System:
PC TORQUE/SAGER NP5720-V Laptop Computer
Specs:
Mainboard (MoBo): Intel 915PM + ICH6-M accepting: Intel® Pentium® M Processors 730/740/750/760/770/780
(1.60 - 2.26GHz, 90nm, FSB533, 2 MB L2 cache, uPGA478)
CPU: Intel Pentium M "Dothan" 770 - 2.13 GHz/2 Mb L2 Cache/533 MHz FSB
GPU: NVIDIA 7800 Go GTX (Ultra) PCI-E 16x W/256 Mb DDR3 Video Memory
RAM: 2 Gb's Dual Channel DDR2 SDRAM @ 533 MHz/DIMMS 2x
HDD: 80 Gb 7200 RPM Ultra ATA 100 HDD
Optical Drive: 8x DVD +-/R/RW - 4x DL+R + Lightscribe
Battery: SMART Li ION 8 Cell Battery
Sound: Integrated AZALIA compliant interface/SRS WOW
Interfaces: Touchpad with Scrolling Slider, 1 Type II PCMCIA Slot
Infra Red, Wireless Fast IR interface, 4 USB 2.0 Ports, 1 IEEE-1394 Port (4 PIN) Other Ports: 1 S/PDIF out/Side Speakers out, 1 Line-in/Rear L/R Speakers out, 1 Headphone/Front L/R Speakers out, 1 Microphone in/Center & subwoofer Speakers out, 1 DVI out port, 1 RJ-45, 1 RJ-11, 1 S-Video TV-out, 1 Serial, 1 TV-in Port (Requires optional TV-Tuner)
Wireless Network Interface: Integrated 10/100/1000Mbps Ethernet
Power: Auto Switch 110/220V AC Adapter, AC power cord
Dimensions: Dimension (H) x (W) x (D) : 1" - 1.8" x 15.6" x 11.5"
Weight: 8.3lbs. w/Battery
LCD: 17.0" WUXGA 1920x1200 Active Matrix LCD Display with Clearview Display, 1.3 Mega Pixel Digital Video Camera
Operating System: Microsoft Windows XP Professional w/SP2
--------------------------------------------------------------------------
Introduction:
We all know that choosing a laptop can be a difficult endeavor. It can require strenuous effort to choose a laptop manufacture one feels comfortable with. For most of us, budget ultimately is the deciding factor for choosing a laptop. But once we get past make and cost or vice versa, intended usage, specification, resale value, portability, style, durability, reputation, quality and value become a paramount concern. Furthermore, anyone of these factors amongst a million other concerns may become your influence in choosing one laptop over another.
For me, the most important factor was quality to cost ratio, not simply value, value can be very misleading if the equipment you're purchasing has fantastic specifications at a low price but is of low quality. Performance is extremely important to me as well because I am a gamer. I subscribe to the theory that if you buy the most powerful laptop you can, you can do just about anything else with it as well. Specifications of high quality were a necessity; durability quite simply falls under quality for me, then style, then resale value, then portability. Reputation, like value can also be a wee bit deceptive, but ultimately it influences us one way or another at some level. However I was cautious with reading too much into reputation as my review will reveal.
In order for any review to have worth, I felt my personal circumstances mentioned above should be known. Therefore any insight derived from my review can be had with the perspective of what is important to ME in a laptop, or simply my circumstances.
*****Please consume this review from the perspective of what is important to YOU in a laptop and NOT what is important to me. In other words, too many folks read negative input regarding a given piece of equipment on these boards and simply negate the technology because someone else say's so (Here we go with reputation again, reputation is deceptive because you can read something bad about an item and 900,000 other folks think the item is great, the opposite can be true as well). Its all perspective, I will do my best to deliver what I consider to be fact and let the reader assemble his/her own conclusion.*****
--------------------------------------------------------------------------
The Ordering Process:
For me, ordering a laptop online is a PAIN IN THE A$$. The reason being is that I know full well I am getting ready to look at the desert tray and I know I can’t have anything off of it because what I’m looking at is fake!!! You all know what I’m talking about, you go into a fancy restaurant and request to see the desert tray and the waiter brings out the tray and all of the deserts are those high quality plastic replicas! You have to order your desert and you can’t just pick from the tray! Annoying as hell to me…
In all seriousness, ordering from PC TORQUE was a pleasant experience, much more so then my past Alienware nightmare (THE PHASE GAME SUCKS!!!). PC TORQUE offers several convenient options for ordering a laptop, from several methods of payment, an intuitive interactive order interface; a responsive sales staff (Actually ordered my lappy from LUKE@PCTORQUE ;-) to special shipping arrangements for the customer. It’s clear why PC TORQUE has developed such a strong following. The best part of the process was the extremely quick turnover time. PC TORQUE provides CA residents the option of having Sager ship their lappy’s to PC TORQUE in TEXAS so applicable CA sales tax can be avoided! How many companies will do this? Very, very few across the board, this is why I decided to bypass Sager direct and go with PCT. Furthermore, the exact same config from Sager costs more then from PCT! I requested a special method of payment and had my lappy within 2 weeks of ordering it. Even though I annoyed Luke by calling him a few times, he was more then happy to deal with my ANAL RETENTIVE A$$! Honestly the order and delivery process was as smooth as silk, not unlike my 6,796 3DMARK05 marks ;-).
--------------------------------------------------------------------------
First Impression:
To be completely honest, after owning an Alienware and seeing the beautiful packaging that comes with an AW, I was very disappointed when I received my 5720-V. I mean, comparing the two is like comparing Apples to Oranges, AW is a high-end boutique reseller, a marketing machine selling a brand, Sager cuts the trimmings and simply delivers the goods. Even though the AW was nothing more than a CLEVO, after AW massages them, they really do take the “Clone” feel out of the machine. This is done with the AW branding and some of the extra steps AW goes through to make their systems a cut above from an aesthetic point of view.
When I received my 5720-V and saw this big ugly brown box with “Made in China” stamped on it (Not that products made in China aren’t of quality, its just that I have also seen those stamps on a million items at my local swap meet and realize the stuff is being resold cheap and wholesaled even cheaper, some of it of questionable quality). I was a little worried about what I might have just spent $2,800 bucks on. I was at least expecting a nice white Sager Midern box, when I opened the box, I was a little taken, the system appeared “Cheap” to me, though I am very biased as I was comparing the system’s packaging to my previous AW. This was a mistake because AW is ridiculously over priced, I knew this when I bought one originally, I bought it because their resale value is second to none. So I put all that aside and was actually very happy with what I saw, the 5720-V is a beautiful machine, and HUGE! (See Photos) I could not believe the size of the sucker. Very big machine… Notice the size of my Logitech Trackball as a perspective reference point to the size of the system. The weight of the system was surprisingly light, compared to a 7700 tank, the 5720-V felt great. The keyboard was the same standard issue I had on my 7700 and I loved it then and still do, best laptop keyboard I have ever used. The 5720-V’s upper chassis (Palm Rest area) did appear a little cheap compared to the Sager 98XX series. However, this is also due in part to keeping the weight of the unit down, you want a prettier chassis, and it’s going to cost you. The upper chassis palm rest area has an almost cheap laptop cooler plastic appearance, however, as I mentioned, this bodes well for weight and heat so it was fine with me. Over all the unit is solid and quite unique looking, I am getting ready to have SMOOTH CREATIONS do some insane paint work to mine, so its all good ;-).
--------------------------------------------------------------------------
LCD:
The LCD was a scary issue for me, after seeing the beautiful AW Area-51m 7700 (Sager 9860) LCD, I was worried I would be a little let down. My 7700 had a 1680x1050 resolution and was absolutely breathtaking to view. By far the best LCD screen I have ever seen. The 5720-V’s massive 1920x1200 LG LCD is beautiful, but not as nice as the 7700’s to be completely honest. The 7700 used a Philips LCD which was just awesome.
All that aside, it’s a beautiful LCD screen in its own right. My screen has what some refer to as the “Sparkles”. An almost glittery sheen, “Sparkles” cover the LCD screen and are much more apparent when colors such as RED are displayed. Being that my last LCD was like a Sony XBRITE, this takes about 3 hours of getting use to and you’re over it, that is unless you needlessly worry about it.
