Upgrading Samsung Q35 (with T7600G?)

I was requested to upgrade this notebook. Owner loves it and doesn't want to buy a new one.

Thankfully, this particular Q35 has 512MB soldered RAM and a SATA HDD connector. As far as I understand that's the best to be expected from this model. Adding 2GB RAM stick and 2 more USB via CardBus is trivial.
Screen seems to be good enough (and there is no option for higher res anyway), battery is performing great despite its' age, so they will remain intact.

However, I have not decided on the CPU yet. T7600G is the best possible option. How do I know whether mainboard supports it and overcloking it by multiplier prior to ordering? Should I extract BIOS somehow and look there? Please enlighten me on the subject. And I know it will need a new power brick if CPU is upgraded.

As for HDD upgrade, I am having in mind an 9.5mm SSHD inside an optibay, with cables hooked up from HDD bay (which, sadly, takes 7mm HDD only). Not very elegant solution, but seems best for a notebook with a single SATA1 connector, and I would prefer to avoid messing with either PATA or mpcie RAID cards.

 

If you're doing this to learn on, and for free to the owner, go ahead.

When a client approaches me with this 'offer', I point them to a new system and/or walk away.

No amount of upgrades will be worth the cost - especially compared to a new system with full warranty and all the platform extras that (7 years?) of progress brings to the table.

 

ztilleroftheearth, I am doing this for free, it's kinda gift. Of course offering to replace Q35 with a new notebook (clevo w230ss and some others) was the first thing I did, but owner insists on keeping it. So I agreed to make necessary upgrades and here we are. On the other hand, I don't want to end up with couple old misfit CPUs in the process.

Talking about years of progress - I can't agree with you. Unlike 90s and early 00s, progress in the industry has slowed down substantially. 5+ years old maxed-out systems still beat new cheap ones to pieces even performance-wise, not mentioning comfort - where the f is this damn industry going?! It's very sad, really. However, not the case for Q35 - its' iGPU was crap even on release, but with some OS tinkering it might stand the test of time until Skylake. Overclocked T7600G would help a lot, thogh.

 

I guess we'll agree to disagree about the progress over the years. A maxed out 5+ year old system is not to be compared to a cheap current one in any real scenario where money will be exchanged and performance is really needed.

Sorry I can't help with the cpu upgrades (any that I've seen done have all been frankenbuilds, with dubious 'improvements').

With your work being 'free', it may be worth doing this (dollar-wise). But your time is not free (even if you don' charge for it) and that is why I suggested to walk away initially.


Good luck.

 

I had a Q35 once upon a time (read about it here) and it was a nice machine but the screen was too small for my everyday usage.

If you have got a 2GB module then try it and see what happens. The chipset can probably address 2GB and, given the on-bard RAM, would need to ignore something. Or it may sulk and not boot.

Double-check your observation of the drive height. I don't think the 7mm drives had been invented in 2007 and my Q35 came with a Hitachi HTS541010G9AT00 which the specs say is 9.5mm. You might also find the optical drive bay has a slower interface.

IIRC, the Q35 comes from an era when the BIOS needed to be programmed to recognise the different CPUs so the answer to your CPU question probably lies with whatever is the fastest CPU that Samsung put in the Q35. Someone here sold a Q35 with a T7600 so that appears to be an option. However, getting that to work may depend on whether you can track down a compatible BIOS. There's one for Sept 2007 on the Samsung UK support site (I would suggest you grab all those downloads before they disappear).

John

 

John Ratsey, it works perfectly with 2GB RAM stick, refuses to boot with 4GB. I will re-check HDD height and update the BIOS. Hopefully, T7600G will work.
tilleroftheearth, yeah, guess there is a couple of things we never agree on... Anyway, good luck to you too. =)

 

That's good to know about the RAM.

I'll be interested to know the outcome. Good luck.

John

 

In January I bought a used Lenovo T500 and a T400. Both came out in 2008, so we are talking 6 years old now. After upgrades, mostly to memory and hard disks, they offer the same performance of cheap laptops today (and some expensive ULV laptops), but offer more RAM, more storage space, much better build quality, and higher resolution 16:10 screens. All-in I am at about $500 for both of them. I would be hard-pressed to get a single comparable new laptop for $500.

