Acer 1410 SU2300 Benchmarks with RAM & HDD Upgrades [surprising results!]

After getting my new 1410, I was interested in finding out just how much a hard drive and RAM upgrade would help. I have a 1410 SU2300 which came with 2 1GB DDR2 Dual Channel 667 Mhz 5-5-5-15 RAM modules and a 160GB Western Digital Scorpio Blue. I made a list of fairly well known benchmarks, and ran them with various configurations using upgraded RAM (2 2GB DDR2 Dual Channel 800 Mhz 6-6-6-18 Corsair) and HDD (320GB Seagate Momentus 7200.4).

For testing, I wiped the drive, installed Windows 7 x64 Ultimate, and applied all the tweaks in the 1410 thread, including installing BatteryBar, RMClock, and a few programs of my own like Office 2007 and Dropbox along with the latest drivers for every component and all of the latest MS updates. Then, I installed the following programs for benchmarking:

PCMark Vantage x64 Professional - Overall performance indicator, provides scores for various areas of computer usage. Instead of running synthetic benchmarks, it runs scripts using standard Windows programs like Media Player and Internet explorer. Accurate and popular with repeatable results; probably my favorite benchmark.
PCMark Vantage x86 Professional - Same as x64 version, shown for 32 bit comparison – can only address 3.5gb of the 4.0 installed
Passmark Performance Test 7 – Synthetic overall benchmark, could not be completely run due to directX issues. Not very repeatable or accurate, I personally dislike it.
IOMeter – Hard drive benchmark, sends repeated IO requests and measures operations completed per second in various configurations. Reasonably repeatable, very accurate, much better at providing an overall performance indicator then synthetic sequential read\write HDD benchmarks.
Everest Ultimate – Memory, CPU, and HDD benchmarking options. Not very repeatable or reliable.
Geekbench x86 – Memory & CPU benchmark. Somewhat repeatable, fairly inaccurate.
Geekbench x64 - Same as x86 version, allows for comparison between x86\x64.
Crystal Disk Mark – HDD benchmarking program using various file sizes. Results given in MB\s. Very accurate and repeatable, my favorite sequential read\write testing program. Used 9 pass 1GB file settings.
3dMark 06 (3dMark Vantage couldn’t run) – 3d graphics\CPU benchmarking program. Repeatable, reliable, accepted standard.
HD Tune Pro – HDD benchmark, requires a non-window partition for write tests. Fairly repeatable and reliable, another ‘staple’ benchmark.
DOOM 3 - OpenGL graphics real world benchmark
wPrime – CPU benchmark. 1024M results accurate and repeatable. Staple benchmark. Results given in seconds.
Windows 7 Performance SDK – Used a customized script to benchmark bootup performance, accurate to the millisecond.
ATTO – HDD benchmarking program, staple benchmark. Sequential read\write only, accurate and repeatable.
HDTach – HDD benchmarking program, somewhat accurate and repeatable.
AS SSD – HDD benchmarking program. Designed for SSDs, useful for HDDs. Very accurate and repeatable.

After installing and configuring all of them, I created a disk image to use for later setups. Then, I ran all the benchmarks and recorded the results. I used 5 different setups; stock, stock + RAM upgrade, stock + HDD upgrade, stock + RAM + HDD upgrade, and stock + 1 2GB stick of RAM + HDD (Single channel due to one stick, similar to the SU3500 setup). All results were recorded in a table, with a column denoting the improvement or decline compared to the stock model in percentages.

The utmost pains were taken to insure accurate results. Most benchmarks were run twice and averaged; several that I deemed important, such as PCMark Vantage, were run 5 times then averaged to insure accuracy. Any time results seemed unusual or out of line, several more runs were recorded and averaged. The exact same configuration was used for all setups, down to the number of icons on the screen.

And finally, the results. For the searchable text, click here for the Google docs version.
Click here for the full color table version.

Startup times used two values. Both were measured starting at the time the HDD was initialized for bootup after POST. One was to finish loading Explorer.exe, at this point the computer was logged in with the taskbar and icons fully loaded, but it took 3-4 seconds afterwords for the hard drive to calm down and background programs to finish loading. The explorer benchmark can generally be compared with timed computer startups. All startup programs was the timing for a complete finish of all startup tasks, including delayed ones. HDD activity ceased ~15 seconds before this time. Times given in milliseconds. Text version included in main Google Docs spreadsheet.
Click here for the Startup time (10 trials + average) table.