I immediately checked my LCD’s clamshell for the crevices that so many of you have reported, thank goodness, my clamshell was fine.
What was surprising what the amazing viewing angle of this LCD (See photos), this was impressive.
Furthermore, the BISON integrated webcam is awesome and full screen to boot.
Overall, this display is everything that the DELL XPS M170 is for a whole lot less cash. You want a Sony XBRITE display, then this LCD may not be for you. Nice screen, believe me…
--------------------------------------------------------------------------
Sound:
The integrated Realtek sound is damn good. It sucked royally on the 7700 and kicks major a$$ here due to the SRS WOW system. This truly makes gaming a pleasure as the SRS WOW system truly immerses the gamer. It’s really cool how the system play’s with one’s auditory system. Very neat and brought to you at an affordable price, you can learn more about SRS WOW here: http://reviews.cnet.com/4531-10921_7-6384149.html
Sound is very good people…
--------------------------------------------------------------------------
Battery:
Battery life is decent and can approach the 2 hour mark if your extremely battery frugal and use a few power saving tricks. Real world, you’re looking at 1 to 1.5 hours of B-time. If you want battery life then don’t buy a gaming notebook, buy a slug and go 6 hours. Most of us will go for the first electrical socket we can find and stay plugged in anyway. So the battery life is OK but who cares!
--------------------------------------------------------------------------
Lightscribe:
This piece of ingenious technology is perfectly melded into Nero’s CD/DVD burning software and is an absolute treat. No more illegible hand writing on CD’s and wasted INK JET ink, Lightscribe technology literally let’s you produce photo realistic laser generated labels on your CD’s and DVD’s. It’s a little time consuming and currently only allows for grayscale reproduction, but it’s definitely awesome and worth every penny. Bitchen!
--------------------------------------------------------------------------
The “Benchies”:
Being an Engineer by profession, I have always been quite skeptical of microcomputer benchmarks. For one, unless there is exhausting “Test-Bed” data, it’s very hard for me to lend too much credibility to benchmarking period. I need controlled conditions, with even more controlled testing methods and baseline validation for me to take any “Benchie” too seriously (Yes I am your typical ANAL RETENTIVE Engineer). Two, it’s a well known fact that benchmarking software, such as Futuremark’s 3DMARK05 has fallen under great scrutiny due to some “Shady” driver tweak’s from NVIDIA in the past. Of course there are two sides of the coin, this issue may have occurred due to some creative exploration of the 3DMARK05 source code by NVIDIA to tune their display drivers to produce artificially high benchmarks to sell more of their product. And the other side of the coin is that Futuremark may have been just as guilty by facilitating NVIDIA to more “Easily” obtain their code, after all, one isn’t too valuable without the other. Kind of makes you wonder why benchmarking software like this could ever be taken seriously in the first place, don’t tell me that companies like NVIDIA and ATI did not kick a few bucks down to aid software development efforts of apps like 3DMARK. This is only my opinion, but co-ops like this are common in automotive benchmarks (Namely dynometer testing, where for instance the insane HP/TORQUE numbers you’re reading about were generated while the car was on RACE GAS versus PUMP allowing for a turbocharged car for instance to run at higher boost levels then when on PUMP gas, and maybe a little undisclosed ECU tuning as well. I personally know several performance shops in LA who have told me stories of the BS that gets printed in car magazines).
With all of that said, keep in mind that most computer benchmarks have a very high degree of variance with regard to actual performance and benchmark performance. Microprocessors (CPU/GPU no matter) are highly volatile in terms of how much load can be placed on a given chip and how well it will perform over time. Every chip varies in its respective ability to perform under a given set of conditions and the chips tolerance to those conditions. This is why you can take two seemingly identical processors and receive completely different results. This is because the processors, though of an identical design specification, actually vary quite a bit in their silicon properties. Thus providing varying results, even if all things were equal, operational conditions, component variance (As mentioned above, identical designs aside, no two microcomputer components can truly be identical due to such factors as manufacturing), temperatures, and quite a few other variables would cause a variance in results no matter how identical two given machines might be.
So without further ado, I am providing the following benchmarks under the following conditions:
Room Temps: 67 to 70 F.
Operating System: Windows XP Professional SP2
Laptop Cooler: Targus Notebook Chill Hub (Fans Off, used as platform only)
NOTE: While I owned my Area-51m 7700 Mobile Desktop, I test several laptop coolers and only “1” got my blessing. The KINGWIN ULTRA COOL UC3 performed hands down the best and perfectly accommodates a 17” lappy. When I sold my 7700 I sold my UC3 with it, so until I get another one, this POS Chill Hub will have to do ;-).
Testing Duration: 4 to 5 hours a day of continuous testing over a period of (2) days.
Benchmarking software/tools: 3DMARK05 Professional, Mobilemeter, Game XP, Diskeeper Professional 9.x, Spy-Sweeper, Ad-Aware, Xoft-Spy, Spyware-Blaster and Microshaft’s free BETA Anti-Spyware software. Spybot is great, but I think I got the Anti-Spyware threat covered pretty well ;-).
Conditions: While some of you tweak your laptops wildly, I felt a more “Real World/Average User” test-bed condition would be more beneficial to most folks on these forums. This is because many of us do not have the time to tweak and tweak our hardware and OS for every last 3DMARK possible. Many of you still in college can do this, but for many of us with careers and demanding time schedules, this sort of tweaking just isn’t possible. Furthermore, some users simply don’t feel comfortable performing major tweaks to their systems. The intention here was to provide a few basic benchmarks with some very cool “One click system TWEAKWARE” and see the results. Not only were the results phenomenal with this software, IT’S FREEWARE! This freeware is known as GAME XP, it’s a simple program that tweaks XP for optimal gaming performance without knowing how to do so manually. It’s valuable to all of us because even for us hardcore gamers that know how to tweak our hardware/software, we know this can be annoyingly time consuming. This cool software makes some XP registry tweaks (Even let’s you back up the specific registry alterations before they are made and also has a complete RESTORE feature) that mod XP’s ability to use more of your Pentium M’s massive L2 cache, optimizes GPU performance, uses no DLL’s, and NO installation is required. This app can be had here: http://www.theorica.net/gamexp.htm . This software is perfect for setting up a custom gaming profile and experimenting with your benchmarking configurations. As always, use this app at your own risk as anytime you modify your computer’s registry in an explicit manner, catastrophic circumstances can occur. USE AT YOUR OWN RISK.
The following benchmarks were run with the following configurations:
1.All standard services XP Pro services loaded
2.No standard services: Only the exceptions, NVIDIA service, Intel Registry service, etc… So very few services.
3.With Game XP registry optimizations
4.Without Game XP registry optimizations
5.With all tasks running in task manager
6.With only the necessary tasks to keep XP stable running
7.Basic XP performance optimizations such as: Power scheme set to “ALWAYS ON”, Plugged in = Nothing ever shuts down, On battery = Nothing ever shuts down (Useless for these benches as my benchmarks were all performed with AC power).
8.Performance options Visual affects: Adjust for best performance.
9.Performance options Advanced: Processor scheduling Applications
10.Performance options Advanced: Memory usage System Cache
11.Swap File: 1.5x system memory is the MS recommended rule of thumb, I have 2 Gigs O’ RAM and my settings are 2046Mb (Initial) and 4096Mb (Max)
12.Screen resolution: 1024x768x32 @ 60 Hz,
13.Video Driver: NVIDIA Forceware 81.94, write combining: ON
14.ATA 100 HDD: “Enable write caching on the disk”
15.Screen Saver: Off
16.ABSOLUTELY NO OVER CLOCKING OF CPU/GPU or PM UNDERVOLTING. (Which I have mixed feelings about anyway).