 

While I'm not familiar with the machine in question, judging by the fact that it was offered with a T7200 to begin with means that it will most likely support a T7600G. The microcode is the same.

Having said that, the exact behaviour that one will get as a result of possibly overclocking the unit remains to be seen...personally, I wouldn't...but that's me.

Good luck.

 

No argument at all. You cannot get a comparable notebook for what you paid for 6+ year old technology (even two of them). (No, I'm not being sarcastic here).

The point though is that a T500 and a T400 in 2008 were (both) about $5K worth of equipment - spend that today and see if any comparison would be possible.

Sure, a great deal, no doubt. But what you're missing in addition to the above point is that improvements in battery life, heat, weight and performance (cpu/monitor/wireless/etc.) with todays O/S's and workflows (even simple browsing, ime) is nothing to be scoffed at either.

While I can see the benefit of saving huge loads of money by buying used - it has no bearing when obtaining maximum performance (in all aspects of mobile computing) today is the goal. Of course, that is my goal; it may not be yours.

Btw;
TP's are one of the few notebooks that I would want to upgrade myself (yeah, they're that good). But even TP's improve with each new iteration, even if raw performance seems to 'only' improve 5-8% with each new generation lately.


What I have been made acutely aware of is that $$$$ by themselves are not the deciding factor when deciding on buying a new system; time is.

$$$$ can be saved up and/or earned again; time is a little more slippery to get a handle on. That is why even a ~10% performance increase is worth ~20% higher initial (one time) cost to me. Because depending on the length of ownership, that performance increase each day far outweighs that one time cost.

When I was young, I thought saving $$$'s was the goal - even at the expense of my time which seemed to stretch out into infinity.

Now, paying extra $$ to save me a few minutes a week looks like a good investment - because time seems to be running out much too fast for me right now.


The above is what my computing advice is based on:

Buy the most computer you can afford (always... and without getting into debt) - upgrade (to a new platform) on a need-to basis (based on testing the new platform is actually more productive in your workflows than what you already have) - and always have the most current/maxed out platform possible, given the above, to be able to enjoy the benefits of the progress we're offered.


Anything else and you're not into performance computing - and if you still think you are; no, you're just spinning your wheels.



A personal example is my upgrading my original T100TA 64GB notebook/tablet to one with a HDD in the keyboard dock.

See:
Asus Transformer Book T100TA, 64GB w/ Keyboard Dock at Memory Express

See:
Asus Transformer Book T100TA-DB14T-CA w/ Atom Z3775, 2GB, 64GB+500GB, 10.1in HD, Win 8.1 at Memory Express


What I didn't realize was that the cpu was upgraded also in the new model (Intel® Atom Z3740 to an Intel® Atom™ Z3775). Wow, what a difference.

See:
PassMark - Intel Atom Z3740 @ 1.33GHz - Price performance comparison

See:
PassMark - Intel Atom Z3775 @ 1.46GHz - Price performance comparison


Most would think that the cost of a second T100 would be 'over the top' - but just to make sure, I also added a 512GB SSD in the keyboard dock.


The performance difference between the two systems is phenomenal - especially when Dropbox and OneDrive are syncing my multi-gigabyte sized files (almost 5-1 time difference between the two with Dropbox syncing, with the eMMC storage vs. the USB3.0 SSD storage being almost 3x faster for the better SSD solution).


Now, (since January, by coincidence) I've just spent over 3x the cost of your two systems of getting one T100 as I want it: light, great battery life, fast and with the capacity I need - and this is almost strictly for browsing/email usage. To me this is worth it vs. the amount of $$$ spent.

An extreme example maybe, but one that shows my point of view clearly (and not with extreme workloads either).


Now, I'm not saying to spend everything you have on a new platform. No.

What I'm saying is that (saving up and) spending 20-30% or more on a current system would be more beneficial than potentially saving that money in the first place. In more ways than you can now imagine.











New O/S, new warranty, hardware that is better matched for the O/S and software currently available, lighter, stronger for day to day usage (even if newer systems 'feel' more flimsy vs. the tanks we were used to a decade ago), cooler, faster booting... and the list goes on, the list goes on...