Conclusion: RAM
The memory disappointed me. It showed no significant improvement in most relevant benchmarks. Even the 667-800 Mhz bump did little to improve scores, even in 3d\gaming. This may have been offset by the increase in latency from 5-5-5-15 to 6-6-6-18. However, two significant factors were not counted in the benchmarks. When multitasking with many programs that exceed 2GB, the extra memory can prevent swap file usage. More importantly, windows 7 uses spare memory to cache frequently used programs, which can improve overall speed. Memory was purchased for $73 new, old memory was resold for $30 after shipping\fees.

Total cost: $43. Verdict: Not really worth it. Computer didn’t seem noticeably faster, even when running several programs at once, and the extra 2GB didn’t seem to improve caching either. At most tasks it seemed as though the slow hard drive and SU2300 were holding it back much more then the RAM. Even for 1x2GB owners, the dual channel really doesn’t seem to do much – all of my reporting tools claimed it was enabled, but it just didn’t really show in the benchmarks.

Conclusion: HDD
The hard disk, on the other hand, astounded me. It dramatically increased HDD benchmark results by nearly 50%, and showed a 5% increase in overall performance with a far greater increase of 25-100%+ in application loading and caching times. It decreased boot times and overall made the system feel considerably more snappy; it was a big, noticeable difference compared to the stock drive. Doubling the capacity also helped propel it to ‘must-have’ upgrade status. HDD was purchased for $56 new, old HDD was resold for $32 after shipping\fees.

Total cost: $24. Verdict: Totally and completely worth it. Made system feel and benchmark much faster, doubled storage capacity. Bonus: Comes with a 3 year warranty, triple what the stock HDD has.

In conclusion, the results were surprising. As a big proponent of extra RAM, I expected the memory to sweep the benchmarks. But it seems that Windows 7 deals with 2GB just fine. The HDD did better than even I expected – if I didn’t need to store a ton of files, I would swap out one of my 30GB Vertex SSDs from my desktop for an even greater boost. For anyone considering an upgrade, go for it – for an amazingly low price of $24, it’s a great bang for the buck.

Overall, the laptop now runs impressively fast and smoothly. I’ve spent a lot of time with it – each set of benchmarks took well over 20 hours; 5 runs of PCMark Vantage took over 8 hours alone, and it was hard to resist using it while it was so new. But a week of heavy benchmarking later, I’m glad I recorded the results to share – it should at least provide a baseline for potential 1410 buyers, as I had a hard time finding info on it when doing my research. For $360 for the SU2300 model + $24 for the HDD, you get an amazingly capable 11.6” ultraportable with a great display , 4.5 hour battery life, a nice keyboard and touchpad, a good design, plenty of features like a webcam and gigabit Ethernet, and performance that easily triples even the latest Atom netbooks. And for what it's worth, I typed this review and formatted the benchmarking table while on it.

Thanks for reading this, please take the time to comment or ask questions if you found this information helpful.

 

..and for an extra $300, you can get an X-25M for sweet performance and 5800mAh battery for 9 hours of run time!

(That's my tip)

 

Ehh, I tried one of my OCZ Vertexes in it (they are very similar to the X-25M; higher write speeds but slightly lower IOPS). I didn't run the full benchmark suite as I needed the drive for my desktop, but the improvement in overall performance over the 7200.4 was about the same as from stock to 7200.4. Hard disk benchmarks were through the roof, but the 7200.4 is one of the fastest laptop hard drives, and effectively reduces hard disk bottlenecks. I don't find that paying nearly 50% more for 4-5% performance increase with 1/4th the capacity is worth it, but those who value speed above all else might.

I thought about the battery, but the long shipping times and problems with the chinese knockoffs made me decide to keep mine. A nice investment for those who need the battery life, though.

 

SSD gains aren't quantified in benchmarks, it's often how your operating system feels. But you're familiar with that...

 

Running 3 of them in RAID0 does tend to give one a feel for them, though I usually only use 2 now

I was referring specifically to PCMark Vantage, which does a good job of quantifying the performance gains - in a desktop with a nice processor, they boost performance by a considerable margin. But the SU2300, though good for a netbook\ultraportable, is the bottleneck in most cases, even in application loading. That's not to say it feels slow; the old drive was certainly the bottleneck before, but the far superior performance of the 7200.4 puts it ahead of the SU2300.