17. 5720-V Function + F2 @ AUTO (Fan Mod)
Results:
Mobilemeter temps were ICY cold:
CPU:
40 C. while using MS word – 45/57 C while running 3DMARK05
HDD:
33 C while using MS word – 36/38 C while being thrashed by 3DMARK05
1. Baseline 3DMARK05 score or “BONE STOCK” w/ Forceware 79.31, All tasks, All services, NO Game XP: 6764
2. NVIDIA 81.94 driver upgrade w/All tasks, all services: 6768
3. 81.94, All tasks, All services, Game XP: GPU = 6764 - CPU = 3816
4. 81.94, All tasks, All services, NO Game XP: GPU = 6766 - CPU = 3869
5. 81.94, All tasks, NO services, Game XP: GPU = 6777 – CPU = 4624 - Very nice here as well
6. 81.94, FEW tasks, NO services, Game XP: GPU = 6778 – CPU = 4816!!!
7. 81.94, FEW tasks, NO services, Game XP: GPU = 6785 – CPU = 4769!!! - Best Overall Score
8. 81.94, All tasks, NO services, NO Game XP: GPU = 6793 – CPU = 4006
9. 81.94, FEW tasks, NO services, GAME XP: GPU = 6796 – CPU = 3960
Summary:
There is far more than meets the eye here, I never did break the 6800 barrier, which bummed me out a little, however I think with the right driver, this is quite possible. I was using the basic NVIDIA Forceware and not hopped up Forceware, so I think I did well. 6796 is pretty darn close to 6800 (I mean the difference is not noticeable) and satisfies me just fine, FOR NOW ;-).
The most amazing finding from my testing was the incredible CPU scores I was obtaining with my 2.13 GHz Pentium M. This is what I was referring to when I said there was more than meets the eye here. During the tests where I scored significant CPU scores, even though the GPU scores do not show it, my frame rates were FAR MORE consistent and sustained then all other tests that I performed. So for instance, instead of my frame rates being 26, 28, 27, 24, 23, 19, 23, 29, my frame rates would be 35 36, 36, 39, 40, 40, 40, 40, 41, 42, etc… So despite the 3DMARK05 score, the consistency and exponentially sustained frame rates were remarkable. Game XP definitely had an effect on my CPU scores, though the 3960 CPU score doesn’t reflect this definitively, my system had been running for 7+ hours when this score was obtained and other factors such as simple operational equipment variances could have attributed for this.
Also of note, closing certain tasks actually hurt my scores. I also noticed this with specific services. Closing down the Intel Registry Service made my scores go down (Duh, go figure). You really need to experiment here and see what works best for you for your machine.
--------------------------------------------------------------------------
NOTE: OVERCLOCK AT YOUR OWN RISK!
NEW - 12/28/05
Overclocking:
Anyone who knows me knows that I have mixed feelings about OC'ing laptops. The confinements, cooling limitations, hardware replacement expense, and overall risk to benefit simply have not equated with me as a whole. After building water-cooled systems and GPUs in the past and pushing systems to their very limit, it's hard for me to OC anything without going all the way. Owning a lappy as my main gaming machine has been a true lesson in dialing myself down a few notches, which is a good thing, it's another technical challenge. That said, I consider myself to be an open minded individual and have come to learn that nominal OC'ing can be safely accomplished in a laptop form factor when performed correctly and responsibly. There is plenty of evidence on these boards from some qualified individuals that OC'ing a laptop does not have to bring certain death to a lappy, one just needs to be more cautious with a laptop then with a desktop. There is clearly less margin for error.
Admittedly, the majority of my OC'ing experience has been with ATI cards and 3DFX Voodoo series cards. Albeit nVidia bought 3DFX out, it took me awhile to turn green... I have been having some trouble OC'ing my 7800 Go GTX (Another thread). I am new to nVidia OC'ing and am learning alot from some very knowledgable folks, kudos to you.
My Overclocking Ideology - To me, in order for laptop OC'ing to have any worth, it must be done with the least invasive methodology possible. I realize some folks will "Pin Mod", "Undervolt", Apply AS5, etc... For my purposes, software OC'ing is as far as I am willing to go with my 5720-V. Thus, the following information is provided from the perspective of mildly intrusive, "Bolt-On" overclocking. The following data will be updated as much as possible as I continue my thirst for nVidia knowledge and pursue my favorite hobby, gaming =).
OC Baselining - For most users, the least amount of effort producing the greatest results while maintaining reliability, component safety, and maximum performance benefits is the goal for OC'ing a laptop, "Bang for the buck". Unfortunately, far too much emphasis has been placed on synthetic benchmark apps like 3DMARK05'. This app takes into account several factors and computates a score based on those factors. It's a multi-dimensional benchmark utility, which some may argue is beneficial. I however disagree, the main reason being is that several GPU performance benchies have been done by some very credible folks while actually gaming (Not running a bench app), and the results tell a far different story. For instance, by default, 3DMARK05' clearly takes into account for your CPU (2D CPU tests) and your overall score is affected by how well your CPU performs during the bench. However, look at this data regarding CPU speed and how CPU speed affects performance in CALL OF DUTY 2 http://www.gamespot.com/features/6140517/p-4.html As you can see, its pretty clear CPU SPEED DOES NEARLY NOTHING FOR YOU IN COD2. However, when going for that awesome 3DMARK05' score, it's imperative everything is OC'd to get the best possible score. From the average user's perspective, apps like this can easily misguide someone into thinking that their system is nothing unless they can achieve a certain 3DMARK05' score. This of course could not be farther from the truth. Real world gaming and 3DMARK05' are two completely different animals. Scoring high in 3DMARK05' tells me that you're technically savvy and can optimize a system under specific conditions, and that's great. I have found that what folks are reporting on 3DMARK05' and the performance they are actually seeing in games is not nearly as profound as you may think.
NEW - 12/28/05
Overclocking Benchmarks:
The following benchmarks will show what I feel are unrealistic improvements in performance from the perspective of "Real World" value. These scores were obtained using the same techniques as above except for one small difference. I accessed "Advanced Chipset Controls" in the 5720-V BIOS and OC'd my 5720-V from 6% - 12%. I felt that starting at 3% was of little value and that OC'ing a full 15% was of little value, which I will explain more.
It breaks down like this, the OC feature in the 5720-V BIOS allows you to OC your 5720's Front Side Bus (FSB). This IS NOT your clock multiplier (16x, etc), this simply lets you adjust the MHz (Your system's clock frequency, or the amount of times the crystal on your MoBo will vibrate per clock cycle (Millions of Hertz = MHz, for you nerds) over your system's default FSB clock of 133MHz in increments of roughly 4 MHz for every +3% you OC in your 5720-V's BIOS. So a 6% increase would OC your FSB to 140.99 MHz, now X this by 16 and you have 2255.84 which equates to 2.26 GHz, (The 2255.84 is Base 2 if you even care). So an easier way to think of it would be like this, you are OC'ing 66.5 MHz to your aggregate CPU clock speed for every 3% you OC. So a 6% increase equates to OC'ing your CPU 133 MHz over your CPU's stock clock frequency aggregate. So if your CPU is a 2.13 GHz PM, you would now be running a 2.26 GHz PM if you chose an OC of 6%, another 3% would mean another 66.5 MHz on top of that, etc. If your still wondering, this should help, say your CPU is a 2.13 PM, your system clock is 133 MHz, multiply this by 16 and you have 2.13 GHz (Actually 2.128, not going to go into Base 2 numbering here, it will confuse some of you). Add another 133 MHz or 6% in your 5720-V's BIOS, and you now have 2.26 GHz. OC'ing a processor by OC'ing a chip's FSB is the proper way to OC a CPU, people all too often play with multipliers and screw everything up. The reason Sager allows this in the 5720-V's BIOS is because the PM has an unreal reputation for OC'ing exponentially without having to touch voltage levels. Read on for a little more on this... I hope this helps some of you that have had questions about the OC feature...
Because I am currently not able to OC my GPU (I think I may need the Sager driver's to do this, I can't with the nVidia driver's so far, even with Coolbits unlocked). I decided to exploit the absurdity of 3DMARK05' scores by showing how much my 3D MARKS went up from simply OC'ing my CPU 6%-12%. The results are as follows:
1. 6% CPU OC to 2.26 GHz, no services, few tasks, game XP - GPU: 6846 CPU: 4084
2. 9% CPU OC to 2.33 GHz, no services, few tasks, game XP - GPU: 6864 CPU: 4239
3. 12% CPU OC to 2.40 GHz, no services, few tasks, game XP - GPU: 6879 CPU: 4365
So here is the quick and easy:
1. +83 3D MARKS over initial baselining.
