Processor/Video Card Upgrade- Toshiba L305-S5955

Hi All!

I want to upgrade my L305-S5955- frankly, it isn't up to snuff!

I am interested in upgrading my graphics and processor. I do not know at ALL which is compatible for my model and search results here have com up with nothing.

I do have CPU-Z so whatever information will be needed... please let me know!

Thanks~

 

Sorry, but upgrading your video card isn't possible with that notebook.

 

you cannot upgrade ur GPU... For CPU , post a few screenshots of the current CPU u have and ur chipset... For better graphics u can try DIY Vidock...

http://forum.notebookreview.com/showthread.php?t=418851...

But for the cash u'll spend , might as well get a proper new laptop and we can suggest some for u here...

 

L305-S5955

Processor and Chipset
• Intel® Celeron® Processor 900 2.20GHz, 1MB L2, 800MHz FSB
• Mobile Intel® GL40 Express Chipset

 

Although I'm positive you won't be upgrading a graphics card on that laptop, upgrading the processor is a possibility. If you are able to access the processor and remove it, you should be able to replace it with the majority of core 2 duos.

Edit

Sorry, it turns out the processor you have is the most powerful one your chipset will allow.

 

If u got GL40 , 800MHz FSB processors will be ur best bet... in that range T9300 will give best price and performance...

 

Don't waste your money upgrading the processor. If you need a better processor and gpu, sell the laptop and with that and the money you would have spent on an upgrades, buy yourself a better computer.

 

How can you tell him to sell his laptop when he can upgrad his processor to a Core 2 Duo T7300 for abour £40 (ebay)

 

He is looking for better graphics and processor, so he probably wants to game, or at least do something 3d. Upgrading everything that he can isn't going to improve any fps one lickety split though. Also the T7300 is probably a step in the wrong direction. Going from a 2.2Ghz single core R0 wolfdale to a 2Ghz dual core merom just doesn't sound like the best idea. A roughly 50% boost in ability from the T7300 doesn't sound so good when it comes at roughly 3x the power consumption.

But if he really doesn't need better graphics and a faster processor would suffice for whatever work he is doing, then I would suggest getting a wolfdale core 2 duo that runs on a 200Mhz bus.

 

i did suggest he get a new laptop especially if he wants to game... for about $1500 u can easily get an Asus G73 which is like fastest single card laptop....

 

Considering the back-order and build quality issues, I might question that "easily" bit.

 

If yall are recommending a new laptop, then I suggest if its just light gaming, then go for one of the ASUS UL-series like the UL50vt.

 

there aren't that many quality issues... also back-order problem solved easily by buying on ebay... some stock there...

 

This CPU is 65nm and for temperature reasons i'd go with an 800MHz FSB T9xxx series, i think the GL40 chipset is restricted personally by Intel so that it's permanently () for any decent upgrades. I use a T7700 in my T61p and really want a T9300 which i would then overclock to 2.8GHz if i could via a BSEL pin mod if it's restricted. The T7xxx series run fairly hot IMO.

 

The Asus G73 owner's lounge would seem to differ with you. And if that's representative (not certain), then the eBay machines might be duff models that are being further passed on. I don't like the design of the machine, myself, so I have no personal interest in it.

 

I don't think it is a step down because:
1- power consumption is the same for both CPU. Their DTP is the same 35W
2- The T7300 has speedstep hence a better battery live
3- The T7300 has 4MB L2 cache and the Celeron 900 has only 1MB
4- The T7300 has two cores while the C900 has only one.

I reckon that in 3D apps the T7300 can be up to 100% even 110% faster then the Celeron 900.

For 2D and 3D apps the graphics card doesn't matter that much. It matters only for rendering.

for gaming I agree, a laptop with a dedicated graphics card or better a desktop is required. For anyting else a CPU upgrade and possibly more ram would do the trick

 

TDP =/= power consumption. The 45nm celeron can use roughly 1/3 the power of the 65nm C2D at full load, maybe even less if both processors are undervolted.

