Last week we wrote about the QuickWeb Supa VZ1 VPS, which sported some amazing results for the price. However, like we mentioned before in the review, we wanted to give you the feeling of how fast their SSD VPSes are. I will tell you right from the start, these VPSes are some of the fastest I have ever used. They're really fast... and I mean it!
The QuickWeb Supa V8 VM2 VPS is quite the bargain for how much you get. For $15 a month, it will give you 512MBs of RAM with 768 swap, two CPU cores, 8GBs of SSD storage, 500GBs of bandwidth, and a 100Mbit port speed (which is actually less than than the Supa VZ1 I have). Of course, granted the price of SSDs nowadays (okay, they are not as much as they were a year ago, but they are still a good chunk of change), you do see some limitations compared to the hard disk based VPSes, like a lower port speed and less storage.
Although I would say for $15 a month, you are getting a great VPS for the money.
As the SSD can download data very fast to storage, I decided to download to the disk than to sending it off to /dev/null. The results are quite amazing.
wget http://cachefly.cachefly.net/100mb.test--2012-02-12 05:37:23-- http://cachefly.cachefly.net/100mb.testResolving cachefly.cachefly.net... 126.96.36.199Connecting to cachefly.cachefly.net|188.8.131.52|:80... connected.HTTP request sent, awaiting response... 200 OKLength: 104857600 (100M) [application/octet-stream]Saving to: `100mb.test'100%[======================================>] 104,857,600 83.2M/s in 1.2s2012-02-12 05:37:24 (83.2 MB/s) - `100mb.test' saved [104857600/104857600]
83mb/s is incredible for a VPS! And as there is indeed less storage on a SSD, I am sure these nodes will be a little bit faster overall as QuickWeb cannot fit as many users as they can with the space hard disks offer.
If you are looking to buy an SSD VPS, you are most likely doing it for the speed. This QuickWeb SSD VPS troops quite a bit!
dd if=/dev/zero of=test bs=64k count=16k conv=fdatasync16384+0 records in16384+0 records out1073741824 bytes (1.1 GB) copied, 5.78825 s, 186 MB/s
root@ssdtest:/downloads/ioping-0.6# ./ioping -c 10 / 4096 bytes from / (simfs /dev/simfs): request=1 time=0.4 ms 4096 bytes from / (simfs /dev/simfs): request=2 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=3 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=4 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=5 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=6 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=7 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=8 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=9 time=0.2 ms 4096 bytes from / (simfs /dev/simfs): request=10 time=0.2 ms --- / (simfs /dev/simfs) ioping statistics --- 10 requests completed in 9009.6 ms, 4281 iops, 16.7 mb/s min/avg/max/mdev = 0.2/0.2/0.4/0.0 ms
One might say that 186mb/s a second isn't that much for an SSD compared to RAID, but if you think about it, this is one drive, not four or five drives combined together (RAID)! Not only are the dd results impressive, but the I/O ping is amazing at a nice 16.7mb/s a second (the VZ1 VPS got a low 0.4 mb/s). That's like 40x faster compared to the RAID config! Hot damn that's good! :P
We know what you're thinking! How's the performance on this baby? As there is one more CPU core on this VPS than the VZ1, I'm thinking it'll do better. Let's see!
