ReadyBoost背后的理论是，它有助于加快具有少量RAM 内存(RAM memory)的旧计算机。但是，没有人知道这项技术的影响有多大，因此我们决定自己进行测量并了解ReadyBoost是否具有积极影响。继续阅读以了解我们的发现：
用于测试比较的 PC(Test Comparison)
我们在这些测试中使用了三种硬件配置和两台计算机：一台装有Windows 8.1 Pro的笔记本电脑和一台装有Windows 7 Ultimate的台式机。两种操作系统都有 64 位版本的Windows。在笔记本电脑上，我们在使用其内置的 4GB RAM和移除 2GB 后进行了测量和测试，以便我们更好地评估ReadyBoost对RAM 内存(RAM memory)量较低的系统的影响。
笔记本电脑是Acer Aspire V3-551，搭载AMD 四核 A8-4500M APU(AMD Quad-Core A8-4500M APU)，集成AMD Radeon HD 7640G图形处理器，与CPU共享 512 MB 内存。这台笔记本电脑的存储是 500GB希捷(Seagate)硬盘驱动器(disk drive)，工作转速为 5400 rpm。默认情况下，这款笔记本电脑配备 4GB 的DDR3 RAM。
台式机是定制的 PC，配备AMD 四核(AMD Quad-Core) A4-3670K APU、AMD Radeon HD 6530D 显卡(video card)，还与CPU共享 512MB RAM 内存(RAM memory)。这台 PC 有 Western Digital Caviar Blue硬盘，1TB 和 7200 Rpm，外加 4GB DDR3 RAM。所有这些组件都由华擎 A75M-HVS 主板(ASRock A75M-HVS motherboard)连接。不幸的是，这台 PC 只有一个4GB 的内存模块(memory module)，我们无法像在笔记本电脑上那样移除 2GB。
操作系统是最新的，它们是大约三个月前安装的。他们还安装了我们的日常应用程序。诸如互联网(Internet)浏览器、媒体播放器等之类的东西。
我们用于ReadyBoost的闪存是HP v135w驱动器，具有 8GB空间和 USB(space and USB) 2.0，使用NTFS 文件(NTFS file)系统格式化。
我们进行的测试
我们对每台PC 和硬件配置(PC and hardware configuration)进行了一系列五项测试，以便我们尽可能准确地了解ReadyBoost对系统性能的影响。
首先，我们使用了PCMark Vantage x64基准测试。
该基准测试在一系列八个测试套件中测量总体系统性能，涵盖数字存储器、(system performance)电视和电影(TV and movies)、游戏、音乐、通信和生产力。至少在理论上，它应该准确地衡量ReadyBoost在执行诸如使用Office应用程序、浏览网页、看电影等任务时对一般系统性能的影响。(system performance)
我们用于此基准测试的设置如下：分辨率(Resolution)- 1024x768， 抗锯齿(Anti-Aliasing)- 无，纹理过滤(Texture Filtering)-最佳和 HDD 套件目标(Optimal and HDD Suite Target)- C。(C. Both) “强制全精度”("Force full precision")和“强制软件顶点着色器”("Force software vertex shaders")渲染选项均未选中，并且每个测试重复 3 次。
然后我们使用了3DMark Vantage基准测试。
大多数人都知道，这个基准测试一般DirectX 10游戏性能。我们在选择性能(Performance)预设的情况下运行了所有测试 3 次，并测量了平均结果。
接下来，我们使用BootRacer来衡量ReadyBoost是否对每个(ReadyBoost)测试系统(test system)的启动时间有任何积极影响。我们在每个系统上测量了三个连续的引导过程并计算了平均值。
此测试非常有用，因为它测量从操作系统(operating system)开始启动到您的计算机完全响应的那一刻所经过的时间，而不增加写入密码所需的时间。
我们用于测量的第四个应用程序是AppTimer。理论上，ReadyBoost可以帮助SuperFetch 系统(SuperFetch system)提高总体系统性能(system performance)。这是通过学习PC上最常用的应用程序并在系统启动(system startup)后预加载它们来实现的，以便在您启动它们时加载更快。我们仅在Acer Aspire V3-551 笔记本电脑(Acer Aspire V3-551 laptop)上进行了这些测量，因为它使用频率很高，而且我们有一个明显的趋势，即每次登录时都使用相同的应用程序。这台笔记本电脑上最常用的程序是Opera Next和Microsoft Visual Studio 2012. 在过去的几个月里，它们每天都在使用。