 

where are you finding the 2300 model for $360?

 

Scouted ebay a few days, picked a cheap target, PMed asking for a few bucks off, bought with Bing Cashback

 

true, true, that'll do it.

 

Amazing deals with 10% cashback. You can buy new and resell a month or two later with no loss or even a little gain. Great for people who always want the the latest thing but are cheapskates

Anyway, I have discovered the secret to smooth 1080p video playback. I started up by loading a fast action scene from the matrix, played great with no audio\video problems but CPU was pushing 90% - I'm guessing a long smooth pan might have had a dropped frame or two. Plus, skipping around took a lot longer then it does on my desktop.

Follow these steps:

#1: Install MPC x64 ( http://sourceforge.net/projects/mpc.../MPC-HomeCinema.1.3.1249.0.(x64).exe/download)
#2: Install the (ancient, year old beta) x64 version of haali media splitter ( http://haali.net/mkv/mkx.y.9.exe)
#3: Install the registry hack to enable hardware acceleration on almost all MKV files ( http://www.bigandfree.com/11384042)
#4: Just kidding, there is no number 4. Download and run 3 executable files and you'll go from a stock windows 7 x64 install to smooth 1080p playback, intel drivers not withstanding.

The most action-packed scene I could find hit a record 36% CPU usage, most action was 15-25, spread perfectly over both cores. Pausing and skipping were far faster. It should have no problems playing smoothly on a SU3500 either. Not bad for 1080p on a netbook.

 

Just for the fun of it, I gave AS SSD a run to see how the Vertexes did. Yea, SSDs are the future. Nuff said.



Ladies and gentleman, the SSDs achieved a performance over 119 times greater then a fast spinning HDD of the same size on the 4-64k benchmark, an important meter of performance. Sadly, the new HDD doesn't do that well on that particular benchmark, but makes up for it in other areas.

 

Hi josh... thanks for sharing your results of extensive testing with us...

I have a question for you...?

I'm planning on getting a 1810t with 320gb hdd... !

So if i remove the hdd from it and replace it with "(320GB Seagate Momentus 7200.4)." that you put into your lappy...

1. Will i see huge difference in everyday tasks...??? & Can u put in words what kind of difference will i see if any...???

2. Will the 7200rpm hdd be more noisy than the Stock 5400 rpm hdd...???

3. Will putting in the 7200 rpm hdd reduce battery life...???

Thanks

 

Have you tried YouTube/Hulu HD? Is it possible to get smooth playback (high frame rate) out of those (especially the 1080p versions)? Adobe just released a new beta of Flash Player 10.1 (beta 3) and while it seems to be mainly an update for GMA 500 and Broadcom Crystal HD users I'm curious if it offers any improvements for the GMA 4500MHD.

 

Rana,

Program launching and other HHD dependent tasks will run roughly 50% faster, translating into 30-40% real world performance boost while launching programs. Pretty much any time your waiting for computer while the HDD light is flickering will happen a lot less frequently.

The 7200.4 is one of the quieter 7200 rpm drives. It has a background whirr not unlike the fan on low/medium, and seeks are slightly louder, but it's only noticable (and not disturbing) in a quiet room.

Battery life will only decrease slightly, around 3-5%.

I haven't tried the new Flash player yet, I'll try it in a few hours. The 10.0 version could play 720p YouTube/Hulu HD fullscreen without issues, but choked on high-action 1080p YouTube scenes.

Sorry for the short post, I had to type it on my phone.

 

I found that Beta 3 decreased performance for the 4500MHD. Beta 1 seems to work much better. Watching the Dark Knight trailer wasn't a perfect 24fps (more like 15-24), it was never really choppy. Beta 3 made explosions and high action choppy.

 

hey josh .... i got a couple of questions....

If you take out the stock hdd and put in a "new" hdd of your own then...

1. Does it void the warranty of the laptop since you opened it..??

2. If there is any problem with any other part of the laptop, say maybe.... keyboard, screen, trackpad etc and you have to send it back to acer wont your waaranty be void...???

