Jason Dunn
09-19-2007, 03:00 PM
<img src="http://www.digitalmediathoughts.com/images/jd-laptop-kingston-valueram-XPS-M1330.jpg" /><br /><br /><b>Product Category:</b> Laptop RAM<br /><b>Manufacturer:</b> <a href="http://www.kingston.com">Kingston</a><br /><b>Where to Buy:</b> <a href="http://www.kingston.com">Kingston</a><br /><b>Price:</b> $400 USD MSRP for 2 x 2 GB<br /><b>System Requirements:</b> A laptop compatible with DDR2/667 RAM<br /><br /><b>Pros:</b><br /><li>Easy to install;<li>Kingston RAM priced well against other brands;<li>Lifetime warranty.<br /><b>Cons:</b><br /><li>Hard to pin down a real speed boost over 2 GB in most scenarios.<br /><b>Summary:</b><br />The Dell XPS M1330 I bought earlier this month is the most powerful laptop I've ever had the pleasure of using, yet it's also among the smallest. With an Intel Core 2 Duo running at 2.2 Ghz, a 200 GB 7200 RPM hard drive, and a 128MB NVIDIA GeForce Go 8400M GS GPU, it's an all-around capable system that simply <i>screams</i> running Windows Vista Ultimate. One of my primary tasks for this laptop is RAW photo editing, so I wanted something with some serious punch. I ordered the system with 2 GB of RAM, but was curious right from the start how much better it would be with 4 GB of RAM.<!><br /><PAGEBREAK><br /><span><b>Making the Jump From 2 GB to 4 GB</b></span><br />The price of RAM upgrades from Dell has always been outrageous, so I opted to give my friends at <a href="http://www.kingston.com">Kingston</a> a call after my M1330 showed up - they sent me 4 GB worth of their <a href="http://www.ec.kingston.com/ecom/configurator_new/partsinfo.asp?root=&LinkBack=&ktcpartno=KTD-INSP6000B/2G">DDR2/667 laptop memory</a>, two sticks of 2 GB each. The RAM currently retails for $200 USD per stick, so that's $400 worth of RAM. Would it improve the performance of my XPS M1330? I've always felt that 2 GB of RAM was the "sweet spot" for Vista - would more make any difference?<br /><br /><b><span>Installing the Kingston RAM</span></b><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-001.jpg" /><br /><i>Figure 1: $400 worth of Kingston RAM. Tasty!</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-002.jpg" /><br /><i>Figure 2: Step one was to remove the back panel on the XPS M1330. This is what I saw: two Hynix-branded 1 GB modules.</i><br /><br />A word about the RAM on the XPS M1330: the entry-level configuration of the M1330 on Dell.com has 2 GB of RAM. In Canada, it's offered with 1 GB of RAM, and moving up to 2 GB costs $100 CAD extra ($96 USD). With Dell.com, it costs $375 USD extra to move from 2 GB up to 4 GB. Ordering from Dell Canada, it would have cost me $400 CAD to upgrade from 2 GB to 4 GB, which is about $375 USD. So as you can see, the upgrade price costs nearly as much as buying 4 GB of RAM from Kingston - the difference is, if you upgrade after the fact you end up with two 1 GB modules that you can sell or use in another system. It just doesn't make any sense to upgrade direct from Dell.<br /><br />The prices work out almost the same for the hard drives (which I wish I had thought about) - moving from the 160 GB, 5400 RPM drive up to a 200 GB, 7200 drive costs $200 USD. Ordering a Hitachi 7200 RPM, 200 GB hard drive from NewEgg costs $205 USD - and you end up with a spare 160 GB drive. Dell is certainly making a healthy profit on their upgrades, but the flip side is that your laptop warranty can get complicated if you've upgraded multiple components on your own.<br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-003.jpg" /><br /><i>Figure 3: After removing the default RAM, I can see how easy Dell makes it to upgrade the RAM: they've even labelled the slots.</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-004.jpg" /><br /><i>Figure 4: Snapping in the 4 GB of Kingston RAM took only a few seconds.</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-005.jpg" /><br /><i>Figure 5: Putting on the back panel of the XPS M1330.</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-006.jpg" /><br /><i>Figure 6: I was a bit worried that something went wrong with the upgrade, because I was staring at this screen for quite a while - the progress bar was going very slowly.</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-007.jpg" /><br /><i>Figure 7: I wanted to see how much RAM the system reported outside of Windows Vista, and this photo shows the answer: all 4096 GB of it!</i><br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-008.jpg" /><br /><i>Figure 8: The XPS M1330 telling me that the RAM had changed.</i><br /><PAGEBREAK><br /><b><span>Benchmarking the Results</span></b><br />First things first: as expected, Windows Vista Home Premium 32-bit didn't allow me access to all 4096 MB of the RAM. Instead, I had access to 3582 MB. While that's not 4 GB, it's much more than 2 GB, so I was still pleased to see the bump in overall available RAM.<br /><br /><img src="http://images.thoughtsmedia.com/dmt/2007/kingston-4GB-RAM-upgrade-009.png" /><br /><i>Figure 9: The actual amount of RAM under 32-bit Vista.</i><br /><br />Then I started benchmarking - and I was surprised by the results. I had hoped that Vista would be able to take advantage of the extra 1.5 GB of usable RAM and boost overall system performance, but that didn't seem to be the case. Let me explain further by going over some of the benchmark results I saw when comparing 2 GB to 4 GB - I also included tests where I had a single stick of 1 GB RAM in there, functioning in single-channel mode.