2. Most substantial gain was obtained at +%6 CPU OC (2.26 GHz) with 50 3D MARKS over previous best.
3. Additional +6% CPU OC (+12% - 2.4 GHz) only netted an additional 33 3D MARKS over +6% CPU OC (2.26 GHz).
Now I was pretty disappointed with the above scores, but I am also aware that if the GPU had been OC'd, I would have seen some pretty incredible numbers with as little as a 6% CPU OC. A 6% CPU OC mated to a marginal GPU Core/Mem OC should easily push me across the 7000 barrier if I so desire. I think the above numbers show, OC'ing your CPU over +6% is pretty much worthless from a risk to benefit perspective.
Back to my point, by simply OC'ing my CPU 12% or to 2.40 GHz, I netted a whopping +83 3D MARKS over my initial baselining. This may not sound like alot to some of you, but remember, my GPU frequencies were absolutely bone stock. So my argument here is that those additional 83 3D MARKS are nearly worthless in "Real World" gaming but look bitchen on an NBF signature. Now at this point you're probably wondering how my system ran @ 2.4 GHz, well I must admit, I was a little scared to take it that high. I had great success at 6% and 9% so I decided to go for it for the following reasons. If you notice, there is no voltage interface in the 5720-V BIOS. This is a little presumptuous, but I don't think Sager would even add the option to OC by as much as %15 without a voltage interface if they weren't fairly certain the PM could do so at completely stock voltage levels. If anything, you might run into system instability issues, but frying your chip is a lot easier when you have access to a voltage interface, that's why it's not there and you don't need it (Read on).
What amazed me the most perhaps was that my system temps at a bone stock 2.13 GHz were within a degree or two at 2.4 GHz according to Mobilemeter. This is a true testament to the amazing Intel Centrino architecture, truly impressive. In the interest of being thorough, I felt it necessary to include some screen captures of the above scores. Futhermore, I have included screen captures that define what I mean by "No Services" and "Few Tasks". Of course services are running, I should have stated "Few Services" but one extra letter is one extra letter less annoying when naming a file, so I used "No Services", lol. You will see the "Few" services I have running in the captures below:
NEW - 12/28/05
In conclusion:
It's my opinion for whatever its worth, that if your going to OC your 5720-V, stick with a "Baby Steps" approach and focus on the refinement of OC'ing your 7800 Go GTX. I think the above data illustrates that OC'ing your CPU to gain points in 3DMARK is pointless. So while you'll add some 3D MARKS, you will gain nearly nothing while actually gaming (As clearly exemplified in the COD2 article). COD2 benefits far more from HT Procs, Dual Procs, and Dual Core Procs then misdirected single CPU frequency manipulation. Unless your using your system for 2D apps (Why would you even buy a 5720-V in the first place?), the benefit from "Real World" gaming from OC'ing your CPU alone simply does not equate. The data before you is represented as a "Snapshot" versus a "Trend" over time. OC like this long enough and you could very well have problems, though I doubt it with the PM. Consider this article snipet from Tom's Hardware regarding the PM 770, "We overclocked the Pentium M 770 to 2.56 GHz by increasing the frequency of the front side bus to 160 MHz, without even having to raise the core voltage." That says it all doesn't it, this explains why there is no voltage interface in the BIOS and why one isn't needed ;-). Remember that while 3D MARKS all make us envious, they really don't mean much as you can see. All in all, I can tell you this with a high degree of certainty: An identically equipped 5720-V with a 2 GHz PM versus my 2.13 GHz PM may score a little differently in 3DMARK05' leading you to believe that the 2.13 is faster, just like a 2.26 PM, however, you set them side by side in a real game and I beg to see a real difference.
The 256Mb DDR3 7800 GTX "Desktop" card clocks in at - 430/1200, to my knowledge. The Go clocks in at - 400/1100, if you can reasonably and safely get your GPU to clock at the same frequency as the 7800 DT version, there is absolutely no reason to go any farther in my opinion...
More GPU OC data as I get there...
---------------------------------------------------------------------------
Gaming:
To me this was the true test, on anticipation of receiving my system; I feverishly went out and grabbed Quake 4. I have loved Quake since day "1" and I knew that this game would push my system to the limit. I also enjoy the PC version of Grand Theft Auto: San Andreas, so I loaded this, along with Doom 3 and its respective expansion pack, Area-51, and the incredibly fun but dated, Battlefield 1942 and expansions.
The system blew me away with butter smooth frame rates and what really got me was playing for hours and the system was as cool as a cucumber with AUTO FANS! I have played with the Dell XPS and have experienced its underside temps when gaming, IT DOES NOT RUN AS COOL AS THE 5720-V in a comparable config. The 5720-V absolutely tore Q4 up, I was very surprised at how it ran, even under conditions of heavy GPU load, the 7800 Go GTX is just blazing fast. I never thought I would live to see the day when a notebook was kicking a desktops a$$ in games like Q4 (Can’t get to cocky though, there are some SLI beasties hitting 13 to 17,000 3DMARKS in 3DMARK05) I was running at 1280x1024 with 2/4 AA and everything maxed with not a hitch. Admittedly, 2x ran better then 4 obviously, but again, I have not had the time to truly tweak. My next experiment is FEAR, this game looks great. Check this review often as I will update my benchmarks and provide game specific frame rates as time goes on… Doom 3 runs butter smooth and really cannot be appreciated unless you’re running on a GPU like the 7800 Go GTX or better. I thought the game kicked a$$ on my old 6800 Go and all I can say now is wow! Gaming is amazing on the 5720-V, this is truly the fastest laptop I have ever seen… Anyhow, stay tuned for more information as I acquire it…
Gaming is incredible all!
--------------------------------------------------------------------------
General productivity and surfing:
The Pentium M has got to be the best CPU Intel has ever produced; it’s also one of the most underrated. Due to the extremely INEFFICIENT design of the P4 Prescott with its elongated pipe lines and excessive heat, my 3.6 GHz P4 w/1Mb L2 cache in my previous AW Area-51m 7700 was flat slow. I remember thinking at the time, "This thing is sluggish", I experienced excessive delay when simply right-clicking on the desktop to modify GPU settings. This was mainly due to heat, the P4 is a furnace of a CPU. I will say however that the chip excelled nicely when multi-tasking, if your a big MT'er, the PM is good but can't touch a desktop P4 for MTing. The best 3DMARK CPU score I obtained with the P4/HT was 5100 and some change. My 2.13 PM’s whopping 3DMARK CPU score of 4816 proves that the 770 is a force to be reckoned with. Folks say this chip does not multitask well, I disagree, as mentioned above, I would rate it "Good-Very Good". It does not compare to a P4 Prescott like I said, however, if you’re a light multi-tasker such as I, you won’t even notice it. Now mind you, the 770 will choke where the Prescott won’t, but the overall efficiency and speed of this chip is unparalleled. It’s a true technological work of art based on Pentium 3 architecture! Now, design a PM with dual core technology, you then have a Masterpiece.
I found my daily tasks ever so pleasing and was blown away at how responsive the chip is to instruction. The chip is super swift because of its short pipeline architecture and massive L2 cache, combine this with its low heat production due to its already low voltage levels and modest clock frequency and the PM is just a treat. The huge L2 and short pipeline architecture is why you don’t have to run at ridiculous clock rates to get the same thing done as a P4 Prescott. Even the 2Mb L2 P4’s don’t impress me, they run too damn hot for my liking and the additional 1Mb of L2 was a weak attempt at making a bad design better on Intel’s part. Yes the MT is amazing, but the chip is not worth the cost as you can still cook an egg on it and the chip still suffers from "Heat Soak".
All my apps run great with the chip, pretty much everything, even Adobe Photoshop CS. Surfing is great, a beautiful processor.
The Pentium M is the greatest commercial chip Intel has manufactured to date, truly outstanding.
