NP9262 arrived, pics, benches, screenshots

That's a lot of good information. Can't wait for mine to arrive

I really appreciate all the benchmarks and info you've been posting. When I get mine next week I can run them with 7200 RPM RAID-0 drives and see what kinda differences we see.

Sabregen if you need more storage space/bandwidth, just shoot me a PM. I can set you up a subdomain on my webhost without any bandwidth issues.

 

Wow, thanks heaps man.That is great info. ESIV used to be a ball breaker on all machines, seems it just eats up htis game like a wad of chewing gum. Great to see as I mod for ESIV.

I am so looking forward to getting my XR5. Btw, I'm getting 3x 200gb (Seagate) SATA300 in raid 5 config for instant drive failure back up/hotswap out the dud if it ever occurs.

Thanks again!!!

-Coors916

 

just got my OS up and running on my new 9262 from xotic. I'm no genius when it comes to these things and had to contact their customer service for a possible driver issue. I must say that all the positive things I've heard about their support are true. Going from dealing with sony over the vaio I sent back to talking to these guys is like waking up from a horrible nightmare and realizing life really isn't that bad after all... now on to the games!

 

Gratz on the new rig. I hope you enjoy

 

Thanks for the info, mine is shipping as I type to San Fran, then it gets brought over to New Zealand, should have it in 9 days, man I can't wait

 

Is it any diference in perfomance between just a 100 GB 7200 rpm HD and 2 x 100 GB 7200 in raid 0?

 

Well umm with raid 0 and 2 100 GB your running 200GB @14,400RPM.

 

2 drives at 7,200rpm is not 14,400 rpm. rotational disk speed is not cumulative. That's like saying a quad 2.4ghz cpu like my q6600 is 9.6ghz, when it clearly is not. If it were, do you not think Intel would use that as the rated speed of the CPU? Disk space in a RAID-0 array is cumulative, rotational speed is not.

 

you should see roughly a 25-40% increase in throughput speed due to the added disk. Keep in mind, this is just an estimate. Actual performance will vary on the program being used, and whether the operation is a read or a write. Reads are faster than writes in RAID-0, so things like level loading and uncompressing images, and archives shoudl be noticably faster. Yes, even in a desktop system, without the benefits of an ASIC or onboard battery backed cache, you will still see a difference between a single and multiple disk configurations.

 

Whatever, all i know is that it will defiinitely be faster than a 42 or 54.

 

The tone in your original post is certainly a "Well duh...jeez, I can't believe I even have to say this...I mean, everyone knows this..." When clearly, the poster was just looking for some help. If they had known, then they likely would not have asked, and your response was not only arrogant, but incorrect, as well.

As far as spindle speed is concerned, yes 7,200RPM is faster than 4,200 or 5,400. 10k and 15k are faster than 7,200. I trust you see the progression. Please preface erroneous information posts with a disclaimer, so people who do not know you're full of it can avoid reading your drivel, and thus, further confusing themselves. You do a disservice to those truly looking to understand something that they don't.

 

Sabre,

What motivated your boss to make this purchase? Just curious what its intended function is, etc (other than being one fantastic machine, that is).

 

for an S Corp, end of year funds must be either used or lost (essentially, more accurately, lost to taxes), so he deicded to get new notebooks for myself and another employee that is constantly on the road. The BB 8830 in my sig was also part of the spending, as was $50k in upgrades to our in-house servers, as we move to heavily virtualized environments, and server consolidation. The laptop is a great machine to test out desktop virtualization hosts, and to a lesser extent, it is also good for server virtualization, although I become increasingly bound by disk performance when doing server virtualization on this machine. He loves Clevo designs, and has both an Alienware P4 based and a Sager P4 based notebook himself. I've always loved Clevo (more specifically, Sager) designs as well, and in my consulting days, steered many clients towards a reboust design to fit their needs from one of Clevo's resellers.

 

Yeah, whatever, the fact of the matter is, is that it will definitely be fast than 4200 or 5400, remember the bigger the number the better, if you can hook up some 10000RPM Raptor drives in a desktop, then you will ultimately be a boss among the ranks.

 

Yes, I agree with this latest post. Keep in mind that there are other governing factors besides spindle speed, but as a generalization, the faster the spindle, the faster the disk. We are getting off topic here, especially in reference to WD Raptors, which is a desktop implentation of entry-level enterprise hard drives. This thread is regarding the NP9262, which can neither use desktop 3.5" drives, nor can it use any other 10k drive, since no manufacturer currently makes them in the form factor required by ht eIntel ICH8R controller. It supports up to SATA 3.0Gbps drives in the 2.5" form factor. There are many manufacturers that make 2.5" 10k drives, but none of them work in this machine. Please stay on topic, and do not repost what has already been said as affirmation of the obvious. You can merely agree...or not.

If you'd really like to argue that WD Raptor 10k drives put you at the top of the heap in terms of desktop HDD I/O Performance, please PM me, and I can show you numerous products that smoke Raptors.

 

Yeah, the 9262 you can only go as high as 7200 for a notebook as far as I know, so yeah, two of those raid 0 you'll be make 9 and 10000 in 3D Mark in no time.