 

tilleroftheearth, you're missing one important thing here. Industry simply does not offer some viable options... anymore!

E.g. I want to have a not-bigger-than-15.4" 16:10 notebook which has really good, preferably scissor-type keyboard, nice sound and can take at least 2x2.5" + msata inside it. There is no such thing, not even sure there ever was. It's either ridiculously storage-crippled MBP, bigger 16:9 Alienware with glossy screen to be replaced right away, or 16:9 Clevo with inferior keyboard and sound - none is good enough.

I mostly write on my notebook, thus comfortable screen and keyboard are vital for me. I often listen to music in the process, sometimes using internal speakers, and carry the notebook with me a lot, including long voyages. I do not want to have loads of devices for different tasks, a single compact phone (phone-pc lately) and a single small-or-midsize notebook is more than enough for me.

Yes, my old Acer performs ~ like modern i3s. However, for my usage pattern it's far superior in all other areas, pure performance aside, than anything around. So I'd rather waste some more time modding it (mSATA & better screen are the last major things to be done) and use it happily for some years after, than buy a new notebook and get annoyed using it every day. If it was easier to drop in some modern mainboard in a compact 16:10 case, I'd certainly go that way - but that is really time-consuming, especially since there aren't any examples around lately.

 

Computers that support SATA but don't support MSATA can be modified to install them. There are always one or two unused SATA channels on the chipset. They are always connected on the motherboard and either go to ground or just terminate without connecting to anything else. These can be tapped (four wires per SATA channel) and connected to any SATA compatible device, including MSATA. The BIOS should detect them perfectly fine. You should be able to find some cases of people doing this.

 

Can't argue with what you want from your system (and I'm not).

But to accept the good of the new, we have to give up some of the good of the old too.

Nothing you mention is a deal breaker in my eyes - even if I may have been as stubborn as you once upon a time.


Change may not be easily apparent as being good at first blush - but don't let yourself be stuck in 'yesterday' like the old man still preferring to walk to town rather than ride in one of those new fangled horseless buggies.

 

Sure, but this means, traces, wasted space and etc. In the most common scenario they are left NC (not connected) and that's it. There is no way to get it outside the chipset. If for some very lucky reason they left them as test points, then yes - mod away. In my case - 5920G has an USB left as 2 test points (D+ and D- you have to get the juice elsewhere), and meant to be second SATA connector but left unsoldered, with all the lines there. Other than that, there is nowhere to be found another SATA line. I've even got to extend that I got faulty MoBo and desoldered the SouthBridge and tested with multimeter if by any chance there's another test point which leads to SATA on the back - no there isn't. Another example is 8740w - premium machine, 6 SATA lines supported by the chipset, 5 SATA lines used (primary slot, ODD, eSATA, dock SATA drive, dock eSATA), 6th left NC. Examples like those are, again, more often than not.

As kinda on topic - I'll take upgrading old and proven machine any day. My next build - 8740w

 

John Ratsey, ajkula66, thanks and good luck to you too. I will update the topic once Q35 is ready, but can't resist continuing off-topic discussion until then. =)

Qing Dao, after a brief look I couldn't find those lanes, there may indeed be not there as triturbo mentioned. Anyway, I would prefer to avoid soldering to the motherboard itself since I'm not very experienced at it, at least yet. I'm planning to either modify esata/usb combo port, or use an mpcie-to-sata card.

tilleroftheearth, how come features I expect from a notebook are yesterday? MBP has great 15.4" displays. Dell Alienware has keyboard and sound. Clevo has best expandability (esp. storage) in any size. I'm totally disappointed by the fact that none of the notebook-producing companies offer all these options in a single package. Older notebooks at least can be modded to have it all - at the cost of i3 or worse performance, but still. I almost don't play computer games recently, maybe that's why I'm not that bothered with it. =\

triturbo, good luck with your 8740W. =)

 

What manufacturer's give is what (they think) customers actually want. At mass quantities (correction; MASS QUANTITIES).

What is that Currently? The cheapest, lightest, touch-enabled notebook that may or may not be what people in the know (you and I) actually know what is best for a great computing experience.