--------------------------------------------------------------------------
PC TORQUE/SAGER NP5720-V Laptop Computer
Specs:
Mainboard (MoBo): Intel 915PM + ICH6-M accepting: Intel® Pentium® M Processors 730/740/750/760/770/780
(1.60 - 2.26GHz, 90nm, FSB533, 2 MB L2 cache, uPGA478)
CPU: Intel Pentium M "Dothan" 770 - 2.13 GHz/2 Mb L2 Cache/533 MHz FSB
GPU: NVIDIA 7800 Go GTX (Ultra) PCI-E 16x W/256 Mb DDR3 Video Memory
RAM: 2 Gb's Dual Channel DDR2 SDRAM @ 533 MHz/DIMMS 2x
HDD: 80 Gb 7200 RPM Ultra ATA 100 HDD
Optical Drive: 8x DVD +-/R/RW - 4x DL+R + Lightscribe
Battery: SMART Li ION 8 Cell Battery
Sound: Integrated AZALIA compliant interface/SRS WOW
Interfaces: Touchpad with Scrolling Slider, 1 Type II PCMCIA Slot
Infra Red, Wireless Fast IR interface, 4 USB 2.0 Ports, 1 IEEE-1394 Port (4 PIN) Other Ports: 1 S/PDIF out/Side Speakers out, 1 Line-in/Rear L/R Speakers out, 1 Headphone/Front L/R Speakers out, 1 Microphone in/Center & subwoofer Speakers out, 1 DVI out port, 1 RJ-45, 1 RJ-11, 1 S-Video TV-out, 1 Serial, 1 TV-in Port (Requires optional TV-Tuner)
Wireless Network Interface: Integrated 10/100/1000Mbps Ethernet
Power: Auto Switch 110/220V AC Adapter, AC power cord
Dimensions: Dimension (H) x (W) x (D) : 1" - 1.8" x 15.6" x 11.5"
Weight: 8.3lbs. w/Battery
LCD: 17.0" WUXGA 1920x1200 Active Matrix LCD Display with Clearview Display, 1.3 Mega Pixel Digital Video Camera
Operating System: Microsoft Windows XP Professional w/SP2
--------------------------------------------------------------------------
Introduction:
We all know that choosing a laptop can be a difficult endeavor. It can require strenuous effort to choose a laptop manufacture one feels comfortable with. For most of us, budget ultimately is the deciding factor for choosing a laptop. But once we get past make and cost or vice versa, intended usage, specification, resale value, portability, style, durability, reputation, quality and value become a paramount concern. Furthermore, anyone of these factors amongst a million other concerns may become your influence in choosing one laptop over another.
For me, the most important factor was quality to cost ratio, not simply value, value can be very misleading if the equipment you're purchasing has fantastic specifications at a low price but is of low quality. Performance is extremely important to me as well because I am a gamer. I subscribe to the theory that if you buy the most powerful laptop you can, you can do just about anything else with it as well. Specifications of high quality were a necessity; durability quite simply falls under quality for me, then style, then resale value, then portability. Reputation, like value can also be a wee bit deceptive, but ultimately it influences us one way or another at some level. However I was cautious with reading too much into reputation as my review will reveal.
In order for any review to have worth, I felt my personal circumstances mentioned above should be known. Therefore any insight derived from my review can be had with the perspective of what is important to ME in a laptop, or simply my circumstances.
*****Please consume this review from the perspective of what is important to YOU in a laptop and NOT what is important to me. In other words, too many folks read negative input regarding a given piece of equipment on these boards and simply negate the technology because someone else say's so (Here we go with reputation again, reputation is deceptive because you can read something bad about an item and 900,000 other folks think the item is great, the opposite can be true as well). Its all perspective, I will do my best to deliver what I consider to be fact and let the reader assemble his/her own conclusion.*****
--------------------------------------------------------------------------
The Ordering Process:
For me, ordering a laptop online is a PAIN IN THE A$$. The reason being is that I know full well I am getting ready to look at the desert tray and I know I can’t have anything off of it because what I’m looking at is fake!!! You all know what I’m talking about, you go into a fancy restaurant and request to see the desert tray and the waiter brings out the tray and all of the deserts are those high quality plastic replicas! You have to order your desert and you can’t just pick from the tray! Annoying as hell to me…
In all seriousness, ordering from PC TORQUE was a pleasant experience, much more so then my past Alienware nightmare (THE PHASE GAME SUCKS!!!). PC TORQUE offers several convenient options for ordering a laptop, from several methods of payment, an intuitive interactive order interface; a responsive sales staff (Actually ordered my lappy from LUKE@PCTORQUE ;-) to special shipping arrangements for the customer. It’s clear why PC TORQUE has developed such a strong following. The best part of the process was the extremely quick turnover time. PC TORQUE provides CA residents the option of having Sager ship their lappy’s to PC TORQUE in TEXAS so applicable CA sales tax can be avoided! How many companies will do this? Very, very few across the board, this is why I decided to bypass Sager direct and go with PCT. Furthermore, the exact same config from Sager costs more then from PCT! I requested a special method of payment and had my lappy within 2 weeks of ordering it. Even though I annoyed Luke by calling him a few times, he was more then happy to deal with my ANAL RETENTIVE A$$! Honestly the order and delivery process was as smooth as silk, not unlike my 6,796 3DMARK05 marks ;-).
--------------------------------------------------------------------------
First Impression:
To be completely honest, after owning an Alienware and seeing the beautiful packaging that comes with an AW, I was very disappointed when I received my 5720-V. I mean, comparing the two is like comparing Apples to Oranges, AW is a high-end boutique reseller, a marketing machine selling a brand, Sager cuts the trimmings and simply delivers the goods. Even though the AW was nothing more than a CLEVO, after AW massages them, they really do take the “Clone” feel out of the machine. This is done with the AW branding and some of the extra steps AW goes through to make their systems a cut above from an aesthetic point of view.
When I received my 5720-V and saw this big ugly brown box with “Made in China” stamped on it (Not that products made in China aren’t of quality, its just that I have also seen those stamps on a million items at my local swap meet and realize the stuff is being resold cheap and wholesaled even cheaper, some of it of questionable quality). I was a little worried about what I might have just spent $2,800 bucks on. I was at least expecting a nice white Sager Midern box, when I opened the box, I was a little taken, the system appeared “Cheap” to me, though I am very biased as I was comparing the system’s packaging to my previous AW. This was a mistake because AW is ridiculously over priced, I knew this when I bought one originally, I bought it because their resale value is second to none. So I put all that aside and was actually very happy with what I saw, the 5720-V is a beautiful machine, and HUGE! (See Photos) I could not believe the size of the sucker. Very big machine… Notice the size of my Logitech Trackball as a perspective reference point to the size of the system. The weight of the system was surprisingly light, compared to a 7700 tank, the 5720-V felt great. The keyboard was the same standard issue I had on my 7700 and I loved it then and still do, best laptop keyboard I have ever used. The 5720-V’s upper chassis (Palm Rest area) did appear a little cheap compared to the Sager 98XX series. However, this is also due in part to keeping the weight of the unit down, you want a prettier chassis, and it’s going to cost you. The upper chassis palm rest area has an almost cheap laptop cooler plastic appearance, however, as I mentioned, this bodes well for weight and heat so it was fine with me. Over all the unit is solid and quite unique looking, I am getting ready to have SMOOTH CREATIONS do some insane paint work to mine, so its all good ;-).
--------------------------------------------------------------------------
LCD:
The LCD was a scary issue for me, after seeing the beautiful AW Area-51m 7700 (Sager 9860) LCD, I was worried I would be a little let down. My 7700 had a 1680x1050 resolution and was absolutely breathtaking to view. By far the best LCD screen I have ever seen. The 5720-V’s massive 1920x1200 LG LCD is beautiful, but not as nice as the 7700’s to be completely honest. The 7700 used a Philips LCD which was just awesome.
All that aside, it’s a beautiful LCD screen in its own right. My screen has what some refer to as the “Sparkles”. An almost glittery sheen, “Sparkles” cover the LCD screen and are much more apparent when colors such as RED are displayed. Being that my last LCD was like a Sony XBRITE, this takes about 3 hours of getting use to and you’re over it, that is unless you needlessly worry about it.