我们对这些应用程序进行了五次启动测试，每个操作之间有 5000 毫秒的延迟。窗口检测模式(detection mode)为Window Name /> ，Window Close Method为Alt + F4。
最后但同样重要的是，我们使用Windows 中的性能监控工具进行了测量。(Performance monitor)
我们为ReadyBoost缓存计数器设置了一个数据收集器集。(Data Collector Set)我们在PCMark Vantage x64 基准测试(PCMark Vantage x64 benchmark)运行时收集了这些信息，因为它是一个运行几个小时的完整测试，因此有足够的时间来查看ReadyBoost的运行情况。如果您想了解如何设置数据收集器集(Data Collector Set)，请查看本教程，并添加ReadyBoost缓存(Cache)计数器。此类计数器仅在您启用ReadyBoost后可用。
ReadyBoost(Impact ReadyBoost Has)对一般系统性能(General System Performance)的影响
在运行PCMark Vantage时，在配备 4GB (PCMark Vantage)RAM的笔记本电脑(Laptop)和配备4GB RAM的台式机上使用(Desktop)ReadyBoost时，分数要高 1-2% ，但令人惊讶的是，在配备 2GB RAM的(RAM)笔记本电脑(Laptop)上使用时，分数要低约 1% 。
我们还观察到，在PCMark Vantage(PCMark Vantage)执行的大多数测试中，使用ReadyBoost时没有任何改进，但有两个例外：Memories和Productivity套件。在这些测试套件中，分数每次都提高了 2%。
Memories套件测试计算机在不同情况下的性能，例如复制、组织或共享照片、视频或音乐。您可以在下面看到Memories测试套件的结果。
Productivity套件在浏览Internet和使用办公应用程序等日常任务中测量系统性能。(system performance)您可以在下面看到Productivity测试套件的结果。
当启用ReadyBoost时， (ReadyBoost)PCMark Vantage中的游戏(Gaming)和通信(Communication)测试记录的分数持续降低 1% 。
游戏套件测试从连接游戏控制器(game controller)到玩在线游戏的游戏场景中的(Gaming)CPU、GPU 和 HDD 性能。(GPU and HDD performance)您可以在下面看到游戏(Gaming)测试套件的结果。
通信(Communication)套件测试计算机在发送电子邮件或通过Internet共享信息等情况下的性能。您可以在下面看到通信(Communication)套件测试的结果。
ReadyBoost(Impact ReadyBoost Has)对游戏性能(Gaming Performance)的影响
在运行3DMark Vantage时，结果表明ReadyBoost对游戏性能没有影响。
此基准测试在运行游戏时测量CPU 和 GPU 性能(CPU and GPU performance)，并且在启用ReadyBoost时分数没有显示任何明显差异。
测量的最大差异为 0,002%，这在统计上是微不足道的。
ReadyBoost(Impact ReadyBoost Has)对Windows 启动时间(Windows Boot Timings)的影响
BootRacer表明ReadyBoost改进了操作系统(operating system)的启动时间。在 4GB 的系统上，改进时间仅为 1 秒，而在 2GB 的系统上，改进时间为 4 秒（% faster）。
我们可以有把握地说，在RAM(RAM)较少的系统上，您会注意到启动时间的显着改善。
ReadyBoost(Impact ReadyBoost Has)对应用程序启动时间(Application Startup Time)的影响
在操作系统(operating system)启动后立即启动应用程序意味着与打开系统 15(application 15)分钟后打开相同的应用程序相比，启动速度较慢。这是因为您一登录Windows就会启动许多服务和后台进程。
提醒一下，这些测量仅在具有 4GB RAM和 2GB RAM的(RAM)Acer Aspire V3-551 笔记本电脑(Acer Aspire V3-551 laptop)上进行，因此您可以轻松比较ReadyBoost在可用(ReadyBoost)RAM很少和当您使用有相当数量的内存。
当您打开一个您经常使用的程序时， ReadyBoost(ReadyBoost)做得很好，一旦您登录。它将应用程序的首次启动时间提高了 10-15%。
