EVDO SuperCharged 5000K |
| Wednesday, 23 February 2005 | |
(Sorry! You can't get EVDO that fast)No, you cannot get 5000kbps via the current EVDO. That headline is there to get your attention! EVDO has a maximum speed of 2.4Mb. However, the average speed is 300-500kbps (unless you check out our Turbo Boost Guide, and you can lock into EVDO towers for faster downloads). You may have seen some wild claims on the Internet boasting incredible EVDO speeds. These companies use benchmarks to "prove" how fast their download speeds are...but we at EVDOinfo would like to tell you the TRUTH about benchmarks! Benchmarks can be skewed to give better looking results than you get in the real world. This article will give you the facts. Updated Information. See how to optimize your TCP settings to get FASTER download speeds. Here is something that we found on the Internet: "1XRTT speeds are at about 400kbps EVDO speeds are at about 3000kbps if you dont get BETTER than these speeds... let us know... we want to get you maximum speed." Here is another wild claim: "my 'bandwidth tests' on 1xrtt - they are showing 480-667 kbps!" First of all, 1xRTT has a TOP limit of 144k bps. Once you take into account the overhead of the transmission and lots of other factors, you will probably get 50K - 110K. So how can someone claim that you will get 400K out of a pipe that CAN ONLY provide 144K? Also, the current EVDO has a limit of 2.4mb, and most people will see an average of 300-500kbps. Many are seeing greater than that, perhaps 500-1200kbps. So how is it possible to get 3000kbps? Time to get educated. Here is a benchmark for 1XRTT with a 3-bar signal strength (we ran the benchmark atDSLReports.com). As you can see, we are getting 98kbps download speed and 62kbps upload speed. This speed is about what you can expect from 1XRTT when you have a 3-bar signal.  Without moving the computer or disconnecting, we fired up and went to performance.toast.net and ran their benchmark. WOW, look at the difference! 449kbps download speed (they only benchmark the download speed)?!  How is this possible? The answer is in how tests are really measured. When Venturi Compression is turned on, every time you send a request, you are going through the Venturi Proxy Server. When doing a download test, Performance.toast.net is actually measuring how long it takes to send the large text file to your browser. When you have a high speed proxy sitting in between things, the remote benchmarking server sends the large text file to the proxy server, very very fast. Once the file has been sent to the proxy, they send their measurement graph. This type of benchmark is measuring how long it takes for the sending server to send the file to your proxy server, not to your browser! This makes for wildly varying and often inaccurate results. After getting these two conflicting reports, we decided to do a real-world test and download a 5mb file. Sure enough, we got around 100-120kbps download speeds, which would be closer to the DSLreports.com benchmark vs. the Performance.toast.net one. Here is another interesting fact. Some of the EVDO towers have a T1 coming into them. If you were sitting under the tower AND nobody else in your area was using the EVDO tower (ideal and rare conditions), the most bandwidth you would get is 1544kbps. You simply cannot get higher speeds than that! From Verizon's Web site: **Maximum possible speed varies. It declines with distance from cell site and is limited to 1.54 Mbps at certain cell sites with backhaul limitations. Number of users on the Verizon Wireless BroadbandAccess network may also affect maximum possible speed. Average upload speeds expected to be between 60 and 80 kbps. Because of the difference in upload speeds, we do not recommend using Verizon Wireless BroadbandAccess for video conferencing or other applications that require faster upload speeds. Some good advice: Verizon has a 14 day return policy. You can try EVDO in the different locations that you need access and do your own tests. Of course, the best test is doing what you normally do:
A thread was started on EVDOForums.com that discusses another benchmarking site (it is very cool). |
|
| Last Updated ( Friday, 18 August 2006 ) |
| < Prev | Next > |
|---|
Mac EVDO