System InformationOperating System Linux 2.6.32-274.7.1.el5.028stab095.1 i686Model N/AMotherboard N/AProcessor Intel(R) Xeon(R) CPU E31270 @ 3.40GHz @ 3.39 GHz1 Processor, 2 Cores, 2 ThreadsProcessor ID GenuineIntel Family 6 Model 42 Stepping 7L1 Instruction Cache 32.0 KBL1 Data Cache 32.0 KBL2 Cache 256 KBL3 Cache 8.00 MBMemory 15.6 GBBIOS N/AIntegerBlowfishsingle-threaded scalar 2192 ||||||||multi-threaded scalar 4699 ||||||||||||||||||Text Compresssingle-threaded scalar 3001 ||||||||||||multi-threaded scalar 5844 |||||||||||||||||||||||Text Decompresssingle-threaded scalar 3256 |||||||||||||multi-threaded scalar 6689 ||||||||||||||||||||||||||Image Compresssingle-threaded scalar 2464 |||||||||multi-threaded scalar 4827 |||||||||||||||||||Image Decompresssingle-threaded scalar 2439 |||||||||multi-threaded scalar 4987 |||||||||||||||||||Luasingle-threaded scalar 4269 |||||||||||||||||multi-threaded scalar 8506 ||||||||||||||||||||||||||||||||||Floating PointMandelbrotsingle-threaded scalar 2837 |||||||||||multi-threaded scalar 5769 |||||||||||||||||||||||Dot Productsingle-threaded scalar 4670 ||||||||||||||||||multi-threaded scalar 9902 |||||||||||||||||||||||||||||||||||||||single-threaded vector 5565 ||||||||||||||||||||||multi-threaded vector 12820 ||||||||||||||||||||||||||||||||||||||||LU Decompositionsingle-threaded scalar 3021 ||||||||||||multi-threaded scalar 6120 ||||||||||||||||||||||||Primality Testsingle-threaded scalar 4885 |||||||||||||||||||multi-threaded scalar 7831 |||||||||||||||||||||||||||||||Sharpen Imagesingle-threaded scalar 11411 ||||||||||||||||||||||||||||||||||||||||multi-threaded scalar 22952 ||||||||||||||||||||||||||||||||||||||||Blur Imagesingle-threaded scalar 8862 |||||||||||||||||||||||||||||||||||multi-threaded scalar 17800 ||||||||||||||||||||||||||||||||||||||||MemoryRead Sequentialsingle-threaded scalar 7839 |||||||||||||||||||||||||||||||Write Sequentialsingle-threaded scalar 12333 ||||||||||||||||||||||||||||||||||||||||Stdlib Allocatesingle-threaded scalar 5595 ||||||||||||||||||||||Stdlib Writesingle-threaded scalar 8842 |||||||||||||||||||||||||||||||||||Stdlib Copysingle-threaded scalar 17438 ||||||||||||||||||||||||||||||||||||||||StreamStream Copysingle-threaded scalar 7279 |||||||||||||||||||||||||||||single-threaded vector 8554 ||||||||||||||||||||||||||||||||||Stream Scalesingle-threaded scalar 7696 ||||||||||||||||||||||||||||||single-threaded vector 8242 ||||||||||||||||||||||||||||||||Stream Addsingle-threaded scalar 8067 ||||||||||||||||||||||||||||||||single-threaded vector 9028 ||||||||||||||||||||||||||||||||||||Stream Triadsingle-threaded scalar 8792 |||||||||||||||||||||||||||||||||||single-threaded vector 6686 ||||||||||||||||||||||||||Benchmark SummaryInteger Score 4431 |||||||||||||||||Floating Point Score 8888 |||||||||||||||||||||||||||||||||||Memory Score 10409 ||||||||||||||||||||||||||||||||||||||||Stream Score 8043 ||||||||||||||||||||||||||||||||Geekbench Score 7547 ||||||||||||||||||||||||||||||
As you can clearly see, the score is about 1.4x faster than with one core. Logically, you would of expected 2x faster, but technology isn't that persistent. Never alas, it's for sure something to not complain about; that's a very fast VPS right there! We have to test the UnixBench scores though to see how it can stand up with consistent load. This should be where the differences are.
TEST BASELINE RESULT INDEX Dhrystone 2 using register variables 376783.7 18826207.6 499.7 Double-Precision Whetstone 83.1 1952.5 235.0 Execl Throughput 188.3 12461.4 661.8 File Copy 1024 bufsize 2000 maxblocks 2672.0 338868.0 1268.2 File Copy 256 bufsize 500 maxblocks 1077.0 96341.0 894.5 File Read 4096 bufsize 8000 maxblocks 15382.0 2533936.0 1647.3 Pipe Throughput 111814.6 2404133.8 215.0 Pipe-based Context Switching 15448.6 774252.4 501.2 Process Creation 569.3 36065.4 633.5 Shell Scripts (8 concurrent) 44.8 1986.0 443.3 System Call Overhead 114433.5 1811852.9 158.3 ========= FINAL SCORE 516.4
With the UnixBench being 1.8x faster than the VZ1 VPS, I'd say the extra processor sped up the VPS by nearly two times, which realistically should happen. The performance for a two core VPS at only $15 a month is not bad at all!
I am very impressed with QuickWeb's v8 VM2 VPS; it well exceeded my expectations on a high-performance VPS. If you are looking for a VPS that will handle database heavy applications with consistent load times, QuickWeb's v8 VM2 VPS is for you! Hard drives might handle sequential reads very nicely, but these SSDs can handle random I/O way better than anything else can!
Overall, I give this VPS an A+ on everything. Despite the slower port speed, it is ironically faster than the one gigabit line as well; I guess that's an SSD for you!
Let @tjasko know on Twitter!