Bought a gskill phoenix pro 60gb SSD and got it installed yesterday. Did benchmark with crystaldiskmark. results for ssd along with 1tb caviar black i wrote up a full impression/review thing on my blog if you dont mind checking that out too. ive never wrote any review or anything before so some comments on it would be cool http://derpumad.blogspot.com/2010/10/gskill-phoenix-pro-60gb-ssd-impressions.html ^ hope its okay to link to that?
Just a word of advice... You really should *not* be testing a SF drive (though arguably any SSD) with CDM using a 1000MB setting as this is randomly hammering the nand - & if you run it numerous times then it's just going to make things worse. You'll get equally accurate results lowering it to 100MB but 'dirtying' only a 10th of the nand -> a much quicker recovery. &, of course, these are the lowest speeds for your SSD since CDM uses highly incompressible data - the optimum speeds will be gotten from ATTO - & the real life speeds will be somewhere in between the two depending on the data you're writing.
oh I see. THanks for telling me that. I don't really plan to test it anymore anyways. It's fast and I like it
No worries - just that i have seen some people getting a little carried away with testing their sparkling new SSDs & then wondering why (& getting a bit miffed when) they may slow down for a bit...
See, when people say, "Whats the point of a SSD?", I say, "Look at the random read and write comparison."
What do you call a Portuguese SSD in a bowl of milk? Nand-o's But seriously, i'm looking to get an SSD as well but i was wondering if it really is worth paying 200+ quid for a 128GB SSD to use as a boot drive and install a few games on. Will I be totally blown away and think i've got the bargain of the century or will I regret it?
You won't regret it. And 120GB SSDs are cheaper than that now. Intel's 80GB is on Scan for £136, 120GB OCZ Vertex (the original one) is £158.
I fully agree with phazed you wont regret it, think of all the precious time you could save pwning no0bz!
Music to my ears. The 120GB OCZ Vertex has a nice balance between read/write speeds of 250MB/s and 180MB/s respectively. I was thinking of going for the 128gb Crucial C300 SSD which has read/write speeds of 355MB/s and 140MB/s respectively. Reckon it's worth the extra 60 odd quid for the C300 or is OCZ Vertex quick as it is? Haha i'll give them a few minutes rest! As much as I really do want an SSD i'm still hesitant on forking out for it. It could be one of those things that I really really want but could live without if i had to, i'm not sure lol. As for bootup and load times, how quick is it compared to say like a laptop waking from standby? Obviously i'm not expecting instant boot and load but just as a comparison. EDIT: Sorry for the hijack! I won't continue for long
i actually meant two completely different things by what i'd written (but hadn't quite explained it that way) & forget that each thread needs to be treated as an independent entity so was wrongly assuming that previous posts in other threads were part of what i'd written... my bad. Anyway, (in the 1st half of the statement) each nand block has a ltd no of read-erase-write cycles, & by randomly hammering it with unnecessary writes will decrease the longevity of any SSD. Since the results on a single SSD will be as accurate with a 100MB CDM test as a 1000MB test then there's no sense in running the latter. [NB the one type of occasion where this doesn't hold is when you're using a multiple SSD raid array, the raid card has on board cache & you want to test the raw speed of the SSDs (ie without the cache). in this situation *only*, you would naturally want to run b/ms using test data at a value higher than the cache to eliminate its effects. Obviously though, this is something of a specialised/unusual situation (albeit one that i'm in) though, even if you're 'affected' by it, you'd still want to bear in mind that you don't want to be running it repeatedly for the reason given.] Then, part of the specific longevity of the SF controller SSDs (which the OP's is) is based on the DuraWrite technology which compresses writes on the fly to use fewer nand blocks than other controllers will. This however is based upon usual writes being a combination of highly compressible data (where the SSD will be fastest & least nand used), (predominantly) averagely compressible data & highly incompressible data (where the SSD will be slowest) -> changing the write profile dramatically to highly incompressible data, which is what CDM does, will significantly reduce the lifespan of the SFs... ...though, obviously, running it once or twice will have less of an affect than the cumulative one of b/ming semi-incessantly (which i have seen some people do in other forums). & (in the 2nd half) separately, whilst an OS & SSD combination may have trim &/or garbage collection, these aren't necessarily either available or 'instant (performance) fixes'. As i was trying to give basic sensible advice & the OP doesn't state what OS they're running then leaping to the assumption that they are using an OS that supports trim isn't justified &, as it's general advice that others may read, to categorically state that "trim will sort your woes" wouldn't be accurate for everyone (including additional people who may be running SSDs in raid arrays naturally). Then, as i've said elsewhere re SSDs in R0, the SFs are less responsive to trim than some other SSDs - instead relying more predominantly on their own 'magical' version of GC (it's apparently not the same as normal GC, but NDAs have limited info). This however, whilst far more robust than something like the C300 in non-trim OSes or raid, strikes a far less aggressive balance on cleaning dirty nand than something like the original Vertex drives (with later f/w)... ...though, again as said in other threads, this can be worked around with the SFs (& most other SSDs if needed) by overnight log off idling. So, an assumption that trim (if available) &/or GC will be an instant fix isn't necessarily true depending on both the OS & the specific SSD itself (+ raid arrays), & relentless b/ming would be a sure fire way to get worse & worse results until whatever recovery method has had a chance to run. Anyway, hope this explanation kind of covers things as i'm now going to be away for the next days... As with everyone else who posted whilst i was typing, the speed of a SSD makes such a huge difference & is strikingly noticable. Not 100% sure that i'd go for the original Vertex, though my 2x V Turbos were pretty damn good at the time (& still going strong). Personally think a pair of 50/60GB SFs are a much better buy for most people (unless you're doing major amounts of incompressible writes), as they'll consistently outperform the 120GB C300 & don't require additional expenditure on a 6Gb/s card... ...even though i have a decent 6Gb/s raid card, i still went for SFs d.t. the performance for my usage needs. Still, we're only a few months away from decent 6Gb/s SSDs & at that point my opinion will be reevaluated.
No doubt about it. But i meant simply the comparison of waking from standby/sleep to a functional desktop screen. Obviously the hardware is totally different but i'm looking at the difference in time so I can get a good idea of what to expect from an SSD. Basically, could an SSD boot a PC quicker than a laptop can go from sleep to functional? I'm not expecting it to but it might give me a better idea of the speed of an SSD since i've never really had the pleasure of owning one. Or is this comparison going to actually tell me nothing about it?
I personally don't own an SSD but I will be buying one in the very very very very near future...about this near | |. From reviews of the Crucial C300 it seems to be the biggest contender due to its massive read speeds, which will certainly slash boot times down drastically.
it's obviously going to make no difference to the bios checks but, excluding that, a decent one (or more in R0) reduces the Windows loading time so dramatically...
The only thing that happened with mine is having a SATA 3 card increased boot times, due to a longer time to get to the Windows bootup.
the boot times are quick but whats really important is that windows is responsive immediately. i'd say any ssd is a decent upgrade, with the last gen at those prices now its worth getting in there
Nice, you're part of the eilte now! I'll read your review when I get home. +1 Still really happy with my X-25 M.