I immediately checked my LCD’s clamshell for the crevices that so many of you have reported, thank goodness, my clamshell was fine.
What was surprising what the amazing viewing angle of this LCD (See photos), this was impressive.
Furthermore, the BISON integrated webcam is awesome and full screen to boot.
Overall, this display is everything that the DELL XPS M170 is for a whole lot less cash. You want a Sony XBRITE display, then this LCD may not be for you. Nice screen, believe me…
--------------------------------------------------------------------------
Sound:
The integrated Realtek sound is damn good. It sucked royally on the 7700 and kicks major a$$ here due to the SRS WOW system. This truly makes gaming a pleasure as the SRS WOW system truly immerses the gamer. It’s really cool how the system play’s with one’s auditory system. Very neat and brought to you at an affordable price, you can learn more about SRS WOW here: http://reviews.cnet.com/4531-10921_7-6384149.html
Sound is very good people…
--------------------------------------------------------------------------
Battery:
Battery life is decent and can approach the 2 hour mark if your extremely battery frugal and use a few power saving tricks. Real world, you’re looking at 1 to 1.5 hours of B-time. If you want battery life then don’t buy a gaming notebook, buy a slug and go 6 hours. Most of us will go for the first electrical socket we can find and stay plugged in anyway. So the battery life is OK but who cares!
--------------------------------------------------------------------------
Lightscribe:
This piece of ingenious technology is perfectly melded into Nero’s CD/DVD burning software and is an absolute treat. No more illegible hand writing on CD’s and wasted INK JET ink, Lightscribe technology literally let’s you produce photo realistic laser generated labels on your CD’s and DVD’s. It’s a little time consuming and currently only allows for grayscale reproduction, but it’s definitely awesome and worth every penny. Bitchen!
--------------------------------------------------------------------------
The “Benchies”:
Being an Engineer by profession, I have always been quite skeptical of microcomputer benchmarks. For one, unless there is exhausting “Test-Bed” data, it’s very hard for me to lend too much credibility to benchmarking period. I need controlled conditions, with even more controlled testing methods and baseline validation for me to take any “Benchie” too seriously (Yes I am your typical ANAL RETENTIVE Engineer). Two, it’s a well known fact that benchmarking software, such as Futuremark’s 3DMARK05 has fallen under great scrutiny due to some “Shady” driver tweak’s from NVIDIA in the past. Of course there are two sides of the coin, this issue may have occurred due to some creative exploration of the 3DMARK05 source code by NVIDIA to tune their display drivers to produce artificially high benchmarks to sell more of their product. And the other side of the coin is that Futuremark may have been just as guilty by facilitating NVIDIA to more “Easily” obtain their code, after all, one isn’t too valuable without the other. Kind of makes you wonder why benchmarking software like this could ever be taken seriously in the first place, don’t tell me that companies like NVIDIA and ATI did not kick a few bucks down to aid software development efforts of apps like 3DMARK. This is only my opinion, but co-ops like this are common in automotive benchmarks (Namely dynometer testing, where for instance the insane HP/TORQUE numbers you’re reading about were generated while the car was on RACE GAS versus PUMP allowing for a turbocharged car for instance to run at higher boost levels then when on PUMP gas, and maybe a little undisclosed ECU tuning as well. I personally know several performance shops in LA who have told me stories of the BS that gets printed in car magazines).
With all of that said, keep in mind that most computer benchmarks have a very high degree of variance with regard to actual performance and benchmark performance. Microprocessors (CPU/GPU no matter) are highly volatile in terms of how much load can be placed on a given chip and how well it will perform over time. Every chip varies in its respective ability to perform under a given set of conditions and the chips tolerance to those conditions. This is why you can take two seemingly identical processors and receive completely different results. This is because the processors, though of an identical design specification, actually vary quite a bit in their silicon properties. Thus providing varying results, even if all things were equal, operational conditions, component variance (As mentioned above, identical designs aside, no two microcomputer components can truly be identical due to such factors as manufacturing), temperatures, and quite a few other variables would cause a variance in results no matter how identical two given machines might be.
So without further ado, I am providing the following benchmarks under the following conditions:
Room Temps: 67 to 70 F.
Operating System: Windows XP Professional SP2
Laptop Cooler: Targus Notebook Chill Hub (Fans Off, used as platform only)
NOTE: While I owned my Area-51m 7700 Mobile Desktop, I test several laptop coolers and only “1” got my blessing. The KINGWIN ULTRA COOL UC3 performed hands down the best and perfectly accommodates a 17” lappy. When I sold my 7700 I sold my UC3 with it, so until I get another one, this POS Chill Hub will have to do ;-).
Testing Duration: 4 to 5 hours a day of continuous testing over a period of (2) days.
Benchmarking software/tools: 3DMARK05 Professional, Mobilemeter, Game XP, Diskeeper Professional 9.x, Spy-Sweeper, Ad-Aware, Xoft-Spy, Spyware-Blaster and Microshaft’s free BETA Anti-Spyware software. Spybot is great, but I think I got the Anti-Spyware threat covered pretty well ;-).
Conditions: While some of you tweak your laptops wildly, I felt a more “Real World/Average User” test-bed condition would be more beneficial to most folks on these forums. This is because many of us do not have the time to tweak and tweak our hardware and OS for every last 3DMARK possible. Many of you still in college can do this, but for many of us with careers and demanding time schedules, this sort of tweaking just isn’t possible. Furthermore, some users simply don’t feel comfortable performing major tweaks to their systems. The intention here was to provide a few basic benchmarks with some very cool “One click system TWEAKWARE” and see the results. Not only were the results phenomenal with this software, IT’S FREEWARE! This freeware is known as GAME XP, it’s a simple program that tweaks XP for optimal gaming performance without knowing how to do so manually. It’s valuable to all of us because even for us hardcore gamers that know how to tweak our hardware/software, we know this can be annoyingly time consuming. This cool software makes some XP registry tweaks (Even let’s you back up the specific registry alterations before they are made and also has a complete RESTORE feature) that mod XP’s ability to use more of your Pentium M’s massive L2 cache, optimizes GPU performance, uses no DLL’s, and NO installation is required. This app can be had here: http://www.theorica.net/gamexp.htm . This software is perfect for setting up a custom gaming profile and experimenting with your benchmarking configurations. As always, use this app at your own risk as anytime you modify your computer’s registry in an explicit manner, catastrophic circumstances can occur. USE AT YOUR OWN RISK.
The following benchmarks were run with the following configurations:
1.All standard services XP Pro services loaded
2.No standard services: Only the exceptions, NVIDIA service, Intel Registry service, etc… So very few services.
3.With Game XP registry optimizations
4.Without Game XP registry optimizations
5.With all tasks running in task manager
6.With only the necessary tasks to keep XP stable running
7.Basic XP performance optimizations such as: Power scheme set to “ALWAYS ON”, Plugged in = Nothing ever shuts down, On battery = Nothing ever shuts down (Useless for these benches as my benchmarks were all performed with AC power).
8.Performance options Visual affects: Adjust for best performance.
9.Performance options Advanced: Processor scheduling Applications
10.Performance options Advanced: Memory usage System Cache
11.Swap File: 1.5x system memory is the MS recommended rule of thumb, I have 2 Gigs O’ RAM and my settings are 2046Mb (Initial) and 4096Mb (Max)
12.Screen resolution: 1024x768x32 @ 60 Hz,
13.Video Driver: NVIDIA Forceware 81.94, write combining: ON
14.ATA 100 HDD: “Enable write caching on the disk”
15.Screen Saver: Off
16.ABSOLUTELY NO OVER CLOCKING OF CPU/GPU or PM UNDERVOLTING. (Which I have mixed feelings about anyway).