I have almost given up on expandability/upgradeability in a notebook beyond RAM, a second drive bay and/or a wireless card upgrade.

But with DDR4/16GB+ RAM SoDimms getting closer to reality for some future new notebooks, along with NGFF giving the real possibility of multiple, fast and 'small' storage subsystem expandability options - along with AC Wave 2 PLUS capable wifi technology, the notebooks I'm currently using will soon be scrapped and/or donated to make way for this future that looks way better than the 'yesterday' you're clinging too.


Sure, if you can make 'yesterday' work at a level that is acceptable, you're there (and good for you!).

But the DC QT (dual core, quad thread) Atom processor I'm using typing this response 'feels' faster browsing the net than any Intel Core 2 Duo I ever owned (or used).

See:
PassMark - Intel Core2 Duo T7600 @ 2.33GHz - Price performance comparison


See:
PassMark - Intel Atom Z3775 @ 1.46GHz - Price performance comparison


And...

It's doing it with 30W less too. (And up to ~8 hours run time with a much smaller/lighter chassis and battery too).


Anyone who thinks that technology has stood still is simply wishing to ignore what is staring them in the face, if they open their eyes and really look.

 

No one. That's exactly why people resort to upgrading older platforms...

Quite a few folks - myself included - have been doing alright with upgrading "yesterday's" tech...

As per the charts that you've linked yourself, the difference in performance is marginal at best: 1335 vs. 1325 and ranking of 1073rd vs. 1080th.

Not all that bad for a 6-year-old CPU. Granted, Atom will have a significantly better battery life.

With all of that said, T7600 is a 1st generation C2D...something like T9900 would do a *lot* better...see here:

PassMark - Intel Core2 Duo T9900 @ 3.06GHz - Price performance comparison

Of course, one can't install a T9900 in a socket M, so that's a non-issue as well as a non-option in OP's set of circumstances.

While technology has most certainly moved along, I'm unimpressed by Haswell and am far from being the only one.

Not to mention that I have yet to see anyone actually proving that a 16:9 aspect ratio is superior to 16:10...but that's a whole another story.

 

Not spec of that laptop but, TriTurbo has upgraded higher res screens before and helped other figure out the cabling to do so on other laptops. for 10 bucks USD you can get an ODD to USB ODD kit. why not a straight up SSD? helped out my battery life. or is storage a must have?

Extra devices like the USB ODD can fit in the laptop bag for road trips. Speaking of which i need to figure something out my self i nthat department since i just went from 15.6" to a big hoking 17.x"

 

I'm not arguing about 16:10 being a better aspect ratio. I agree 100%. As to being unimpressed with Haswell; your call - but I'll touch on this a bit more further down.

The passmark scores I linked to were to show that with 8.5 times less power needed, the same (raw) performance is reached. But was is even more to the point is that the Atom architecture is a 2 core 4 thread processor which leaves the Core 2 duo in the dust (even with nominally same 'scores').

As for your first point? I think you missed what I wrote?


Anyway, when a once premium 6+ year old cpu architecture is effectively surpassed with a $35 USD SoC - I think it is better seeing what is in front of us rather than trying to show how upgrading and OC'ing is somehow better (even if it does allow one to continue using a 16:10 screen).


See:
Intel Atom Z3775 Tablet SoC - NotebookCheck.net Tech



An acquaintance was showing me a very nicely done up Japanese 'sports' car he had just finished, saying who needs a Porsche/Ferrari when you can just bolt on more power than a Porsche will ever have for less than the cost of the Porsche. The ride around the block was exciting and yes, the car did move very fast.

The difference is in the end, it is still no Porsche, let alone the Ferrari the owner thought it could go up against.


With monitors surpassing 1200 dpi vertical resolution the 'case' of needing a 16:10 aspect ratio monitor is wearing thin.

With systems that weigh less, get better battery life and (shock) yes, are built better (not to mention higher usable/sustained performance in our mobile systems): bringing around a Model T to a client shows more than just a talent with keeping a clunker alive - it also shows that just maybe, the grasp on technology we might think we have, we don't.