<br /><br /><b>Windows Experience Index: Memory Operations Per Second</b><br />4 GB RAM: rating of 4.8<br />2 GB RAM: rating of 4.8<br />1 GB RAM: rating of 4.5<br /><br />The interpretation of this data? Having RAM running in dual-channel mode helps with the speed a bit, but if the RAM is all the same speed the results are going to be identical regardless of how much there is in total.<br /><br /><b>PC Mark 2005</b><br />4 GB RAM: 5263 PCMarks<br />2 GB RAM: 5223 PCMarks<br />1 GB RAM: 5167 PCMarks<br /><br />4 GB RAM: 4583 PCMarks Memory Only Test<br />2 GB RAM: 4566 PCMarks Memory Only Test<br />1 GB RAM: 4513 PCMarks Memory Only Test<br /><br />A very slight difference between the three - almost not worth mentioning because it's so slight. The difference between dual channel and single channel is less in this instance than I would have thought. PC Mark is a great benchmarking program because it <a href="http://www.futuremark.com/products/pcmark05/tests/">tests so many things</a>, so I'm a little surprised that there isn't a bigger difference between 2 GB and 4 GB, and shocked that the 1 GB benchmark results aren't lower.<br /><br /><b>PassMark PerformanceTest</b><br />4 GB RAM: PassMark Rating of 93.8<br />2 GB RAM: PassMark Rating of 94.6<br />1 GB RAM: PassMark Rating of 104.2<br /><br />This is a new application I've only recently started using, but it allows for a high degree of control when testing. This was one of the only applications in which I saw much of a difference, and only in one test: the "Memory - Large RAM" test. This test is described as follows: "This test measures the ability to allocate very large amounts of RAM and the time taken to read this RAM. The test is designed to measure the ability of the system to support applications that use very large amounts of RAM." Given the description of that test, it's not surprising that the 1 GB configuration scored 183.1 and the 4 GB configuration scored 816.7. The 2 GB config? 410.4.<br /><PAGEBREAK><br /><b><span>Searching for a Real-World Benefit</span></b><br />Since the benchmarks didn't show much, if any, improvement between 2 GB and 4 GB, I decided to try some real-world tests. First I tried my <a href="http://www.digitalmediathoughts.com/index.php?topic_id=12620">Nero Recode test</a>. The results? Absolutely no significant performance difference between 1 GB, 2 GB, and 4 GB. Video transcoding is almost completely CPU-based, so throwing more RAM at the problem has no impact. It's possible that with a slow 4200 RPM hard drive I might have seen a difference, but the 7200 RPM drive was fast enough to keep the RAM full and feeding the CPU.<br /><br />Next, I tried the <a href="http://www.digitalmediathoughts.com/index.php?topic_id=12620">DxO RAW photo test</a>, and the results were more telling: with 4 GB of RAM, the XPS M1330 took 5 minutes and 44 seconds to convert the 20 RAW images to JPEG. With 2 GB, it was within seconds of the same time. With 1 GB, however, it took nearly twice as long: 9 minutes and 50 seconds. It seems DxO is heavily RAM-based and 1 GB just wasn't enough for it, but 2 GB was.<br /><br />Thinking perhaps I was onto something with the RAW files, next I tried going for a much bigger data set, thinking it might show a significant difference between 2 GB and 4 GB: I took 1000 RAW images and converted them to JPEG using Adobe Lightroom 1.2. The results? The 2 GB and 4 GB configurations were within 60 seconds of each other, which isn't significant when the overall time was 26 to 27 minutes.<br /><br /><b><span>Conclusion</span></b><br />Windows Vista uses RAM much more effectively than Windows XP - it will aggressively cache frequently used programs to RAM rather than let it sit empty. This is why even when nothing is open or running, Vista will report you have almost no RAM free. <a href="http://www.codinghorror.com/blog/archives/000688.html">This is a good thing</a>, because RAM sitting there doing nothing doesn't give you any benefits. Vista can easily fill up 3582 MB of RAM - the XPS M1330 has 2709 MB cached and only 13 MB of RAM free as I write this. Benchmarking this and getting hard numbers is a much more difficult thing - I ran the PCMark benchmark over and over, hoping Vista would cache some of the program and I'd see some speed improvements, but they were very minor (and perhaps not even a result of memory caching). Hopefully someday we'll see benchmarking software that's optimized for Vista.<br /><br />While I was unable to see any significant performance difference between 2 GB and 4 GB of RAM with the types of scenarios I was running, there's one specific instance where I know it would make a difference: Virtual Machines. If you're running virtual OS software where you allocate a specific amount of RAM to the virtual machine, that RAM becomes unavailable to your main OS so you'll want every MB of RAM you can squeeze into your computer to ensure decent performance.<br /><br /><i>Jason Dunn owns and operates <a href="http://www.thoughtsmedia.com">Thoughts Media Inc.</a>, a company dedicated to creating the best in online communities. He enjoys mobile devices, digital media content creation/editing, and pretty much all technology. He lives in Calgary, Alberta, Canada with his lovely wife, and his sometimes obedient dog. He loves his XPS M1330 with 4 GB of RAM, even if it doesn't seem any faster than 2 GB.</i>