如果你关闭它然后再次启动它，ReadyBoost仍然会改进，但幅度较小。当您在登录 PC 后立即启动常用应用程序时，它的影响最为明显。
ReadyBoost如何使用闪存驱动器(Flash Drive)来提高性能(Performance)？
我们在性能监视器中使用了(Performance Monitor)数据收集器集(Data Collector Set)，以便了解ReadyBoost如何使用我们插入的 HP 闪存以提高性能。我们查看了以下指标：
-
MB Cached - ReadyBoost从主 RAM 内存中复制了多少数据到记忆棒上，以加快速度。
-
已使用缓存- (Cache Used)ReadyBoost在运行 PC Vantage 时使用了多少先前缓存的内存，以提高性能。它使用的缓存内存越多越好。
-
Skipped Reads/Second - ReadyBoost从闪存驱动器读取数据而不是从硬盘驱动器读取数据的次数。更频繁地从闪存驱动器读取意味着更快的读取和更好的性能。
测量是在Acer Aspire V3-551 笔记本电脑(Acer Aspire V3-551 laptop)上进行的，当时它有 4GB 的RAM，然后是 2GB 的RAM，因此我们可以看到ReadyBoost对具有不同内存量的同一系统的影响。
结果非常相关，每个计数器都证明，当笔记本电脑有 2GB内存(RAM)时，ReadyBoost使用闪存的次数至少是笔记本电脑有 4GB(ReadyBoost)内存(RAM)时的两倍。我们可以看到，ReadyBoost在笔记本电脑只有 2GB内存(RAM)时缓存了两倍的总数据量，使用了两倍的闪存存储容量，并避免使用硬盘(storage capacity)驱动器(disk drive)从闪存读取数据来读取数据6(memory 6)比它拥有 4GB RAM时的性能高出几倍。
上图您可以很容易地注意到ReadyBoost 对内存量(ReadyBoost)较低的系统的影响越来越大。
摘要(Summary)- ReadyBoost 提供什么？
回顾一下，这里总结了在RAM 较少的系统上启用ReadyBoost时将获得的改进：(ReadyBoost)
- 打开照片、音乐或视频(music or video)等媒体文件的速度稍快（约 2%）。
- 网页加载和Office应用程序的使用稍快（约 2%）。
- 您的系统的启动时间得到改善（高达 7%）。
- 您最常用的应用程序启动速度更快（10% 到 15%）。
当您玩游戏或运行CPU 或 GPU(CPU or GPU)密集型应用程序时，您不会注意到启用ReadyBoost有任何区别。(ReadyBoost)当您添加更多RAM时，改进也会变得更小。在具有超过 4GB RAM的系统上，您不会注意到任何改进。ReadyBoost在(ReadyBoost)RAM为 2GB或更低的系统上最为有效。此外，将SSD 驱动器添加到您的计算机意味着使用(SSD drive)ReadyBoost毫无意义。添加SSD后，我们进行了一些非常快速的测试，我们很快注意到ReadyBoost变得无关紧要。
Does ReadyBoost Work? Does It Improve Performance for Slower PCs?
The theory behind ReadyBoost is that it helps speed up older computers with low amounts of RAM memory. However, nobody knows how much of an impact this technology has, so we decided to make our own measurements and learn whether ReadyBoost has a positive impact or not. Read on to learn what we have found:
The PCs Used for the Test Comparison
We used three hardware configurations for these tests and two computers: a laptop with Windows 8.1 Pro and a desktop with Windows 7 Ultimate. Both operating systems had 64-bit editions of Windows. On the laptop, we have made measurements and tests while using its built-in 4GB of RAM and then after removing 2GB so that we can better evaluate the impact ReadyBoost has on systems with low amounts of RAM memory.
