Recent real world tests of Samsung's RAPID tech?

[OrionMaster]

Not looking standard for benchmarks, we know Rapid looks great there. I saw some real world tests but they were from last year. Does anyone have realistic tests from the latest Rapid Version (1.1 IIRC)?

 

[BonzaiDuck]

Not looking standard for benchmarks, we know Rapid looks great there. I saw some real world tests but they were from last year. Does anyone have realistic tests from the latest Rapid Version (1.1 IIRC)?

What sort of decision are you facing?

I couldn't say anything about "real world tests:" I was impressed with the benchmarks (6,000+ MB/s seq-read); I could "feel" a slight improvement.

I looked into other "caching" options, particularly for an old laptop I acquired. Some will say "not worth it;" others offer more positive remarks like mine.

Ultimately, my decision to use RAPID (or anything else as I did with the lappie) depended on reliability and stability. The RAPID feature hasn't made anything "worse." Similarly for the lappie and Primo-Cache. In that situation -- definitely "better."

Maybe look at it another way. Recently there was a comet approaching Mars, supposedly traveling 135,000 miles an hour. Light travels at 186,000 miles per second. For the human, a blink of the eye is just . . . a blink of the eye. We're getting to a point where it's harder to notice the difference, even if there is one.

Using RAPID is not likely to cost you or complicate anything. Neither is not using it, or so you may likely perceive.

 

[OrionMaster]

Actually, the real world tests from last year did show that rapid made things worse (for some things, def not all). the author even said he didn't recommend turning it on

http://techreport.com/review/25282/a-closer-look-at-rapid-dram-caching-on-the-samsung-840-evo-ssd

 

[BonzaiDuck]

Actually, the real world tests from last year did show that rapid made things worse (for some things, def not all). the author even said he didn't recommend turning it on

http://techreport.com/review/25282/a-closer-look-at-rapid-dram-caching-on-the-samsung-840-evo-ssd

Interesting review, although tests performed and review written before the newer version of RAPID. The original version limited the cache to 1GB.

That difference shows up in the benchmarks as a considerable improvement. In my case, with 16GB of RAM, the seq-read rate went from 1,100 to 6,000 plus. But again, I'll be the first to say "that's just a benchmark!"

Of course, the real promise of any improvement would show in the 4K-Random read tests. That's where your desktop end-user is going to "feel" snappier response. I get less improvement in the 4K tests on my sig desktop than I get with my Primo-Cache-enabled laptop -- a factor of 3x better. It also depends on the test. AS-SSD gives one set of numbers; Crystal-Disk-Mark gives another; Samsung gives something else.

Put it another way in a relative comparison. When ISRT was implemented on the Z68 (and later) motherboards, it definitely gave you 80% of native SSD speed for accelerating a run-of-the-mill SATA-II HDD. (Or the appearance thereof). I used it for 30 months.

You wouldn't notice this similar arrangement between RAM and SSD nearly as much.

I really can't make a definitive pronouncement about it. I just haven't found any "down-sides" to using it over the last six months.