System Device Port Testing

ATTO Disk Benchmark

To validate that the board’s device ports were functioning correctly, ATTO was configured to test against transfer sizes from 0.5 to 8192 KB with Total Length set to 512 MB against the Thinkpad Twist's primary SATA drive. Lenovo uses a priority technology on the Twist dubbed RapidDrive. The RapidDrive technology combines the SATA III drive and the 24GB micro SSD drive into a hybrid drive, resulting in increased data access speeds.

rapiddrive-pic.png

RapidDrive Hybrid drive illustration
Courtesy of Lenovo

External device testing was done against the included USB 3.0 ports using conversion cables to connect to an OCZ Vertex 3 90GB SATA III SSD. Using ATTO, All tests were run 3 times with the highest repeatable read and write scores recorded in MB/s values. The SSD selected for testing has a maximum read throughput of 550 MB/s and a write throughput of 500 MB/s on a SATA III port and a maximum read throughput of 280 MB/s and a write throughput of 260 MB/s on a SATA II port .

sata-perf.png

According to various other testing sites, the real-world performance maximum of USB 3.0 peaks at 200-250MB/s (about 60-75% of its rated 350MB/s throughput).

Apparently, Lenovo did something right with its RapidDrive technology, as evident in the massive read numbers on the SATA III controller. The internal hybrid drive came in with a throughput number of 3825 MB/s, over 30x that of the write performance of the same drive and almost 10x that of an SSD drive on the same controller. This performance was theorized to be a result of the data being read existing in the cache portion of the hybrid drive. The USB 3.0 port performance numbers fell exactly as expected, with both read and write drive performance approaching that of a SATA II device.

SoftPerfect Research NetWorx Speed Test

In conjunction with Windows Performance Monitor, SoftPerfect Research NetWorx Speed Meter application was used to measure the upload and download performance of the motherboards integrated network controllers. Speed Meter was used to measure average network throughput in MB/s with Windows Performance Monitor used to measure CPU utilization during the tests.

Realtek GigE controller

For the wired network adapter, the testing consisted of copying two file sets from and to a remote system directly connected to the local system via a crossover cable. Use of a crossover cable eliminates the possibility of throughput loss due to router passage. The two file sets used consisted of a single 3 GB archive file and a folder containing 3 GB of audio files.

gige-speed.png

gige-cpu.png

The Realtek controller performed well, with upload and download speed during the large file tests within 6 MB/s of each other. The small file transfer tests were slower than the large file tests, with the spread between them at almost 10 MB/s. When compared to the theoretical maximum throughput for gigabit Ethernet (125 MB/s), the large file transfer tests performed better than the small file transfers. The average CPU utilization remained below 15% during most tests with the exception of the small file download tests where utilization averaged over 20%.

Broadcom 802.11n wireless adapter

For the wireless network adapter, the testing consisted of copying two file sets from and to a remote system connected via router. The target system was connected to the router via a wired GigE link to eliminate the possibility of throughput loss due to wireless transmission. The two file sets used consisted of a single 200 MB archive file and a folder containing 200 MB of audio files.

wireless-speed.png

wireless-cpu.png

The wireless adapter performed well with speed averaging 5 MB/s during all tests. This is well below the theoretical maximum of 18.75 MB/s for 150Mbps rates wireless throughput, but remains well within expectations due to overhead and packet loss inherent in wireless-based systems. The CPU utilization remained below 10% during most tests with it going just above for the duration of the small file download test.

Broadcom Bluetooth adapter

For the Bluetooth network adapter, the testing consisted of copying two file sets from and to a remote system directly connected via the Bluetooth link. The two file sets used consisted of a single 10 MB archive file and a folder containing 10 MB of audio files.

bluetooth-speed.png

bluetooth-cpu.png

The Bluetooth adapter performed very well during both the large and small file tests with performance measured during the large file transfers seen to closely approach the 375 KB/s (3.0 Mbps) rated throughput of a Bluetooth-based connection. Due to its wireless nature, Bluetooth-based connections can suffer the same overhead and packet loss issues of other wireless networking types. CPU utilization remained below 5% during all tests.

« PreviousNext »