The laptop is an Acer Aspire V3-551, powered by AMD Quad-Core A8-4500M APU, with an integrated AMD Radeon HD 7640G graphic processor, which shares 512 MB of memory with the CPU. The storage for this laptop is a 500GB Seagate hard disk drive, working at 5400 rpm. By default this laptop comes with 4GB of DDR3 RAM.
The desktop is a custom built PC with an AMD Quad-Core A4-3670K APU, an AMD Radeon HD 6530D video card that also shares 512MB of RAM memory with the CPU. This PC has Western Digital Caviar Blue hard disk, with 1TB and 7200 Rpm, plus 4GB of DDR3 RAM. All these components are connected by an ASRock A75M-HVS motherboard. Unfortunately this PC has only a memory module with 4GB and we could not remove 2GB like we did on the laptop.
The operating systems are up to date and they were installed about three months ago. They also have our daily applications installed. Things like Internet browsers, media players, etc.
The flash memory we used for ReadyBoost is an HP v135w drive with 8GB of space and USB 2.0, formated using the NTFS file system.
The Tests We Have Performed
We ran a series of five tests on each PC and hardware configuration, so that we get an overview that's as accurate as possible about the impact of ReadyBoost on system performance.
First, we used the PCMark Vantage x64 benchmark.
This benchmark measures general system performance in a series of eight test suites covering digital memories, TV and movies, gaming, music, communications and productivity. At least in theory, it should accurately measure the impact ReadyBoost has on general system performance while doing tasks like using Office applications, browsing the web, watching movies, etc.
The settings we used for this benchmark are the following: Resolution - 1024x768, Anti-Aliasing - None, Texture Filtering - Optimal and HDD Suite Target - C. Both "Force full precision" and "Force software vertex shaders" rendering options were unchecked and each test was repeated 3 times.
Then we used the 3DMark Vantage benchmark.
As most of you know, this benchmark measures general DirectX 10 gaming performance. We ran all its tests three times, with the Performance preset selected, and measured the average results.
Next, we used BootRacer to measure if ReadyBoost has any positive impact on the boot times of each test system. We measured three consecutive booting procedures on each system and calculated the average.
This test is very useful, because it measures the time elapsed since the operating system began booting to the moment your computer is fully responsive, without adding the time it takes to write your password.
The fourth application we used for our measurements is AppTimer. In theory, ReadyBoost helps the SuperFetch system improve general system performance. This is achieved by learning the applications that are most used on your PC and preloading them after the system startup, so that they load faster when you start them. We performed these measurements only on the Acer Aspire V3-551 laptop because it is used very often and we have a clear trend of using the same applications each time we sign in on it. The most used programs on this laptop are Opera Next and Microsoft Visual Studio 2012. They have been used every day for the last couple of months.
We ran five startup tests with these applications, with a delay of 5000 milliseconds between each action. The window detection mode was the Window Name/> and the Window Close Method was Alt + F4.
Last but not least, we made measurements using the Performance monitor tool in Windows.
We have set up a Data Collector Set for the ReadyBoost cache counters. We gathered the information while the PCMark Vantage x64 benchmark was running, because it's a complete test that runs for a few hours, so there is enough time to see ReadyBoost in action. If you want to see how to setup a Data Collector Set, please check this tutorial , and add the ReadyBoost Cache counters. This category of counters is available only after you enable ReadyBoost.
The Impact ReadyBoost Has on General System Performance
While running PCMark Vantage, the scores were 1-2% higher when using ReadyBoost on the Laptop with 4GB of RAM and the Desktop with 4GB of RAM, but surprisingly, it was about 1% lower on the Laptop with 2GB of RAM.
We also observed that in most of the tests performed by PCMark Vantage there were no improvements when using ReadyBoost, with two exceptions: the Memories and Productivity suites. In these suites of tests, the scores were improved every time by up to 2%.
The Memories suite tests a computer's performance in different situations, like copying, organizing or sharing photos, videos or music. Below you can see the results for the Memories suite of tests.
The Productivity suite measures system performance in usual tasks like browsing the Internet and using office applications. Below you can see the results for the Productivity suite of tests.
