ZIF equipped notebook owners now have a great value performance storage upgrade option. Runcore ProIV ZIF PATA SSD combines the high performance Indilinx Barefoot controller as used in OCZ's very popular Vertex, adding a sata-to-pata bridge chip for use on ZIF systems. Performance testing shows an average of 3.4-times-faster performance than a ZIF HDD. It's available for purchase from MyDigitalDiscount or SolidStateCentral (Aus).
1: ProIV ZIF internals. Indilinx Barefoot controller, Samsung NAND flash, Jmicron JM20330 sata-to-pata bridge chip
2: Package contents: USB enclosure, instructions, SSD drive, 3 ZIF cables, USB cable, 2 screwdrives and stickers
3: ZIF SSD/HDD width comparison. ProIV circuit=~2mm, ProIV chassis=5mm, Toshiba 80GB HDD=8mm
4: ProIV ZIF SSD as installed in HP 2510P
Power consumption idle/active:
0.5/0.83W (measured 0.5/2.0W)
Installation
Installation was easy. A USB enclosure is provided to house the SSD supplied with Acronis Trial EasyMigrate to clone existing data. Initially I had a hard time inserting the ZIF cable into the ZIF socket until I realized the black strip on the ZIF socket is a stiffener as shown here. Flicking the stiffener into an upright position allowed very easy insertion of the ZIF cable.
Performance Comparison: Runcore ProIV ZIF SSD versus Toshiba ZIF HDD
Tested platform: XP SP3, HP 2510P U7600-1.2 2GB ICH8M UDMA5/ATA100 I/O.
^1 - measured using powertop in Linux/Recovery console. Toshiba ZIF HDD's 1.1W rated Active was used as a reference point.
^2 - industry standard used to measure performance.
^3 - link shows drive hparm output and partition alignment details
^4 - 4kb-64thrd a good reflection of overall os/app responsiveness
The above benchmarks compare the HP supplied 4200rpm 1.8" ZIF HDD against the Runcore ProIV ZIF SSD. An identical XP partition image was used to give back-to-back comparison of bootup time. The image and partitioning was optimized for best performance on both the SSD and the HDD by setting it to be (i) defragmented using Perfect Disk (ii) it was aligned to a 512MB boundary for best SSD performance. WinBootInfo gives the precise bootup time data. The actual 26GB of total data occupied the top third portion of the 1.8" ZIF HDD so the benchmarks are better than average and will worsen as the drive fills and seek times increase.
The Runcore ProIV ZIF SSD is a very noticable improvement in performance over the 1.8" ZIF HDD. Going from a coffee break inducing 80.2 seconds to 21.6 seconds XP boot time, or taking 122 seconds versus 350 seconds to read and write a 2GB file. That's an average of 3.4 times faster performance, the sort of disk I/O activity that can be said to extend to all disk activity. There's no ticking noise, Firefox doesn't have momentary seek delays when scrolling windows or reading/writing cache. Applications just popup instantly. Once experiencing this sort of speed it's hard to go back to using the ZIF HDD.
The OCZ Vertex in a sata-to-pata optical bay caddy is included for comparison. We see it performs very similarly to the Runcore ProIV. It's virtually the same setup but in 2.5" factor and uses the Marvell 88SA8040 sata-to-pata chip. Back-to-back comparisons of the both bridge chips is in the optical bay caddy review.
Power Consumption
We do see the SSD consumes more idle and active power consumption than the 1.8" ZIF HDD, though it's higher performance means more time spend in idle mode. So far it appears to be at least on par with perhaps a slight battery life improvement than the supplied 1.8" ZIF HDD.
Garbage Collection
I'll also add that at one point in my testing, after what must have been a complete write to all the cells on the SSD, CrystalDiskMark write performance halved, exactly as described here. This was solvable by either (i) installation of firmware 1916 which delivers internal, transparent garbage collection and Win7 trim support or (ii) older versions of firmware without GC/trim can manually run wiper.exe the write performance was restored to "as new" levels per OCZ's recommendation of running wiper.exe 2-3 times a week. It took less than 30 seconds to do it's work when I ran it.