 

Shouldn't be a problem. They can't even tell if you opened it - there's no tracing paint on the screws or other such markers. In fact, most RMAs require that you remove the HDD before sending it in so it isn't damaged during shipment and there arn't any issues with data. It's similar to RAM - an easy and common upgrade that they don't mind as it's hard to do wrong. If the problem is obviously caused by a bad replacement (torn cable, screwdriver marks on the motherboard or other obvious physical damage) you may have problems, but I've RMAed faulty laptops with non-stock RAM etc. with no problems.

And again, like I said, you generally remove the HDD before shipping it for repair - so there's virtually no chance you'll have issues with it.

 

Hi josh..

thanks for the swift reply... though i got a few more questions for ya..

1. Why did you pick this particular 7200rpm hdd...??

"320GB Seagate Momentus 7200.4)"

Where did you buy it from...??

Is it the best 7200 rpm hdd...? Did u read some reviews about it comparing to other 7200 rpm hdds before selecting that particular one. i dont remember but i read somewhere that there were Reliability Problems with Seagate hdds and that lot of them fail. I'm under the impression that WD are the best and most relaible. Would you say that would be true...?

2. Why did you buy the Acer with the "SU2300 model"...??

How much better would be a SU7300 or a SU9400...??

I ask as i'm considering these...


http://www.acer.ca/acer/productv.do...ctx1g.c2att92=474&ctx1.att21k=1&CRC=505027361

http://www.futureshop.ca/Search/Sea...re%20Site&pcname=MCCPCatalog&AllowSecure=True

It has there chioces... all hdds are 5400 rpm.

1. SU4100/250hdd/2gb Canadian $ 549
2. SU7300/320hdd/4gb Canadian $ 649
3. SU9400/400hdd/4gb Canadian $ 749

So i'm now confused as to whether i should go with cheapest option "1" and put in a 7200rpm hdd, but it has only 2 gb ram....or i should go with option "3" Fastest Processor and put in 7200rpm hdd...?? or middle option "2" and put in a 7200rpm hdd...??

1. How much difference in performance will be 2gb vs. 4gb...??
2. Which Processor i should choose..??
3. Which of the above 3 options i should go with & why...??

Money is not a obstacle... i can afford any of the 3 mentioned but if Option 1 with SU4100/2gb gives same performance as option 3 with SU9400/4gb then i dont want to be stupid to spend extra just because i can... That money can be used elsewhere in life...!!

However if Option 3 with SU9400/4gb will be much better in performance over Option 1 with SU4100/2gb then i wanna get Option 3.

So as you can see i'm really confused...

My usage is as follows…

1. Internet Surfing… about 15-20 tabs open in Firefox.
2. Itunes playing in the background at the same time.
3. Watching videos on Youtube & VLC Media Player.
4. Editing Photos in Picassa.
5. I do not do any Gaming at all.

You seem to be very knowledgeable about the Technical Aspects of computers/laptops and i will really appreciate your honest feedback on my queries...!!

Thanks,

Rana

 

Yes, I did some research before picking the 7200.4. According to tomshardware at http://www.tomshardware.com/charts/2009-2.5-mobile-hard-drive-charts/Mobile-Performance,1138.html, the 7200.4 series nearly tied for first place while being one of the cheapest 320gb hard drives out there, as well as the cheapest 7200 RPM. So at a cost roughly half what most of the other reviewed drives were, and only $14 above the stock cost, it was really a no-brainer - especially since I'm doubling capacity. I bought it on ebay, using Bing's cashback option combined with buy it now to get it brand new for only $56.

As you can see in my benchmarks, switching from 2GB to 4GB of RAM doesn't really provide a measurable performance boost; I am assuming that if you do major multitasking (5-6 unique, resource-heavy programs at once) you may see an advantage from 4GB, but there is little benefit if you don't plan on that level of multitasking. I picked the SU2300 because it was very cheap ($360) while retaining a nice dual core 1.1ghz ULV CPU that will easily triple the performance of the 'atom' crop of netbooks.

The SU7300 and SU9400 are comparable to the SU2300, with a mild bump in CPU speed; 1.1 to 1.3 or 1.4 will be somewhat noticeable in real world usage but not a drastic speed bump. The SU4100, though it is listed as a pentium, is a 1.3 ghz dual core that is fairly similar as well.

For all 3 of those, I would sell the stock HDD and replace them with a 320\500 gb 7200.4 HDD; the 320 seems to be a better value, as the 500 costs roughly double.