 

Interesting, because I'm getting 9600+ using two 5,400 RPM 160's in RAID-1 (which is notoriously bad for writes). I would venture to guess that even with my lowly 5,400RPM drives in a RAID-0 I can bust 10k in 3DMark06, and I'm running Vista Business x64 as well. XP users get an almost automatic points gain, due to lack of overhead that Vista brings with it.

Also, the limitation of drive choices, as far as rotational speeds for hard drives is concerned, that is a limitation of the market, not of the laptop, it's Intel ICH8R chipset, or power requirements for internal disk drives. It is a limit of the market. THere simply is no drive in current existence, available through retail channels, that surpass 7,200RPM in a 2.5" SATA 1.5 or 3.0Gbps package. If there were, this laptop would very likely be able to support it. I've even gone as far as to dig through the white page articles on the ICH8R, Call Brian Law at Sager, and Sager technical support so that I can rest on the fact that there is no way to change out the backplane connections on the motherboard for ones that will support SAS drives. If that was possible, they do make 2.5" 10k SAS drives, I would surely opt for those. The MTBF on SAS drives is far longer, and they usually come with warranties to support this, as well.

I do not know why you continue to state the obvious as if you just realized it, and it is gospel. More bothersome than that, is that I cannot refrain from replying to your posts to point out this fact.

Can someone else please interject here and take the mic for a second? I could use some fresh thoughts or questions.

 

Yes Sabre, question! Mic to me

Another guy on this forum told me that you shouldn't get the 9262 with all the three drives in raid 0 because it used what he called software raid. Not "real raid". And according to him, this software raid could handle three drives in raid 0, BUT the performance would be less than if you only had two drives in raid 0. Since the software was optimized for 2. There would also be alot more HDD failures.. Not only because the risk increase with 3 drives but also from the software..

You seem to know alot about this. I don't Do you have any answers?

 

Whoa, settle down its not that serious, okay then lets change it up,

Heres a question, what is a solid state drive.

 

A drive that doesn`t use power spining (yes it does not spin) and essentially the fastest one around,increasing boot and load time drastically. Basically.But they`re ultra expensive , have extremely low disk space(32,64 gigs etc) and are limited in read/write cycles...

 

Okay, here's the deal. What this other guy is referring to as "real RAID" is an ASIC with a battery backed cache, as would be typical to find in a RAID array for a workstation, or for a server. Almost all desktop implementations of RAID are "psuedo RAID," "fake RAID," "software RAID," etc. They do not have a true "controller" that handles the interrupt requests, the calculation of the parity bits, or hardware logic.

Most desktop RAID controller do have a chipset, so let me get that out there right now. If you get a board that has a JMircon or SIIG, or SiliconImage RAID controller, you likely will indeed find a chip on the board that has the company logo on it. Yes, this is a RAID controller chip. However, these chips are usually nothing more than a controller BIOS with a firmware on it. All of the RAID levels that are supported are supported through an extensible BIOS with a software image on it. All of the calculations for the array are actually being done on the CPU.

With all of the caculations being done on the CPU, and not actually on the "controller" chip for the RAID solution, you're taking a hit from overhead on the CPU performance. Now, that's not to say that the performance hit will be significant at all, but for true performance afficianados, a "software" or "fake" RAID solution is simply unnacceptable. This is largely true when you get into the HPC (high performance computing) environments and scientific environments, because the people using these machines need every CPU cycle possible to run algorithms. Their compute levels are so high, and the clusters are so large, that even a 0.01% difference in CPU usage can lead to a difference in task completion of weeks.

To summarize, a real RAID controller has dedicated CPU, and battery backed cache to assist in the calculation of the parity bits, and also to maintain a cache of the files that it has already accessed (if I need to explain how cache works, I'm guessing you're pretty lost reading the rest of this). All of these things accomplish the following:

1.) True calculation of array geometry
2.) True parity calculation that is checked for errors with ECC memory, and battery backup, in case of power loss
3.) the ability to move the array to another physical machine (just put controller and drives in the other enclosure)

Performance with a real RAID controller is much better than with a "psuedo" RAID solution. Also note, that most dekstop RAID solutions only provide RAID levels 0, 1, and JBOD, with a few supporting RAID level 5. However, your rebuild times on any of the RAID levels that provide redundancy will be very long on a desktop implementation. This is because most desktop systems can only rebuild the array once a driver initializes the rebuild, which means that the OS has to be running as well. If the OS has to be running in order to properly initialize a rebuild, then you're losing the CPU cycles that the "fake" RAID implementation could be using to rebuild, because the OS is constantly generating requests for I/O and...they all go through the CPU.

By comparison, a drive failure, and subsequent rebuild, even on an entry level dedicated, or "real" RAID solution will be much faster, as it is able to allocate all of it's resources to verifying and rebuilding the array's data.

Here's a simple example:

PC Games are written for a very broad spectrum of hardware, and as such, to get the best performance out of a game, it is usually recommended that you build the fastest and most powerful system that you can afford, in order to run the game at the highest level you can afford. Sometimes, even after several thousands of dollars later, you find that a new game cannot be run at maximum detail settings, and you consider an upgrade.