That's really doubtful. And believe it or not, by default the celeron is downclocked by lowering the FSB through cool n quiet. Also at idle the celeron can consume much less power than the C2D. The wolfdale is a lot more than just a die shrink of merom.

The amount of cache plays little role in the performance of a processor unless you are talking 512k, which does impact performance in games.

True. The C2D will have more processing power, no doubt about that, but:

It will be significantly less than a 100% improvement. Going by clockspeed x cores, it would be an 81% improvement in processing power. However, the two cores wouldn't be 100% efficient in balancing the load so take that down a % or two. The wolfdale is known to get slightly better performance per clock than the merom as well, so take that difference down a few more percent. In games though the extra cache might give the C2D a few % boost, clock for clock, over the celeron.

I have no idea other than in games, where the graphics card is vital.

If the OP really wants to upgrade the processor even though he will still be stuck with the lowly graphics chip, he should get a T4200, T4300, or T4400, all of which are better in all respects to a T7300 and probably cheaper.

 

1- The max DTP of both CPUs is the same (i.e 35W). Please check document
320289.pdf from intel's website. In Page 6 it is stated:
"Dual-Core and Single-Core Standard Voltage (DTP 35W)"

In page 23 of 320289.pdf shows that the max power consumption of the Celeron 900 (C900) is 47A
In Page 30 of 31674505.pdf shows that the max power consumption of the T7300 41A

2- Can you undervolt your C900 with a software like RMclock, NHC or something else? I don't think you So. A T7300 can be easily undervolted (I'm not tolking about pin mod)

3- So you're telling me that a battery live would be extended by three times if a laptop with a T7300 is downgraded to C900 ?
That's impossible. 1) The T7300 can down clock to 800Mhz, and at this speed its voltage is 0.85v. The celeron M900 doesn't downclock and always runs at 2.2Ghz. 2) The T7300 can shut down some of its L2 cache when not in use and thus decreses further its power consumption.

1- The celeron doesn't have cool and quiet since "cool and quiet" is an AMD technology.
2- The Celeron 900 runs at 2.2GHz even when idle (if you don't believe check this). None of the Celeron supports Speed Step.

1- No much different between 512k and 1MB, or 1MB and 2MB, or even 2MB and 4MB but here we are talking about 1MB and 4MB.
2- More and more applications are been optimised to use more and more cache. Otherwise AMD and Intel would save their money and not make CPUs with 8MB and more cache such as the Core i7-920.

1- I think you are confusing black and white. 'wolfdale' is the code name for desktop CPUs such as the E3x00, E5x00 and so on. the T7300 is code named Penryn and the C900 is also code named Penryn. The only difference between the two is in the manufacturing process, which is 65nm for the T7300 against 45nm for the Celeron 900. The architecture of both processors is the same.
Intel uses a two cycles development process. They first introduce a new architecture. Then, while keeping the same architecture the move into reducing the manifacturing process and so on.
That's to say if you take 2 CPUs with the same multiplier, same FSB, same cache, but one manufactured with 65nm and one with 45nm the only difference between them would be in the amount of heat they generate. Their processing power is the same. Their power consumption might or might not be the same.
2- In games I would say your C900 @ 2.9Ghz is as powerfull as a T7300 if not more. @2.2 Ghz well it all depends on the game.
3- In applications such as 3D rendering (which by the way have had support for multi-processors and multicores for decades), audio/video encoding, file compresion and decompresion (such as winrar) the advantage is to the T7300, because of its number of cores and its larger cache. Since most of those are already optimised for Core 2 Duos with their large L2 cache I wont be surprised if a T7300 can complete a rendering task in half the time necessary for a C900.