To be honest, the three of those seem like a big price jump for only a mild speed boost over the 1410; they range from 50% to 100% more in cost, yet real world usage isn't going to be very different. The 1.4 Ghz processor would likely run 15-20% faster in real-world use, noticeable but not a big deal - and IMO, really not worth paying double the money. I haven't personally monitored eBay for those laptops, but I would suggest you find a few different models and visit the ebay search pages for them a few times to see what they tend to go for; remember, a 8-10% bing cashback means that an ebay buy it now will likely be the cheapest route.

The 1410, quite simply, seems to have struck a perfect balance. It's the price of a netbook with the screen size and power of an ultraportable. The only place it is lacking is in battery power, but if you really need it you can sell the stock 6 cell for $35-40, and buy a 9 cell for $80 to get 6-7 hours of battery life. It's not going to beat an Atom with a similarly sized battery, but that huge boost in power (2-3x) will prove enormously beneficial in everday tasks like web browsing.

I'm typing this while visiting relatives on my 1410, and I've been using it a lot the past few days. It simply works amazing for web browsing, video watching etc. - most 'underpowered' netbooks that I've used take a while to render pages, have trouble scrolling smoothly, hang for quite a while every time I try to play videos etc. etc. and all of those problems are vastly reduced or eliminated with the superior power of the 1410.

After reviewing your purchase options, I'm assuing you live in canada where you may have issues with cashback, making traditional options better. If that is indeed the case, I would probably go for the SU7300; the processor won't be a huge boost, but the RAM will go for $40 more and the HDD will probably get another $10-20, so the price bump won't be that big.

Hope this helps, let me know if you have more questions

 

Hi josh, thanks for the detailed reply...

however i didnt quite understand this part... could you re-phrase it... Specially the underlined part...

"After reviewing your purchase options, I'm assuing you live in canada where you may have issues with cashback, making traditional options better. If that is indeed the case, I would probably go for the SU7300; the processor won't be a huge boost, but the RAM will go for $40 more and the HDD will probably get another $10-20, so the price bump won't be that big."

Thanks...

 

Plus the larger capacity battery (5600mAh 6-cell) comes with that model.


Josh, You mentioned a 9-cell battery??? Were you actually referring to the larger capacity 6-cell or is there a 9-cell option out there? I would be interested in a link if you have one. Thanks.

 

I meant that even though the CPU\battery upgrade weren't worth $100, you could sell the other parts (like RAM) to recoup some of the difference and make it worth it.

I was incorrect, it is a higher capacity (5600\5800 instead of 4200) 6 cell battery available for $90.

 

so you meant....

get option 2 with SU7300/4gb/320hdd...

Sell 2 gb ram & 320 gb hdd... ( how & where do i sell them..?? )

& buy seagate 320 gb 7200.4...

right..??

 

Yes, that's what I meant. You should be able to sell them on eBay easily; just look at the prices of similar items already listed, then offer yours for a few bucks less under buy it now.

 

I'm happy with my upgrades did both at the same time but am guessing the 4GB didn't buy much, but the 40GB Intel X25-V was totally worthwhile. SSD + 1810tray makes it about as quiet as you can get, had to get used to the silent operation, and for what I'm using it for, storage is not an issue.

The 160GB that came with my SU2300 is an absolute joke... I could tell in the past anytime the machine lagged, you'll see the HDD light on constantly, and it used to basically freeze on occasion with any amount of heavy disk access.

 

josh... thanks a lot.... you've been very helpful... i really appreciate it...!!

 

Yes, the 160GB is a joke. But the 7200.4 is really quiet - it makes a gentle whirr that's completely different from the annoying grinding most drives make. It reminds me of white noise (I use a white noise generator to sleep on long car trips). Benchmarks with my Vertex showed the additional performance gain over the 7200.4 is quite small; the CPU is the bottleneck for most tasks - even application loading didn't see that much of a benefit from the 7200.4 to SSD. That's not to say there wasn't a noticeable speed increase, it's just that it only occurred during a few specific HDD-heavy tasks, and didn't feel like as much of a difference as moving from the 160GB to the 7200.4

Plus, the 7200.4 is $14 extra and double the space, the X-25 is $100 extra and 1/4th the space. It's worth it if you have the money, but I would spend it in other areas (like upgrading the HDD), and it's a dealbreaker if you need to store a decent amount of programs\data. 40GB is just too small for most systems.

