way to fix it.

See:
http://forum.notebookreview.com/threads/why-over-provision.760922/#post-9766709

Free space is a poor way to OP'ing...

And depending on the use of the system in question, will never match OP'ing for maximum storage subsystem performance (almost no matter how you use your drive).

 

Hey
tilleroftheearth Wisdom listens quietly...

You need to practice what you preach

I still stand by what I said as from experience and many articles like this one that discusses trim, yet at the end of it says the exact same thing I was saying:
Why Solid-State Drives Slow Down As You Fill Them Up
http://www.howtogeek.com/165542/why-solid-state-drives-slow-down-as-you-fill-them-up/
They found that “minimum performance improves substantially once you hit 25% spare area for these [consumer] drives.” Their final recommendation was that you should “plan on using only about 75% of [your drive’s] capacity if you want a good balance between performance consistency and capacity.”

If you have a solid-state drive, you should try to avoid using more than 75% of its capacity. Buy a larger drive with more storage than you need and you’ll ensure that you always have consistent write performance. Luckily, SSDs are gradually becoming much cheaper, so this isn’t as expensive as it once was.

 

To clarify for the benefit of those not familiar with the finer points of terminology: I think the point here is that free unpartitioned space is much more effective for maintaining performance than free space within partitions (which may not be contiguous and may be cluttered by file fragments).

John

 

You need to read for comprehension.

Spare area is OP'ing. It is not free space.

Learn some more and come back for more.

 

All I will say is I don't see anyone else backing you up in your posts
I'm outta here, unsubscribing here & the other SSD one and will post elsewhere...

 

I already tried to help you. (Read for comprehension).

Come back when you have something to offer.

I'm not here to always be 'right' or to 'win' anything, but to help where I can and learn too where possible.

 

Ignore him, if he doesn't want to listen, it's his harm... You have given plenty of evidence over the years to back your posts! (I do however think that 40% OPing excessive, 10-20% tops should be the max)...

 

Thanks TomJGX.

Excessive depends on what you want from your storage subsystem.

33% is the level where I find most SSD's in most (workstation class and below) usage scenarios start to level off (benefits vs. lost capacity), but up to that point, gains are to be had when the pursuit is having the 'fastest storage subsystem over time (almost) no matter what the load'.

I've mentioned many times in my desktop systems, I routinely use 50% or more OP'ing - but those drives are just for scratch disk use and... I usually have 3 or more SSD's installed in the same system.

 

As we've discussed previously, in light to moderate workloads free space is just as effective as unformatted space. Reducing partition sizes to allow for unformatted space only has a real benefit over formatted free space when using heavy workloads (as you do) or for those who can't manage their digital storage and fill every drive up to capacity.

So he's not wrong. He's just not seeing things from the viewpoint of a power user.

 

Sorry, but that is simply not true. Don't cloud the conversation by my workloads. I have access to others' systems and their workloads too which I would consider 'very light'. You may recall my post of a client who wanted to throw out an older workstation (complete w/SSD) but finally listened to me and allowed me to OP his setup?

Yeah; free space is not OP'ing. Ever.

A system with free space will have 'touched' each nand cell (making it 'dirty' and needing a read/erase cycle before it can be used again) before long - even in a very light workflow - eventually (but much faster than some would guess or admit to).

A system with unallocated capacity will always have that capacity as ready to write to - that is how the controller is designed - (almost) no matter how you use the SSD.

If anyone believes that free space = OP'd space, they're deluding themselves. Don't listen to me. Go find proof of that from others on the web.

My sources (which I can't track anymore) were from videos (indications of future projects) from Intel themselves and others in the SSD industry (including Samsung, btw) which state the same things I do. OP'ing is the way to get the most from an SSD.

What I did with that information back in 2011/2012 is buy (again) an SSD to test those theories and indications and see if it would make a difference in my workflows. The rest, as they say, is history (at that time; with over 50% OP'ing on a desktop system). With my data today showing that 33% being that point where any extra capacity used for OP'ing is not rewarded as greatly with extra performance and/or possible nand longevity.

Using an SSD without OP'ing is simply a waste of performance potential. Especially sustained performance (my use case), but even for 'normal' users too. All the way from a 7 year old that noticed her system being 'faster', to the client that wanted to trash his old, but once great workstation because his of the stutter, pauses and general feeling of working in molasses - on an Intel SSD.

Up to a point in time (varies with many factors, agreed, but in no situation have I seen it being longer than a couple of months), free space may seem to be as effective as unallocated capacity.

(After that, I think most people just don't notice their drives slowing down or how much better they worked 'fresh'...).

But with many systems having effective lifespans of a half decade to a decade or more for many years now, not OP'ing from day zero of ownership is like buying a whale tail or painting racing stripes (for top dollars) for your compact econo car and thinking you're actually driving in the fast lane.

Take care.