1- The T4400 might have a tiny advantage when speed matter but gain the T7300 has four time more cache. I would say that they are equal.
2- T4400 in ebay for £45 ( linK)
T7300 in ebay £39.99 ( link)
I bought a T7300 for a friend 4-5 months ago from ebay for £35

 

1. TDP is not a measure of power consumption. Also they arrive at those current levels by dividing the TDP by the maximum voltage, so they go out the window also.
2. Ok fine, a celeron cannot be as easily undervolted, but it still can be, and an R0 revision penryn at such low clock speeds can decrease its voltage a lot more than a merom can.
3. I'm talking about the processor's power consumption, not the computer's total power consumption. According to Intel, their 45nm core 2 processors use 30% less power than a comparable 65nm core 2 processor. We are talking a single core 45nm cpu vs a dual core 65nm cpu. Also at idle the celeron, as do all penryns, has a lower power state than the 65nm merom. Read about it here: http://blogs.intel.com/technology/2007/04/penryn_update.php
4. The celeron actually does downclock. The mobile 965 chipset and newer support a dynamic FSB to throttle the processor. You never notice it because on any processor with speed step, the speed step overrides it. But if you pop in a celeron, unless you switch off throttling, you will see the speed fluctuate between different levels depending on load.

1. The biggest difference is actually between 512KB and 1MB because that is doubling the cache, and there are quickly decreasing returns from more cache.
2. When it's running something straightforward and linear, cache doesn't have much effect because the computer will know what it needs to get from the ram before it needs it. Some of the most cache dependent apps are games because it can't as easily predict what it will need to do next. Cache does generally increase performance, but some times more of it makes no difference. Usually the difference is from 1-3 percent. In some games it can get to be over 10%. The first increase has the most effect, but after that there really are diminishing returns. An extreme example is that I used to run a conroe dual celeron with 512KB l2 cache at 2.8Ghz paired to an overclocked HD4890. I upgraded the processor to a 6MB l2 cache wolfdale, and clocked to the same speeds I was getting 40% more fps (with all graphics settings at minimum in farcry 2).

1. All falling under the Core 2 architecture, there are conroes, meroms, wolfdales, and penryns, all available in different revisions and two L2 cache levels. Conroes and meroms are the desktop and mobile 65nm incarnations of the core 2 architecture. Wolfdale and penryn are the desktop and mobile 45nm incarnations of the core 2 architecture. There is more to it than a simple die shrink. Within the 65nm and 45nm versions, each have had 3 die revisions, each one running a little cooler and able to run at slightly less voltage than the last. These revisions are just minor tweakings. When moving from 65nm to 45nm, the architecture was improved upon even more than in a basic revision. The 45nm chips support features that are not present in the 65nm such as deep sleep and the sse4 instruction set. Other tweaks made the 45nm chips perform better clock for clock than the 65nm chips. Here is a comparison between a conroe and a first revision wolfdale at the same clocks: http://www.tomshardware.com/reviews/wolfdale-shrinks-transistors,1773-13.html
2. I really don't think that even at 2.93Ghz my celeron would be able to beat a stock T7300 in any games that support multiprocessing. Most games have been for several years or more already, and in fact most games from the last year or two do well with quad cores.
3. I agree.

The T4400 isn't just a bit faster, but penryns are a bit stronger clock for clock than meroms. But yeah, I guess in some cases the cache difference might let the merom edge out. However, the difference in power consumed will be huge once you undervolt.

Don't get me wrong, I'm not a merom hater. I just bought a merom on ebay and I have three others sitting on my desk with me. It's just that the revision R0 wolfdale/penryn is such a big improvement over conroe/merom. The distinction is a lot more apparent when you start overclocking on the desktop, however.

 

Yes I know

I've made a mistake in my last post. The T7300 is a merom...I guess this what happens when we stay up late

As I said before the max power consumption of the Celeron 900 (C900) in Amper is 47A, and the max power consumption of the T7300 is 41A. The C900 default vid is between 1.00V-1.250V. The T7300 default vid is 1.0375 - 1.30V.
Do you agree that the power consumption in watt W is equal to V x A? That gives a Max power consumption of between
1.0375 x 41 = 42.58W to 1.300 x 41 = 53.30W for the T7300
1.0000 x 47 = 47.00W to 1.250 x 47 = 58.75W for the C900
As you can see from the above the max power consumption of the T7300 is below the C900.