17. 5720-V Function + F2 @ AUTO (Fan Mod)
Results:
Mobilemeter temps were ICY cold:
CPU:
40 C. while using MS word – 45/57 C while running 3DMARK05
HDD:
33 C while using MS word – 36/38 C while being thrashed by 3DMARK05
1. Baseline 3DMARK05 score or “BONE STOCK” w/ Forceware 79.31, All tasks, All services, NO Game XP: 6764
2. NVIDIA 81.94 driver upgrade w/All tasks, all services: 6768
3. 81.94, All tasks, All services, Game XP: GPU = 6764 - CPU = 3816
4. 81.94, All tasks, All services, NO Game XP: GPU = 6766 - CPU = 3869
5. 81.94, All tasks, NO services, Game XP: GPU = 6777 – CPU = 4624 - Very nice here as well
6. 81.94, FEW tasks, NO services, Game XP: GPU = 6778 – CPU = 4816!!!
7. 81.94, FEW tasks, NO services, Game XP: GPU = 6785 – CPU = 4769!!! - Best Overall Score
8. 81.94, All tasks, NO services, NO Game XP: GPU = 6793 – CPU = 4006
9. 81.94, FEW tasks, NO services, GAME XP: GPU = 6796 – CPU = 3960
Summary:
There is far more than meets the eye here, I never did break the 6800 barrier, which bummed me out a little, however I think with the right driver, this is quite possible. I was using the basic NVIDIA Forceware and not hopped up Forceware, so I think I did well. 6796 is pretty darn close to 6800 (I mean the difference is not noticeable) and satisfies me just fine, FOR NOW ;-).
The most amazing finding from my testing was the incredible CPU scores I was obtaining with my 2.13 GHz Pentium M. This is what I was referring to when I said there was more than meets the eye here. During the tests where I scored significant CPU scores, even though the GPU scores do not show it, my frame rates were FAR MORE consistent and sustained then all other tests that I performed. So for instance, instead of my frame rates being 26, 28, 27, 24, 23, 19, 23, 29, my frame rates would be 35 36, 36, 39, 40, 40, 40, 40, 41, 42, etc… So despite the 3DMARK05 score, the consistency and exponentially sustained frame rates were remarkable. Game XP definitely had an effect on my CPU scores, though the 3960 CPU score doesn’t reflect this definitively, my system had been running for 7+ hours when this score was obtained and other factors such as simple operational equipment variances could have attributed for this.
Also of note, closing certain tasks actually hurt my scores. I also noticed this with specific services. Closing down the Intel Registry Service made my scores go down (Duh, go figure). You really need to experiment here and see what works best for you for your machine.
--------------------------------------------------------------------------
NOTE: OVERCLOCK AT YOUR OWN RISK!
NEW - 12/28/05
Overclocking:
Anyone who knows me knows that I have mixed feelings about OC'ing laptops. The confinements, cooling limitations, hardware replacement expense, and overall risk to benefit simply have not equated with me as a whole. After building water-cooled systems and GPUs in the past and pushing systems to their very limit, it's hard for me to OC anything without going all the way. Owning a lappy as my main gaming machine has been a true lesson in dialing myself down a few notches, which is a good thing, it's another technical challenge. That said, I consider myself to be an open minded individual and have come to learn that nominal OC'ing can be safely accomplished in a laptop form factor when performed correctly and responsibly. There is plenty of evidence on these boards from some qualified individuals that OC'ing a laptop does not have to bring certain death to a lappy, one just needs to be more cautious with a laptop then with a desktop. There is clearly less margin for error.
Admittedly, the majority of my OC'ing experience has been with ATI cards and 3DFX Voodoo series cards. Albeit nVidia bought 3DFX out, it took me awhile to turn green... I have been having some trouble OC'ing my 7800 Go GTX (Another thread). I am new to nVidia OC'ing and am learning alot from some very knowledgable folks, kudos to you.
My Overclocking Ideology - To me, in order for laptop OC'ing to have any worth, it must be done with the least invasive methodology possible. I realize some folks will "Pin Mod", "Undervolt", Apply AS5, etc... For my purposes, software OC'ing is as far as I am willing to go with my 5720-V. Thus, the following information is provided from the perspective of mildly intrusive, "Bolt-On" overclocking. The following data will be updated as much as possible as I continue my thirst for nVidia knowledge and pursue my favorite hobby, gaming =).
OC Baselining - For most users, the least amount of effort producing the greatest results while maintaining reliability, component safety, and maximum performance benefits is the goal for OC'ing a laptop, "Bang for the buck". Unfortunately, far too much emphasis has been placed on synthetic benchmark apps like 3DMARK05'. This app takes into account several factors and computates a score based on those factors. It's a multi-dimensional benchmark utility, which some may argue is beneficial. I however disagree, the main reason being is that several GPU performance benchies have been done by some very credible folks while actually gaming (Not running a bench app), and the results tell a far different story. For instance, by default, 3DMARK05' clearly takes into account for your CPU (2D CPU tests) and your overall score is affected by how well your CPU performs during the bench. However, look at this data regarding CPU speed and how CPU speed affects performance in CALL OF DUTY 2 http://www.gamespot.com/features/6140517/p-4.html As you can see, its pretty clear CPU SPEED DOES NEARLY NOTHING FOR YOU IN COD2. However, when going for that awesome 3DMARK05' score, it's imperative everything is OC'd to get the best possible score. From the average user's perspective, apps like this can easily misguide someone into thinking that their system is nothing unless they can achieve a certain 3DMARK05' score. This of course could not be farther from the truth. Real world gaming and 3DMARK05' are two completely different animals. Scoring high in 3DMARK05' tells me that you're technically savvy and can optimize a system under specific conditions, and that's great. I have found that what folks are reporting on 3DMARK05' and the performance they are actually seeing in games is not nearly as profound as you may think.
NEW - 12/28/05
Overclocking Benchmarks:
The following benchmarks will show what I feel are unrealistic improvements in performance from the perspective of "Real World" value. These scores were obtained using the same techniques as above except for one small difference. I accessed "Advanced Chipset Controls" in the 5720-V BIOS and OC'd my 5720-V from 6% - 12%. I felt that starting at 3% was of little value and that OC'ing a full 15% was of little value, which I will explain more.
It breaks down like this, the OC feature in the 5720-V BIOS allows you to OC your 5720's Front Side Bus (FSB). This IS NOT your clock multiplier (16x, etc), this simply lets you adjust the MHz (Your system's clock frequency, or the amount of times the crystal on your MoBo will vibrate per clock cycle (Millions of Hertz = MHz, for you nerds) over your system's default FSB clock of 133MHz in increments of roughly 4 MHz for every +3% you OC in your 5720-V's BIOS. So a 6% increase would OC your FSB to 140.99 MHz, now X this by 16 and you have 2255.84 which equates to 2.26 GHz, (The 2255.84 is Base 2 if you even care). So an easier way to think of it would be like this, you are OC'ing 66.5 MHz to your aggregate CPU clock speed for every 3% you OC. So a 6% increase equates to OC'ing your CPU 133 MHz over your CPU's stock clock frequency aggregate. So if your CPU is a 2.13 GHz PM, you would now be running a 2.26 GHz PM if you chose an OC of 6%, another 3% would mean another 66.5 MHz on top of that, etc. If your still wondering, this should help, say your CPU is a 2.13 PM, your system clock is 133 MHz, multiply this by 16 and you have 2.13 GHz (Actually 2.128, not going to go into Base 2 numbering here, it will confuse some of you). Add another 133 MHz or 6% in your 5720-V's BIOS, and you now have 2.26 GHz. OC'ing a processor by OC'ing a chip's FSB is the proper way to OC a CPU, people all too often play with multipliers and screw everything up. The reason Sager allows this in the 5720-V's BIOS is because the PM has an unreal reputation for OC'ing exponentially without having to touch voltage levels. Read on for a little more on this... I hope this helps some of you that have had questions about the OC feature...
Because I am currently not able to OC my GPU (I think I may need the Sager driver's to do this, I can't with the nVidia driver's so far, even with Coolbits unlocked). I decided to exploit the absurdity of 3DMARK05' scores by showing how much my 3D MARKS went up from simply OC'ing my CPU 6%-12%. The results are as follows:
1. 6% CPU OC to 2.26 GHz, no services, few tasks, game XP - GPU: 6846 CPU: 4084
2. 9% CPU OC to 2.33 GHz, no services, few tasks, game XP - GPU: 6864 CPU: 4239
3. 12% CPU OC to 2.40 GHz, no services, few tasks, game XP - GPU: 6879 CPU: 4365
So here is the quick and easy:
1. +83 3D MARKS over initial baselining.