Console's by comparison, use much cheaper hardware, in a very defined environment, and all software (that's legal) that runs on them is designed to take specific advantage of the exact specifications of the said console. As a result, most games for the console offer a good gameplay experience, with few dropped frames, and have feature sets that are in line with other titles on the same system. You can reasonably expect a console game to perform well, due to the restriction of the level of hardware available. The same cannot be said of all PC Games.

desktop or "fake" RAID controllers are like PC games. It's a mixed bag, and just because one RAID controller supports a feature set, doesn't mean the next one will. "Real" RAID controllers are like console games. They have a defined task, and are not expected to do anything outside the nature of their design, and as such, are much faster at this specifically designed task. They are a special purpose controller, just like a gaming wheel, a sound card, a Physx processor, or a Killer NIC. They're much better at what they do, but you have to be willing to pay for it, over the "bundled" feature already on every board out there.

As far as the ICH8R controller performing slower with a 3 drive RAID-0 stripe than a 2 disk RAID stripe, I'm inclined to call BS, but I'd love to see some data. RAID-0 typically scales very well, with more spindles. It only stands to reason that a 3 drive array where there is no redundancy to calculate, and no parity either, would scale with more drives in the array. If he happens to have any HDTach or IOMeter benches to back up these claims, I'd be interested in taking a look.

Hope that helped.

 

A solid state drive is no different (technologically) from what we have been using for years...USB thumdrives/keys.

A solid state drive is merely an array of non-volatile memory on a PCB with an industry standard drive interface connection (instead of USB). They are very low voltage, often only requiring 0.1w to idle, and thus, also extend battery life tremedously in mobile applications. Because they have no moving parts, they are slo extremely good choices for environments where data reliability is paramount, such as in vehicles, military applications, or in aircraft, where high vibration levels would quickly cause a normal disk drive to deteriorate and lose operational status very quickly.

Also, because they have no moving parts, they are not subject to the same levels of potential damage from multiple G-force impacts. Also, due to the lack of a rotational read head or spindle speed, their delay time to access and/or write data is in order of magnitudes better than traditional magnetic disk drives. Their latency is usually measured in "ns," just as system memory, instead of "ms" for normal hard disk drives, and optical media access times.

The throughput on these devices are usually good, although some models have had issues with write performance, with reads being excellent. As such, solid state disk drives are not yet ready for the prime time, but certainly have their uses. Currently, the orther governing factor to their adpotion is the same as every emerging technology: the cost of adoption.

Relative to traditional disks, which can be just as fast, or faster, and cost much less, there's not a lot of people willing to spend thousands of dollars for bragging rights.

The last two issues seem to take a back seat to the cost of ownership, and those are: form factor, and storage space. Currently, most solid state drives are in the 1.8" form factor, with several now emerging (after CES 2008) in the 2.5" form factor (for use in laptops). Currenty available capacities are closer to what one might deem to be traditionally more in line with SCSI storage sizes, and are in mulltiples of 32GB: 32, 64, and 128GB of total capacity, before formatting.

 

^nice. hope the guy can understand even a quater of it. >_>

 

Thank you! Helped alot. Rep power to ya!

The other guy became angry on alot of people and left this forum a couple days ago unfortunaly. But I will come back to you if I find anything that will back him up

(I know what cache is )

 

That part wasn't specifically for you, but the proverbial "you," for anyone else reading and trying to understand.

 

Thats awesome about the solid state, but I dont think i will be needing that.

 

I tried to buy the sager NP9260 from xotic to UK but i found out that ill have to pay taxes when it arrives but the sales suport cant specify how much taxes ill pay the all price without the suposed taxes will be 3500 $ can any1 give me a final price because im geting upset in how hard it is t get a freaking laptop just t play!
ty for ur help

 

Maybe someone else can chime in on this request. I am in the SouthWest USA, so I can't say for sure what sort of taxes or fees you'd be looking at to get it to the UK.

 

If you live in Sweden like I do, you'll have to pay 25% VAT when you pick up your computer. Somtimes you'll have to pay duty too, but not for laptops i think. One expensive computer!

If you go to: http://customs.hmrc.gov.uk You'll find the information you need if you click on "individual & employees" and then "internet shopping".

I've checked and it states that a laptop to UK is duty free but you'll have to pay 17.5% VAT

On the same website you can learn that "Massage apparatus" is duty free and "are liable to VAT on importation at the zero rate (0%)." Why play on a computer when you can play with yourself for free

 

If you know anyone in the states then that person can help you to recive your order, repack it in a new box and then send it to you. If you want to be really safe then send the software and documentation seperatly. But that's illegal if you don't pay the VAT so don't You'll have to pay for the shipping but for a 4000$ computer you'll save some money.

 

I heard when you do it through airmail its perfectly legal.

 

Hey Sabregen,

Check out my hdtune score....

http://forum.notebookreview.com/showthread.php?t=184849&page=2&highlight=hdtune+score

Scroll down a bit...

This is with 3 x160 7200rpm drives....

I do alot of virtualization and the drives make a world of a difference...

Joe