I think you're talking about pin modding here. There is no doubt that the celeron can undervolt better than the T7300. This is because 1) T7300 had 2 cores hence requires more power to run stable. 2) The T7300 has four time more caches and requires more power to use all this extra cache.

I learned not to trust every word that comes from the mouth of the marketing guys at intel
Assuming that there is really a gain of 30% in power consumption between a CPU made using 45nm and one using 65nm. It still does not matter since the T7300 when idle down clock to 800Mhz while the C900 doesn't down clock even when idle (waiting to be proven wrong ).

Correct me if I'm wrong but doesn't speed step = downclocking?

Honestly you have to explain how? since all the info I'm able to find point to:
no speed step = no downclocking. I have two reason to doubt (but I'm open minded if/when proven wrong )
-/ First my friend has a Acer Aspire 4315 with an intel i965 chipset. His laptop had initially a Celeron M530. I played with his laptop few days and did not waitness any down clocking with the Celeron M530.
-/ Second I bought myself Acer Aspire 4315 as a replacement for my gateway. If had a Celeron T1400 (before I replaced it with a T5250) and I didn't see once the T1400 downclock.

I know that the Celeron M530 and T1400 do not share the same architecture as the C900 but I still believe that no speed step = no downclocking.

The is one case though where I saw a celeron down clock. It was in Advent/eisystem laptop. This laptop had a feature in its bios (which can be activited at will) that allowed the laptop's FSB (and consequently the CPU's FSB) to clock down a bit. That's allowed the celeron on that laptop (I think it was an M430) to downclock from 1.73 to 1.2Ghz. But note that this is a feature built in that laptop not in its CPU.

I saw on the net few benchmarks showing the improvment gained from increasing the size of the cache. So I'll come back to this later when if manage to find them.

1. I'm not denaying the above. But you seems to be messing an important factor: The OS. "deep sleep" is something very good to have but if the OS (Windows XP or Vista for most of us) cannot manage it then it is somehow pointless. Vista is known to have a bad managment for C3 to C6 state (if I recall well)
2. I don't know if I shoud agree or disagree with you here. Games and applications that don't run in parallel need to either be a) built/programmed as small threads, or b) compiled with a compiler that can divided them internally into small threads. Now the problem here is that for most application a step C for example cannot be completed untill step A and B are completed. And if step B needs data from A then B has to wait for A to be completed too. In this scenario having a Celeron 900 @ 2.93 is definetly faster then a T7300.
Besides, the above is what has led Intel to introduce the Turbo Boost feature in the core i5, i7...

I'm not really convinced that the T4400 has a better power consumption. The T7300 has a minimum vid of 0.85v (with superlfm). Does the T4400 support superlfm? What's its minimum VID?

Who said anything about hating merom... at the end of the day the only thing that matters is that one owns a computer (desktop/laptop) that services their needs for the best price possible.

I suggested the T7300 because I spent a lot of time in the past two months looking in ebay for a penryn dual core CPU, so I knew that price wise the T7300 is unbeatable

Anyways it was nice exchaging opinions with you.

P.S.
1- I'm still curious to know how the C900 can downclock without supporting speed step.
2- If we were comparing a C2D (65nm) to a C2D (45nm) I don't think there is too much to argue about .

 

As much as I love reading both of your mile long posts, I don't see anything in these last several posts that pertain to answering the op's question. So please let's get back on topic.

 

It is clearly on the topic of upgrading the OP's laptop, comparing his Celeron 900 to a 65nm Core 2 Duo or a 45nm dual core chip as an upgrade.

 

Understood

 

Maybe they could move the secondary discussion to another thread and continue it as I know I would like to see how it all shakes out. One of the moderators should be able to do that for us.