SilverStone FT03: Nothing Else Like It
by Dustin Sklavos on April 28, 2011 12:25 AM EST- Posted in
- Cases/Cooling/PSUs
- SilverStone
Testing Methodology
Part of the reason we took so long in getting new case reviews up on the site is because we decided to go back and revise our case testing procedures. Moving forward we're standardizing our testing methodology in a way that will hopefully feel more concrete and useful than our previous reviews did. I'm going to take the opportunity to break down how things are going to work for this and future reviews, but note that these procedures aren't set in stone. Anand stressed to me in conversation that one of our greatest assets is essentially the mountain of free feedback we get from people who comment on our reviews, and I agree with him. We pride ourselves on being fairly agile and adaptable to the needs of our readership, so if you feel there's something we can change and it's feasible, rest assured that we do listen and will at the very least take it into consideration.
First, we've standardized on two testbeds: a Mini-ITX testbed that's low-powered and doesn't generate too much heat, but isn't too miserly either, and a full ATX testbed that's indicative of the high-end of what most of you are running. Since this review uses our Mini-ITX bed, I'll save discussing the full ATX one for that review. The essential thinking here is that a Mini-ITX board can be used for Micro-ATX builds like this one without sacrificing much (we're not testing multi-GPU scenarios for cases this small since they're seldom designed for it), but we still use a processor that produces enough heat to merit thermal testing. Before I go any further, it may be prudent to just lay out the hardware we're testing with:
| Mini-ITX/Micro-ATX Test Configuration | |
| CPU | Intel Core i3-530 (73W TDP) |
| Motherboard | Zotac H55ITX-WiFi |
| Graphics Card |
Intel HD Graphics (IGP) Zotac NVIDIA GeForce GTX 580 (244W TDP) |
| Memory | 2x2GB Crucial Ballistix Smart Tracer DDR3-1600 |
| Drives |
Slimline DVD+/-RW Drive Kingston SSDNow V+ 100 64GB SSD Western Digital Caviar Black 1TB SATA 6Gbps |
| CPU Cooler | Zalman CNPS8000A with Cooler Master ThermalFusion 400 |
| Power Supply | SilverStone Strider Gold 750W 80 Plus Gold |
The Zotac motherboard is included as one of the most fully-featured Mini-ITX boards available, and it works with the Core i3-530 we had on hand. That i3-530 is also fairly typical of what to expect in smaller builds. From there, the Crucial Ballistix Smart Tracer memory has a unique feature particularly suited to our needs: it includes thermal diodes that can tell you how hot the memory is running. The SSD is used to gauge how easily a 2.5" drive is installed and to provide a stable baseline, while the Caviar Black gauges how easily a 3.5" drive can be installed and gives us a secondary reference point for drive temperatures. The Zalman cooler is quiet, low-profile, and effective, giving us an alternative to the Intel stock cooler. And finally, SilverStone's power supply is modular, small, and efficient, making it ideal for case testing.
Of course, the GeForce GTX 580 is going to raise some eyebrows. In this instance, we wanted a video card for case testing that would push thermals without overpowering acoustics. Not all cases we test with are going to support it, but enough are to warrant its inclusion. As a result, Mini-ITX and Micro-ATX systems are tested in two configurations: with and without dedicated graphics.
Moving on, acoustic testing is standardized on a foot from the front of the case, using the Extech SL10 with an ambient noise floor of ~32dB. For reference, that's a silent apartment with nothing running, testing acoustics in the dead of night (usually between 1am and 3am). A lot of us sit about a foot away from our computers, so this should be a fairly accurate representation of the kind of noise the case generates, and it's close enough to get noise levels that should register above ambient.
Finally, thermal testing is run with the computer having idled at the desktop for fifteen minutes, and again with the computer running both Furmark (where applicable) and Prime95 (less one thread when a GPU is being used) for fifteen minutes. I've found that leaving one thread open in Prime95 allows the processor to heat up enough while making sure Furmark isn't CPU-limited. We're using the thermal diodes included with the hardware to keep everything standardized, and ambient testing temperature is always between 71F and 74F. Processor temperatures reported are the average of the CPU cores.
Thank You!
We do have some thanks in order before we press on, though:
- Thank you to Crucial for providing us with the Ballistix Smart Tracer memory we used to add memory thermals to our testing.
- Thank you to Zalman for providing us with the CNPS8000A heatsink and fan unit we used.
- Thank you to Kingston for providing us with the SSDNow V+ 100 SSD.
- Thank you to CyberPower for providing us with the Western Digital Caviar Black hard drive.
- And thank you to SilverStone for providing us with the optical drive and power supply.
