Ummmm... but without testing your own personal exact usage over time then all b/ms are artifical... ...to a greater or lesser extent. Then, whilst you're saying you agree with me... ...'if' you are after very specific usage types then Anand's drop down tables would be incredibly useful... (as for some *very* specific data types most half decent SSDs will beat another one/be less good) ...however, for almost all *consumers*, (again) you do want to be looking at the light & heavy b/ms within the reviews themselves (as linked to). Well, to take an easy example, as a normal OS/apps/games environment will have both a small %age of r/ws which 4K & almost none are either QD64/QD32 or QD1... ..then a b/m that showed that brand banana is better than brand apple on one or t'other is irrelevant for almost everyone. Whilst you may have ~25%(+) as 4Ks in a typical consume usage then both ~75%(-) will be some other file size &, for the 4Ks themselves (& large no's of other r/ws) the average QD will be 3-7/8 - so neither very high or very low QDs are accurate or are representative of the general data that's read or written. Then, as said already, you add on all of the extra benefits that, whilst not instantly speed related, can either effect longer term speeds &/or longevity &/or error recovery &/or capacity... ...& you value them against your personal budget.
Well, I was pretty close to ordering a pair of vertex 3 max iops, however went back to the review of the corsair performance pro while checking to see if there were any new reviews, and it turns out it uses 'Toshiba toggle mode 34nm NAND', same as the max iops I believe. Thoughts? Is the V3 max iops still the better drive? It seems that the performance pro is slightly better for incompressible, but since most data on an os drive will be compressible the V3 mi is probably still the better choice. The review.
Right - the change in nand then is a reasonable justification for the price over the M4 & that should compensate, to a reasonable extent, for the -ve effects of the faster write speeds on the nand longevity... ...though it will still have a much higher write amplification than the SFs -> higher cycle usage -> lower longevity & a greater need for trim &/or GC to do it's thing. 3 obvious things to note though - 1. it's naturally the 256GB that's tested - not the 128GB that you were looking at. 2. there's no info as to whether GC has actually improved compared to the M4, 510, etc - as you were looking at 2x 128GBs in R0... Though it is *very* unlikely to be as good as the SFs (esp with their lower write amplification coming into play). 3. the 'best' of the b/ms used (unless you've got a very specific usage type) was the PC Mark Vantage one... [NB there is a potential problem with this comparative table though as the system they are using for this drive is different from the one used to test, for example, the V3 - see here - which does not mean that the results are necessarily fair either way... Naturally, if there'd been years in between, then i wouldn't expect them to be using the same system - i'm not trying to suggest that they should be hooked up to something ancient - but... ...well, Anand used the exact same setup & driver versions to test, for example, the V3 in March, the M4's 0009 in August & the Octane in November so the results are clearly comparable.] ...however (assuming that they were fair) that does base a noticeable amount of its testing on things like video editing & recording which are not the best use of SSDs & heavily skew the results in the favour of highly incompressible writes. As it then says in the conclusion - "The question then becomes one of whether an educated consumer will select the SSD with better performance results when transferring large files or the SSD that can transfer large amounts of incompressible data such as music, movies and photographs at a very quick speed. This is where some should look closely at their needs. For the typical consumer, however, there will be no visible performance increase or decrease whatsoever and that is when the pocketbook is watched closely." [NB why "SSD enthusiasts have made it clear that they want an SSD with superior incompressible data performance results and now they have it" is a bit beyond me... (though it's clear, from lots of posts on this forum & bittech's reviews & whatnot, that many people do wrongly believe that things like AS-SSD & CDM results will have a considerable bearing on a general r.l. usage) ...that's not to say that anyone should want shonky incompressible results, but their conclusion, as quoted verbatim does appear to agree with me. Of course some data will be more compressible than others - & this is the reason why i like the fact that Anand's b/ms include data types to cover things such as photo editing (which would, naturally, penalise the SFs to some extent) - but to suggest that most users would be writing huge quantities of 99.9% compressed data is clearly quite foolish.] Anyway, whilst i would personally wait for Anand's light & heavy b/ms as they are much superior for most people's usage, provisionally, this is another premium version of the M4 - but at a cheaper cost than the 510. it's appears to be slightly slower than the M4 for reads, but unless you were using it as a (comparatively) static data drive (ie games installations), the faster writes more than make up for it - making it a better OS/apps/etc drive than the M4. Whilst there is likely to still be a lower longevity than an M4 of the same capacity for the same actual writes (inc write amplification, block combining & wear levelling), the use of 3Xnm nand allays my overall concerns about the effect of the faster write speed &, unless you had a very high cycle usage (& assuming there's nothing inherently flawed about it in general or the specific one that someone received), it should survive the 3 year warranty. Now, as a single drive, (whilst it would noticeably affect the longevity) 'if' someone were want to actually be using it for very significant quantities of incompressible file converting/transcoding/etc (as the primary use) it would be a better drive than the max iops... ...however this is not what most people would sensibly use a SSD for (or if they did, for professional reasons, they'd be looking at higher rated drives &/or much quicker replacement to, amongst other things, deal with the longevity issue) & so the 240GB max iops will be noticeably quicker than the 256GB Corsair than the PC Vantage test suggest... ...plus having the advantages of lower write amplification, better error correction, etc. You, however, are personally looking at a R0 setup so there's then the proven added robustness of the SF GC to add in which, combined with (esp) the much lower write amplification, would be the tipping point for me. But again, i am not saying that using non-SFs in R0 is doomed to failure - simply that there are significant advantages to the SFs in non-trim that, imho, justifies the price difference. it is your money to spend though, & so whilst i've tried to explain things as best as i can throughout all of these posts...