2. Most substantial gain was obtained at +%6 CPU OC (2.26 GHz) with 50 3D MARKS over previous best.
3. Additional +6% CPU OC (+12% - 2.4 GHz) only netted an additional 33 3D MARKS over +6% CPU OC (2.26 GHz).
Now I was pretty disappointed with the above scores, but I am also aware that if the GPU had been OC'd, I would have seen some pretty incredible numbers with as little as a 6% CPU OC. A 6% CPU OC mated to a marginal GPU Core/Mem OC should easily push me across the 7000 barrier if I so desire. I think the above numbers show, OC'ing your CPU over +6% is pretty much worthless from a risk to benefit perspective.
Back to my point, by simply OC'ing my CPU 12% or to 2.40 GHz, I netted a whopping +83 3D MARKS over my initial baselining. This may not sound like alot to some of you, but remember, my GPU frequencies were absolutely bone stock. So my argument here is that those additional 83 3D MARKS are nearly worthless in "Real World" gaming but look bitchen on an NBF signature. Now at this point you're probably wondering how my system ran @ 2.4 GHz, well I must admit, I was a little scared to take it that high. I had great success at 6% and 9% so I decided to go for it for the following reasons. If you notice, there is no voltage interface in the 5720-V BIOS. This is a little presumptuous, but I don't think Sager would even add the option to OC by as much as %15 without a voltage interface if they weren't fairly certain the PM could do so at completely stock voltage levels. If anything, you might run into system instability issues, but frying your chip is a lot easier when you have access to a voltage interface, that's why it's not there and you don't need it (Read on).
What amazed me the most perhaps was that my system temps at a bone stock 2.13 GHz were within a degree or two at 2.4 GHz according to Mobilemeter. This is a true testament to the amazing Intel Centrino architecture, truly impressive. In the interest of being thorough, I felt it necessary to include some screen captures of the above scores. Futhermore, I have included screen captures that define what I mean by "No Services" and "Few Tasks". Of course services are running, I should have stated "Few Services" but one extra letter is one extra letter less annoying when naming a file, so I used "No Services", lol. You will see the "Few" services I have running in the captures below:
NEW - 12/28/05
In conclusion:
It's my opinion for whatever its worth, that if your going to OC your 5720-V, stick with a "Baby Steps" approach and focus on the refinement of OC'ing your 7800 Go GTX. I think the above data illustrates that OC'ing your CPU to gain points in 3DMARK is pointless. So while you'll add some 3D MARKS, you will gain nearly nothing while actually gaming (As clearly exemplified in the COD2 article). COD2 benefits far more from HT Procs, Dual Procs, and Dual Core Procs then misdirected single CPU frequency manipulation. Unless your using your system for 2D apps (Why would you even buy a 5720-V in the first place?), the benefit from "Real World" gaming from OC'ing your CPU alone simply does not equate. The data before you is represented as a "Snapshot" versus a "Trend" over time. OC like this long enough and you could very well have problems, though I doubt it with the PM. Consider this article snipet from Tom's Hardware regarding the PM 770, "We overclocked the Pentium M 770 to 2.56 GHz by increasing the frequency of the front side bus to 160 MHz, without even having to raise the core voltage." That says it all doesn't it, this explains why there is no voltage interface in the BIOS and why one isn't needed ;-). Remember that while 3D MARKS all make us envious, they really don't mean much as you can see. All in all, I can tell you this with a high degree of certainty: An identically equipped 5720-V with a 2 GHz PM versus my 2.13 GHz PM may score a little differently in 3DMARK05' leading you to believe that the 2.13 is faster, just like a 2.26 PM, however, you set them side by side in a real game and I beg to see a real difference.
The 256Mb DDR3 7800 GTX "Desktop" card clocks in at - 430/1200, to my knowledge. The Go clocks in at - 400/1100, if you can reasonably and safely get your GPU to clock at the same frequency as the 7800 DT version, there is absolutely no reason to go any farther in my opinion...
More GPU OC data as I get there...
---------------------------------------------------------------------------
Gaming:
To me this was the true test, on anticipation of receiving my system; I feverishly went out and grabbed Quake 4. I have loved Quake since day "1" and I knew that this game would push my system to the limit. I also enjoy the PC version of Grand Theft Auto: San Andreas, so I loaded this, along with Doom 3 and its respective expansion pack, Area-51, and the incredibly fun but dated, Battlefield 1942 and expansions.
The system blew me away with butter smooth frame rates and what really got me was playing for hours and the system was as cool as a cucumber with AUTO FANS! I have played with the Dell XPS and have experienced its underside temps when gaming, IT DOES NOT RUN AS COOL AS THE 5720-V in a comparable config. The 5720-V absolutely tore Q4 up, I was very surprised at how it ran, even under conditions of heavy GPU load, the 7800 Go GTX is just blazing fast. I never thought I would live to see the day when a notebook was kicking a desktops a$$ in games like Q4 (Can’t get to cocky though, there are some SLI beasties hitting 13 to 17,000 3DMARKS in 3DMARK05) I was running at 1280x1024 with 2/4 AA and everything maxed with not a hitch. Admittedly, 2x ran better then 4 obviously, but again, I have not had the time to truly tweak. My next experiment is FEAR, this game looks great. Check this review often as I will update my benchmarks and provide game specific frame rates as time goes on… Doom 3 runs butter smooth and really cannot be appreciated unless you’re running on a GPU like the 7800 Go GTX or better. I thought the game kicked a$$ on my old 6800 Go and all I can say now is wow! Gaming is amazing on the 5720-V, this is truly the fastest laptop I have ever seen… Anyhow, stay tuned for more information as I acquire it…
Gaming is incredible all!
--------------------------------------------------------------------------
General productivity and surfing:
The Pentium M has got to be the best CPU Intel has ever produced; it’s also one of the most underrated. Due to the extremely INEFFICIENT design of the P4 Prescott with its elongated pipe lines and excessive heat, my 3.6 GHz P4 w/1Mb L2 cache in my previous AW Area-51m 7700 was flat slow. I remember thinking at the time, "This thing is sluggish", I experienced excessive delay when simply right-clicking on the desktop to modify GPU settings. This was mainly due to heat, the P4 is a furnace of a CPU. I will say however that the chip excelled nicely when multi-tasking, if your a big MT'er, the PM is good but can't touch a desktop P4 for MTing. The best 3DMARK CPU score I obtained with the P4/HT was 5100 and some change. My 2.13 PM’s whopping 3DMARK CPU score of 4816 proves that the 770 is a force to be reckoned with. Folks say this chip does not multitask well, I disagree, as mentioned above, I would rate it "Good-Very Good". It does not compare to a P4 Prescott like I said, however, if you’re a light multi-tasker such as I, you won’t even notice it. Now mind you, the 770 will choke where the Prescott won’t, but the overall efficiency and speed of this chip is unparalleled. It’s a true technological work of art based on Pentium 3 architecture! Now, design a PM with dual core technology, you then have a Masterpiece.
I found my daily tasks ever so pleasing and was blown away at how responsive the chip is to instruction. The chip is super swift because of its short pipeline architecture and massive L2 cache, combine this with its low heat production due to its already low voltage levels and modest clock frequency and the PM is just a treat. The huge L2 and short pipeline architecture is why you don’t have to run at ridiculous clock rates to get the same thing done as a P4 Prescott. Even the 2Mb L2 P4’s don’t impress me, they run too damn hot for my liking and the additional 1Mb of L2 was a weak attempt at making a bad design better on Intel’s part. Yes the MT is amazing, but the chip is not worth the cost as you can still cook an egg on it and the chip still suffers from "Heat Soak".
All my apps run great with the chip, pretty much everything, even Adobe Photoshop CS. Surfing is great, a beautiful processor.
The Pentium M is the greatest commercial chip Intel has manufactured to date, truly outstanding.
--------------------------------------------------------------------------