Facebook
Twitter
RSS
81 Comments
View All Comments
Rick83 - Thursday, April 28, 2011 - link
Well, that's why I recently removed all mechanical disks from my tower and into the network. One SSD is all a system really needs, and I don't want spinning disks in my proximity. Of course, I was running ancient 80G and 200G disks which were proper noisy, so that may have been traumatizing. But in general, I don't want storage on my local machine, there's just no point, and to get where I'd want, I'd need a huuuuuge tower. This way, my disks reside in a nice Stacker, well cooled and as much out of earshot as possible, and my desktop system is reduced to a single 2.5" SSD that makes zero noise and has no trouble with getting a bit warmed up, as there are no mechanical tolerances that are impacted. I strongly recommend that approach for anyone who has more than one hard disk that they use...Also, with 3HDDs, why bother paying extra for less and getting a micro-ATX board? Most cases that fit 3 HDDs are so big, that they don't really take advantage of the micro-ATX form factor. (There are some, like the bigger of the Lian-Li cubes, and your P182, but these are already quite "big" cases...Going to a midi-tower is only about 3 cms of height away...
Spoelie - Thursday, April 28, 2011 - link
Most NAS solutions have less than ideal performance (when moving >10GB files around) and less than ideal noise characteristics (high pitched 40mm fans, really??). My current flat is not that big so there is no out-of-earshot location ;). But to be honest, haven't explored that option very well.Instead I have one main, relatively silent desktop as the gaming/htpc/file/printer server, and smartphone/tablet/notebook "clients" for leisure computing. All 3 drives are 1TB WD Caviar Greens (EADS) though, and are a hell of a lot quieter than my older 250-500GB drives. In the current mounting mechanism, inaudible.
mATX boards provide everything+kitchen sink nowadays, ATX form factor is really relegated to multigpu and some htpc configurations with specialized addin boards. The premium is not that much IMO, just a tad over a $100 is what I paid.
I don't really have a genuine need for another case, but am always looking for improvement - and I have to say, something less heavy and less bulky with top mounted connections sounds mighty appealing. The P182 is a real back breaker to tug along. I just don't want to give up too much silence/cooling/mounting capacity.
Rasterman - Thursday, April 28, 2011 - link
I did that too with my latest build, SSD only, NAS for storage. It works fine, but anytime I need to move large files its slow 15-25MB/s. I have been reading the green drives are very quiet and am considering one, but desktop is about as silent as it can get though (3 nexus fans at 500rpm), and ANY noise would be audible. The other thought I had was using a usb3 netbook for a NAS, that should provide much better performance than my synology and be: expandable, easily configurable, can host anything, while not using hardly any power, plus it has built in battery backup.rabidsquirrel - Thursday, April 28, 2011 - link
I think I just threw up a little in my mouth...Seriously, that thing is hidious. Hopefully its thermal properties make up for that for some users.
kevith - Thursday, April 28, 2011 - link
Ehm, do You seriously mean, that 45 dB of noise in an enclosure without a discrete GFX is "not obnoxious" or "dealbreaking.I´m a musician, 45 dB is LOUD, man!
Spivonious - Thursday, April 28, 2011 - link
Thanks for including more useful noise info. 45dBA is loud.Was there any explanation for the odd tilted fan mounts? You allude to Silverstone's exhaustive airflow engineering but never go into details.
Rick83 - Thursday, April 28, 2011 - link
45 db at one foot. Sound diminishes by the second power with distance.Commonly such things are measured at one meter distance. One foot is pretty close, much too close in most desktop scenarios, unless you have the box sitting right on the edge of the desk, and are bent forwards. Tripling distance leads to about a factor 1/9 for the sound attenuation, which results in a decrease of almost 10 decibel. (unless my physics are rustier than I thought they were)
Additionally, no mention was made whether certain directions are noisier than others. I'd expect most of the noise to come out of the top, due to this case's design.
Spivonious - Thursday, April 28, 2011 - link
I'm not sure of the rules for distance and attenuation, but it would mean that at 3 feet the noise level would be 40dBA (8/9 of 45dBA). Still pretty loud.I would love to see what Antec would do with this design. They could probably get it down to 35dBA with no impact on cooling.
Spivonious - Thursday, April 28, 2011 - link
Looked it up http://www.csgnetwork.com/decibeldropdistancecalc....You were right that the attenuation would be around 10dBA (taking it to 35dBA). Much quieter and easily acceptable with a noise floor of 32dBA.
Voldenuit - Thursday, April 28, 2011 - link
Sound intensity decreases proportional to 1/r^2, but sound *pressure* (which our ears are directly sensitive to) decrease proportionally to 1/r. Most sound level meters measure sound pressure level (SPL) and it is the most common way of reporting sound loudness on tech sites.I've owned several Silverstone cases - a TJ08 and a KL01, and neither have been particularly silent or even quiet. Even swapping out the crappy stock Silverstone fans for 800 rpm Scythe Slipstreams didn't help much with the TJ08 - the KL01 unfortunately used a proprietary connector for its front panel fan, so I was stuck with its obnoxious sound volume and characteristics.
If you want *quiet*, Antec P183s and Fractal Design R3s can be made to hover around 14-20dB at 1m, and Puget Systems' custom builds live around 11-14 dB.
I love Silverstone's design and build (for the most part - the FT03 looks like a trashcan to me), but outside of a few exceptions like the FT02 and RV02, they are not exactly silent cases out of the box.