SSDs have been touted as the hard drives of the future, promising fast boot times, low power consumption, noiseless operation, resistance to harsh conditions and more.
Only recently have there been SSDs at a price point where normal people can purchase them. One such example is the SSD in this review, the Patriot Warp V2 64GB SSD. The drive in the review was purchased at a price of 180$ CAD. Most reviews out there are of the 128GB edition, so we thought it would be a good idea to actually review something people would be more likely to purchase, due to a more reasonable price point.
- Here are the highlights of this product:
- Available in 32GB, 64GB and 128GB capacities
- Interface: SATA I/II
- Raid support: 0, 1, 0+1
- Dimensions: 99.88 x 69.63x 9.3 mm
- Weight: 91g
- Sequential Read: up to 175MB/s
- Sequential Write: up to 100MB/s
- Shock Resistant: 1500G/0.5ms
- Vibration Resistant: 20G/10~2000Hz with 3 Axis
- Operating Voltage: DC 5V
- Power Consumption: 280mA~330mA
- Operating Temperature: -10Â°C~70Â°C
- Storage Temperature: -55Â°C~125Â°C
- MTBF: >1,500,000 Hours
- Data Retention: > 5 years at 25Â°C
- Data Reliability: Built in BCH 15-bit ECC
- O/S Support: Windows 2000/XP/Vista Linux, and Mac OSX
- Certification: FCC/CE/RoHS
- 2 Year Warranty
In summary, this particular drive is based on the infamous JMicron controller, which everybody has been criticising for very good reasons.
This particular controller and drives built on this controller have a very serious problem, stuttering and random freezes. Some SSD drives even will not allow Windows Vista to install. Why? Because this particular controller is not too happy with random writes, particularly, the access times it takes to write a file randomly. These kinds of drives will easily put over 200 milliseconds on random writes, per random write. To put this into perspective, a standard laptop hard drive will put about 20 milliseconds on the slower 5400 RPM drives per read or write operation. This means that on random writes, an SSD Based on the JMicron controller, the JMF602, will be more than ten times slower than your average el-cheapo hard drive. We will be running benchmarks to see if this drive suffers the same problems as have been noted on other similar drives.
Our test platform for this drive had the following specifications:
- Processor: Core 2 duo E8400 @ 3.6GHz
- Memory: 4GB OCZ DDR2 800MHz Ram in Dual Channel
- Motherboard: Gigabyte EP35-DS3P (ICH9r Controller)
- Power supply: Enermax 550w
- Graphics Card: BFG Nvidia GeForce 8800GT OC 512MB
- OS: Windows Vista Ultimate 64bit SP1
HDTach is a very popular hard drive testing benchmark. The graph is self explanatory. Comparing this drive to a normal hard drive, using only HDtach, one can imagine super fast boot times, and a super sped up system. Hold on till you see the rest of the benchmarks though before concluding that this hard drive is fast.
This is pretty self explanatory, so on to the next benchmark.
This one actually needs some explaining. The constant drop offs after the good start is due to the JMicron Controller not doing very well. Once the data exceeds the cache, the drive performs very poorly, worse than a standard hard drive. Continuing alongâ€¦
This particular Test used the driveâ€™s buffer, so this is basically a test to see how fast the buffer can read. Not necessarily an important test. But it is interesting to note that this drive is actually able to exceed the manufacturerâ€™s specifications, but these are for â€śsequentialâ€ť which is not the same.
This test is more interesting, as this represents the actual speed that one can expect of the drive. It is sort of disappointing to see that it is not able to achieve the manufacturerâ€™s rating on average. Regardless, it is still quicker than most desktop hard drive available today.
Read access times are VERY Important, and is the Main reason why we can see SSDs booting windows faster than conventional drives, since it is the time required to find a file, the lower the better. The numbers in the sub 0.2ms range are very SSD Like.
Buffered Write is where we see how fast the buffer can be written to. Sort of like a standard hard drive cache. Having high numbers here is good for a few small writes not exceeding the cache, since it helps expedite the process of â€śwritingâ€ť to the disk.
And here is where the Horror Show Begins. The graph starts of looking decent, but once the cache gets flooded, the JMicronâ€™s performance is just simply lacking, causing dips all over the place. This is where the â€śstutteringâ€ť issues reported comes from. What is more important is the next graph:
Here again, it LOOKS to start off okay, but the cache is quickly exceeded, and each random write access takes a whopping 221ms on average. Just so you know, even the minimum 21ms access time is already as slow as a 2.5â€ť 5400 RPM laptop drive. Once we get to 221 ms, we are effectively over 10 times slower than a desktop hard drive. To put this into perspective, most modern day 7200 RPM hard drives have times of 13-15ms. Basically, once the cache is flooded, you can expect slow performance, slower than a conventional desktop or laptop drive.
As you can see from the benchmarks, this drive looks like it suffers from the notorious â€śstutteringâ€ť. However, subjectively testing, this drive is still one of the properly designed ones based on the JMicron controller. I did not have any issues installing windows vista on this drive as others have had on other branded SSDs. As far as stuttering is concerned, I have not encountered stuttering that much in everyday use, only when I was pushing the SSD to do too much, more than a standard drive would be able to handle anyways, did I encounter stuttering on this drive. Rather than stuttering, this drive gave me very brief periods of â€śpausesâ€ť, where the mouse would still be active but the system would remain unresponsive. These periods ranged from a fraction of a second to maybe 3 seconds. This might sound like a serious problem, but under normal usage, I never encountered this.
Long Term Testing
This drive has been in use for approximately 4 months, and no data corruption has occurred. The system has remained very stable and as far as performance degradation is concerned, these drives do suffer from it to a certain extent. A few suggestions have been floating around as to how to restore the driveâ€™s original performance. One was to backup and completely wipe the drive using cleaning tools and to then restore the information. We have done this to see what would result:
This is the Used Drive:
And here is after the wipe:
We can also compare HDTach results:
The drive was wiped with Acronis Disk director suite, one pass.
As you can see from the pictures, before the wipe, the peak reads were actually higher, whereas after the wipe, the drive performed better on writes. I did not understand this behaviour, so I re-ran the test three times, each time with identical results, where the writes were indeed quicker, but the max read speeds had fallen. This was done on the exact same computer as well. This behaviour is only noted on the Atto Test, not on the HDTach Test. I think that the differences are not worth bothering though.
In the end, I am still happy with the SSD. My original hard drive, a Seagate 100GB 7200 RPM, failed due to high vibrations and shaking due to my laptop randomly turning on in my backpack. SSD solved this problem.
As far as power consumption is concerned, this drive actually seems to consume more power than the average laptop hard drive. It will consume, as per the manufacturer ratings, 1.4 watts under idle and will consume 1.6 watts under load. These numbers are rather high, considering the amount of reports saying that SSD should lead to lower power consumption and better battery life, since there isnâ€™t a motor spinning. We can compare SSDs to a processor. A processor, although it has no moving parts, still consumes energy, since electrons must flow. The same can be said of a SSD. A better comparison would probably be RAM, which also consumes energy, but is not normally considered as power hungry, for very much the same reason that a processor consumes energy.
Interesting SSDs are those like the Intel and Samsung SSDs, which consume far less power, and can potentially boost battery life.
Here is a final treat:
Given that we have OCZâ€™s very good and well supported OCZ Vertex, which does not have any of the stuttering issues, I cannot really recommend this drive anymore. The only reason to purchase one of these today would be due to prices. Do not get me wrong, this drive will do fine as an OS Drive for a quiet system such as a Home Theatre PC, just that if you are looking for performance, there are better alternatives out there.