It would seem that if I went SSD for OS I would, as of this moment, notice no obvious change in performance between SATA2 and SATA3 based on drive performance. Then what is the motivation for moving to SATA3 as a standard? Since SATAII/SATA 3.0Gb/s/SATA-300 (are these all the same?) Seem to offer theoretical transfer rates of obviously 3.0 GIGAbits per-second, and since SSDs look like they top out at a around 400-500Mb/s for Sustained Read/Write. Reminds me of Coriolis force? I would run a short stroke but I need the space+transfer rates of all my HD files. Thanks for the advice, that makes perfect sense about the discrepancies in angular velocity. (Just google short stroke.Īs mentioned - SSD are the way to go to improve performance. You live the remaining 70 % unused - ofcoarse for the gain in performance, you lose disk space. On Your Raid0, if you want to get the best performance then use "short stroke" This is where when you create the array, you only use about 30 % of the avalable space. It primarily exists as a memory cache, but recently has gained support for a second layer SSD cache (a bit like the ZFS L2ARC), which can be arbitrarily large.
Fancycache vs supercache software#
As the platter is filled the closer you get to the inner part the the platter the angular velocity decreases. There are however, pure software implementations too: Fanc圜ache It works like you might expect performing read and write caching at the block layer. Writting to the drive starts at the outside edge of the platter and this will be where the highest throughput will be (highest angular velocity. Max throuhput = RPMs as it translates to angular veolicy, The density of the magnetic domains on the platter. Is it that while my Mobo and SATA connections themselves can go "up to" 3.0Gb/s my actual HDDs top out at 170Mb/s?ĭoes anyone actually know what the maximum (even theoretical) Read/Write rates are of 7200 SATA 3.0Gb/s HDDs?Īnother Google crawl on the subject revealed more obfuscation.
Fancycache vs supercache windows#
I can only assume this topic has been hit dozens of times before, so I apologize in advance. Fanc圜ache -hard drive block level caching program, usefull for systems with 4GB+ of ram and even more helpfull to windows XP 64bit, with an outdated cachin. I did a Google crawl and was inundated with data. Is this all I can get out of these drives? I have brand new SATA cables as well (though in most cases that won't matter). i havent tried it yet, so i dont know if you can simply assign 0 MB to RAM and use L2 cache exclusively. Esiste principalmente come cache di memoria, ma recentemente ha ottenuto il supporto per una cache SSD di secondo livello (un po come ZFS L2ARC), che può essere. I checked the channels between these three drives as well, and regardless if the Barracuda shares a channel with either Constellation, it still stays at 130Mb/s when writing to the Barracuda. yes, fancycache allows you to cache anything with anything, provided that there is a cache in the memory as well. Esistono tuttavia anche pure implementazioni di software: Fanc圜ache - Funziona come ci si potrebbe aspettare eseguendo la cache di lettura e scrittura a livello di blocco. I know transfers are set by the slowest drive, so either the write rate of the Barracuda is lower than it's read (by 35+ Mb/s?) or it's the reverse for the Constellations. If I am reading off to the Barracuda the Array settles down at around 130Mb/s. That's surprisingly only when I am writing TO the Constellation Array from the Barracuda. I can only hit these numbers between one brand-new ST32000DM001 Seagate Barracuda (64MB Cache) and two RAID0 Seagate Constellation ES (32MB Cache) drives, and only even then when I make sure the two RAID0 Constellations don't share the same channel. The fastest transfer rates I can hit are 167Mb/s. costs.12 Gb.So I have a few HDDs, all at 7200 and all are SATAII/SATA-300/SATA 3.0Gb/s rated. Have you considered how INSTEAD of using "Cache B" or *any* other "Cache" program, in order to lessen writes to hard disk of "PC A" when running "program C" using a RAMDISK would be better/more efficient?īla.bla. You assume that on your particular "PC A" configuration using a Cache program, let's call it "Cache B" will be effective to run a flight simulator, let's call it "program C".ĭo you know what alternative to "Cache B" Cache program I could use to have "program C" do less writes on hard disk of "PC A"? For aggregate performance on tracking/recording, I would expect a RAID5 to give the best tradeoff between throughput and data protection, but does call for three or more drives in the array. Fanc圜ache, now here renamed, appeared as a reaction to linux bcache from googler which in turn was a clone of things before it to bring linux on par with stuff like Dillon’s FreeBSD swapcache and of course ZFS tiered and metadata caching which is slightly different, but the primary use-case, and what they want to bring btrfs up to.
Fancycache vs supercache Pc#
You have a PC with a given configuration, let's call it "PC A". Now 32 GB DDR3 costs as much as ~12 GB DDR2 -no make sense additionally change the CPU, motherboard and memory, do not need this at this time