 

cdoublejj, 1TB SSD is too expensive, and OP'ed 512GB is only 350GB, while I have couple SSHDs around. Small SSD + HDD is not an option unless I put sata controller in mpcie slot, resulting in usb-wifi - kinda too much hassle.

tilleroftheearth, car is something one's life depends on, while notebook is merely a tool, even when it earns one's living - it's pretty the same as comparing a car with a hammer or a toothbrush, quite pointless. With time and effort one can make even Ford T into a cool sportcar (OK, cool ratrod =) ). Oh, and I prefer old Ferrari/Porsche to new pimped japanese sportcars... even though japanese do build great cars for domestic market, much better than the crap they sell abroad. =) Finally, car market is more customer-friendly. You have the money - you can get whatever you want. Sadly, notebook market has become a consumer market, ruled by marketing people who don't give a damn about usability and ergonomics.

 

Different people have different needs.

Personally, I own an Ivy Bridge 16:10 machine - the last one still being built apart from Macs - and intend to use it until both Windows 7 and myself are ready to retire which should be just about the same time...

I really don't see this as being a proper comparison.

What I'm arguing is that for quite a few people, an 80s W126 Benz - due to its superior build quality - is still a viable daily driver.

Once the scaling in Windows-related software actually gets to the stage of being fully usable when scaled - as it is on Macs and I'm not a Mac fanboy by any stretch of imagination - your point might become valid. The way things stand right now, as nice as the 3K Panasonic IPS LCD in current ThinkPads T540p/W540 is, it's still best viewed at 100% and most people's eyesight is nowhere nearly good enough for that.

I don't know. My CF-52 gets a lot of "wow" from my clients. No one - but no one - laughs at a brand new ToughBook. Or any ToughBook for that fact...

Whether newer machines - even the expensive ones - are actually built better is something I'd very much question...I have yet to see a better-built notebook than the old T61 series - construction-wise - in the non-rugged segment of the market. On the other side of the spectrum, battery life has improved by a huge margin since those days, no questions asked.

While many people can put a Haswell-based laptop with 32GB of RAM to good use, a huge portion of the PC-literate population has no *real* need for such a system. Those who actually do already know that they won't be able to get away by using a maxed-out T500.

My $0.02 only...

 

FlashInTheNight86 and ajkula66,

I think we're now saying (basically) the same things; given our obviously different personal/personality twists and unique ways of expressing ourselves.


One thing I believe to my core is that without a 'time' component when analyzing if a course of action is 'worth it' (when 'tech' is the subject), the decision is only a guesstimate, at best. The 'time' side is the equation makes the decision 'defendable' in my eyes. But, I agree that not every decision needs to be defendable for everyone else.

 

tilleroftheearth, time-wise choosing and buying a new notebook may or may not take more time & effort than reseacrhing upgrade possibilities and actually upgrading used one. I believe the less person knows about a notebook (or other tech), the easier it is for this person to make a final decision. Such person may simply go to a nearest shop and buy the one he/she finds more suitable among sold there, without reading any reviews or researching upgrade possibilites - just simply buy the damn thing and use it happily... until something goes wrong. =)

 

Easy choice does not equal a good choice; again random luck.

and then... 'used happily'? Lol... if that were true, a lot of my clients wouldn't be asking me for recommendations.


I get it that simply buying is easier than actually choosing - but less knowledge is never better than more.

 

triturbo, out of curiosity, what are you going to build out of that 8740W?

 

It has to arrive (next week I hope) and I have to meet it with a screw-driver to tell you for sure, but from all the pictures and threads, I see some unexploited potential. If you know the story of my Acer, then you pretty much know what to expect

 

OK, this evening I've put the Q35 together, except for CPU swap - and will return it to owner tomorrow. Specs are 12.1" 1280*800 / T5500 / 2.5GB DDR2 / Intel X3100 / 750GB 7200rpm SSHD / DVD burner / 4xUSB2.0, FireWire400, Wi-Fi, Bluetooth / Win 8.1 Pro. So far I've spent $5.3 and some parts on it =) . $70 for better CPU, maybe some $30 for PSU if I won't be able to modify one 90W power brick I already have, couple hours rebuilding and undervolting once CPU arrives - and I'm done with it. =)

 

$105.30 total?

Not too bad!

 

Excellent work. You have to feel good about putting life back in that machine