The Gaming and Communication tests in PCMark Vantage recorded scores that were constantly 1% lower when ReadyBoost was enabled.
The Gaming suite tests CPU, GPU and HDD performance in gaming scenarios from attaching a game controller to playing an online game. Below you can see the results for the Gaming suite of tests.
The Communication suite tests computer's performance in situations like sending an e-mail or sharing information over the Internet. Below you can see the results for the Communication suite of tests.
The Impact ReadyBoost Has on Gaming Performance
While running 3DMark Vantage the results have shown that ReadyBoost has no effect on gaming performance.
This benchmark measures CPU and GPU performance while running games and the scores did not show any noticeable difference when ReadyBoost was enabled.
The biggest difference that was measured was of 0,002% which is statistically insignificant.
The Impact ReadyBoost Has on Windows Boot Timings
BootRacer has shown that ReadyBoost improves the boot timings of the operating system. On the systems with 4GB the improvement was of only 1 second while on the system with 2GB it was 4 seconds (7% faster).
We can safely say that on systems with low amounts of RAM, you will notice a decent improvement in boot timings.
The Impact ReadyBoost Has on Application Startup Time
Starting an application immediately after the operating system has booted means a slower startup when compared with opening the same application 15 minutes after turning on the system. That's because lots of services and background processes are launched as soon as you log into Windows.
As a reminder, these measurements were performed only on the Acer Aspire V3-551 laptop when it had 4GB of RAM and when it had 2GB of RAM, so that you can easily compare the impact ReadyBoost has when you have little RAM available and when you have a decent amount of memory.
ReadyBoost does a great job when you open a program that you use very often, as soon as you log in. It improves the first startup time of an application by 10-15%.
If you close it and then launch it again, ReadyBoost will still improve things but by a smaller margin. Its impact is best noticed when you start frequently used applications as soon as you log in to your PC.
How Does ReadyBoost Use a Flash Drive to Improve Performance?
We used a Data Collector Set in Performance Monitor so that we get a view of how ReadyBoost uses the HP flash memory we plugged in order to improve performance. We looked at the following indicators:
-
MB Cached - how much data did ReadyBoost copy on the memory stick from the main RAM memory, to speed things up.
-
Cache Used - how much of the previously cached memory was used by ReadyBoost while running PC Vantage, in order to improve performance. The more cached memory it uses, the better.
-
Skipped Reads/Second - how many times did ReadyBoost read data from the flash drive instead of reading from the hard drive. Reading from the flash drive more often means faster reads and better performance.
The measurements were made on the Acer Aspire V3-551 laptop when it had 4GB of RAM and then when it had 2GB of RAM so that we could see the impact ReadyBoost had on the same system with different amounts of memory.
The results are pretty relevant and every single counter proved that when the laptop had 2GB of RAM, ReadyBoost used the flash memory at least twice as much as when the laptop had 4GB of RAM. We can see that ReadyBoost cached double the amount of total data on when the laptop had only 2GB of RAM, used double the storage capacity on the flash memory, and avoided using the hard disk drive to read data by reading it from the flash memory 6 times more than it did when it had 4GB of RAM.
The the graph above you can easily notice that the impact ReadyBoost has grows on systems with lower amounts of memory.
Summary - What Does ReadyBoost Deliver?
To recap, here is a summary of the improvements you will get when enabling ReadyBoost on a system with low amounts of RAM:
- Opening media files like photos, music or video is slightly faster (approximately by 2%).
- The loading of web pages and the use of Office applications is slightly faster (approximately by 2%).
- Your system's boot timings are improved (up to 7%).
- Your most used applications start faster (by 10 to 15%).
You will not notice any difference with ReadyBoost enabled when you play games or you run applications that are CPU or GPU intensive. Also the improvements get smaller when you add more RAM. On a system with more than 4GB of RAM you will not notice anything getting better. ReadyBoost is most effective on systems with 2GB of RAM or lower. Also, adding an SSD drive to your computer means that there is no point in using ReadyBoost. We've done some very quick tests after adding an SSD and we quickly noticed that ReadyBoost becomes irrelevant.