Pros
one of the best performing ZIF SSDs as shown by 4kb random and sequential i/o performance
an average of 3.4-times faster bootup and sequential read/write performance improvement over a 1.8" ZIF HDD
Win7 trim garbage collection maintains write performance [ requires firmware 1819 or newer ]
Appendix: Notes about realistic performance from the ZIF PATA interface
Many candidate units for the Runcore ProIV ZIF SSD such as the Dell D420/D430, HP 2510P/2710P, Toshiba R500, HP Mini 1000, Lenovo U110, Asus U1E will have the Intel ICHxM UDMA5/ATA100 PATA I/O interface rated at 100MB/s read and 88.9MB/s write. With transfer overhead this decreases to approx ~90MB/s read and ~80MB/s write. There are some exceptions that can do UDMA6 which is limited to 133MB/s reads.
You may ask, well that interface speed is a lot less than SATA-II's 3Gbps or ~250MB/s read. Yes, SATA-II SSDs do perform better at sequential read/write operations such as hibernate or multimeda file reads, but your OS/App loading time is predominantly affected by 4KB random I/O transfer rate. The OS reads multiple small files on your filesystem. 4KB read speeds are only just crossing 30MB/s by Intel X25M's SSD, so the ZIF interface still has plenty of bandwidth to accomodate advances in SSD technology to come. 21MB/s 4KB/s reads puts it within reach of the fastest consumer SATA SSDs.
Runcore ProIV ZIF SSD: usability as a native sata product? as posted on Runcore Support Forum.
Since the ProIV ZIF SSD is using a Indilinx SATA SSD controller then a JMicron JM20330 chip to bridge from sata to pata (ZIF), I was curious whether the product could be easily converted back to being a native sata product?
Left: JMicron JM20330 TX/RX sata pins used for continuity testing
Right: supposed J2 microsata header used for continuity testing
My curiousity was such that I asked a Runcore engineer what the jumper were used for. I was told J2 was native SATA pins with a pinout as shown.
Great future upgradability bonus I thought. Just solder a cable to the J2 header and a 1.8" microsata connector/lead and I could reuse the Runcore ZIF SSD as a microsata device, or even as a 2.5" sata drive using a 1.8" to 2.5" adapter.
So then I performed continuity testing against the J2 pinout indicated against the JM20330's TX/RX pins. This was not successful. The RX/TX pins just read high impedance (not connected) with only the 3.3V and GND pins passing testing. To double-check in case the product was doing some fancy switching of signals when powered on, I jumpered the J2 SSD TX+/TX- directly to each other via a resistor. This would mean the SSD should fail to operate when powered. So I switched on the unit finding the SSD continued to operate and boot the OS. Clearly the pictured board, version ZIF18_VL03, does not have a J2 with microsata pins.
Even so, all is not lost. It is still feasible to solder 4 wires of an 80-pin IDE to the Jmicron's sata TX/RX pins directly and another 2 to the 3.3V and GND pins to give all the wires needed for microsata. Not as easy as soldering onto the nicely spaced J2 pins, but still possible with the right equipment and some skill.
I would be curious to know if anybody else's ProIV ZIF SSD's J2 does pass sata TX/RX pin continuity testing. If so, what board revision do you have?
Standard warning: Exercise care if probing the circuit board to avoid short circuiting components. Kiss your warranty goodbye if doing any soldering.
With the old firmware, I wasn't able to run the wiper tool.
Now that I have upgraded the firmware thanks to your link on the support forum, Nando, I can execute the wiper tool. But after nearly 24h, It's only at 0.44%.
At this rate, it will take 6 months. Maybe that's because I've been using the SSD for 4 months without TRIM, I don't know....
Do you know another way to recover my performance, a faster way, even If I loose my data and I must reinstall my OS ?
Do you know another way to recover my performance, a faster way, even If I loose my data and I must reinstall my OS ?
Ensure your SSD is partitioned to use an erase block size boundary as discussed here. A fresh Win7 install will align to a 1MB boundary which will do the trick. Wiper will then run in seconds/minutes.
If reinstallation/repartitioning is not feasible, can try a Tony Trim instead.
Well now I have a big problem.
I have started to run perfect disk 10. And after 30mn it crashed my OS. And I cannot reinstall it, the installation program says that no drives ar found. Even Ubuntu doesn't find any disk.
I fear that Tony TRIM has killed my SSD.
Edit : my laptop is working again. I didn't do anything except letting it rest for a few hours. Even Windows is working without a reinstallation. Strange.
TechTarget provides enterprise IT professionals with the information they need to perform their jobs - from developing strategy, to making cost-effective IT purchase decisions and managing their organizations' IT projects - with its network of technology-specific Web sites, events and magazines.
Runcore ProIV ZIF SSD User Review