It's the choice of the user, if the money isn't that big of a deal and you don't mind the smaller space, it's certainly worth doing. It'll extend battery life and make the notebook nearly silent. But also keep in mind that SSDs don't like to be totally filled; you shouldn't exceed 30GB on a 40GB drive, or performance will start to suffer and lifespan decrease as garbage collection and wear leveling run out of room.

Glad I could help

 

I did this awhile back last year, and didn't find #3 necessary. Are you suggesting MPC HC + Haali can't be DXVAed by 4500MHD w/o hack?

Regarding OP, thanks for the write up. Although I didn't run benchmarks, I didn't notice any significant speed jump w/ the the Samsung Spinpoint m7 500GB. But w/ my particular 1410, the vibration in the chassis was excruciating to me. So I then opted for the Hitachi 5K500.B w/ lowest power consumption in class to get slightly better batt life at a minimum.

Ram wise, i haven't run anything to really make a palpable difference in stock, but the OCZ 4GB I put in (800mhz) was basically free, after rebate and sale of original 2GB so that was a no brainer for me.

 

I'm not surprised that ram had no effect. You were using 6-6-6-18 and should have tried to find 5-5-5-15 or 4-4-4-12 DDR2-800. The DDR2-800 looks like overclocked DDR2-667 to me. Your ram is like going from a 5400rpm to 7200rpm HDD or like ATA100 vs ATA133. If you found lower latency DDR2-800 it might be more like a move from HDDs to SSDs. Same max transfer rate but lower latency means more throughput by allow high data rate transfers to be sustained longer.

Also I don't know if you used it but in device manager if you check the properties of your various drives and click the policies tab, you will see more options to boost performance. I always put the check in the option to "turn off drive cache buffer flushing". Makes the drives work faster if it doesn't have to spend time flushing the cache buffer onto the hard drive regularly. It's perfectly safe to do this if the system has backup power (battery) because a power failure would prevent the cache buffer from getting a chance to update the stored drive data.

 

I do extensive overclocking on my main desktop, and from experience I can say that RAM latencies have FAR less effect on even RAM benchmarks then MHz does, and even a move from 10-10-24 to 6-6-16 registered a barely-noticeable 3% increase on a few memory-dependent real world benchmarks. The RAM was binned as 1800 MHz by Corsair, and CPUz verified that the RAM was indeed running at the proper speed. Spending considerably more on RAM with even fewer measurable benefits is unlikely to be a worthwhile prospect.

 

I'm not an expert but looking into thaiphoon burner at ram timings there's those 4 states that are advertised and then there's a bunch more dram latencies/timing states that are hardly ever changed which I imagine have an effect on the module's performance. My guess is those advanced timings make a big difference between high performance and low performance in the same DDR2 data rate?

 

They really don't. I've fiddled with every setting imaginary while trying to squeeze a few more MHz out of my RAM. There's a reason they are archaic, well-hidden settings that you can only adjust on a completely unlocked BIOS - they are good for little besides insuring stability, and every one of them combined can be tweaked with no measurable performance gain or loss. MHz is by far the most important gauge of performance, and going from, say, DDR3-1200 to DDR3-1800 can have a 20%+ boost in raw memory performance, while going from 12-12-24 to 6-7-6-18 can have a 5-6% boost. Tweaking any settings besides those won't produce any measurable results on performance (although bumping the Row Address to Column Address Delay fall time a few miliseconds can let you squeeze 20-30 MHz more while remaining stable).

 

Ok thanks, I always found it strange that the Acer 1410 BIOS lists the ram speed as 0667 or 0800 instead of just 667/800. It makes me think there's a way to get DDR2-1066 to work on it.

 

http://shop.ebay.com/i.html?_nkw=dd...70.l1313&_odkw=1066+notebook&_osacat=0&bkBtn=

I don't think DDR2 1066 notebook RAM exists. I doubt the laptop would support it anyway, you'd probably need an unlocked BIOS to overclock it up to 1066.

 

I always thought DDR2-1066 notebook ram existed in high end gaming laptops a few years ago, I could be mistaken. The internet says JEDEC added the specs to it but the existence of such modules is supposed to be rare.

interesting find [POST]6166299[/POST]