T43p graphics obsolete?

[aamsel]

I just ran the 64MB 9600 based 2378FVU on the identical version of 3dmark05. 256MB would not complete the test (out of video memory error).
With 768MB I got a 688, just below what you got with 512MB on the X300.

Andrew
Austin, TX

I am using the most recent version of 3dmark05, version 1.2.0. I just ran the benchmark again and I received a score of 712. I have not received my extra memory stick yet so this is basically a stock system. I am not sure if more memory will help or if there is anything I can do to increase the number but this performance seems about 300 points less than what was posted in that other link for an X300 based system.

Intel 2.0, 512mb ram, sxga+, 80gb HD (5400rpm), Multiburner

trikster2 said: Craig. You need to be using 3DMark 2005. If you are using an earlier version the numbers won't match

[baraider]

the question now is what is the score of your t43 when you receive it...if with 512mb stock ram and score more than 700 then x300 is better than 9600

[aamsel]

And I got the T43 tonight from UPS, so I will post that in the morning. My guess is that the 9600 and the X300 are going to be virtually identical.

Andrew
Austin, TX

the question now is what is the score of your t43 when you receive it...if with 512mb stock ram and score more than 700 then x300 is better than 9600

[baraider]

nice, make sure you post Lots of PICs....do some review...etc...can't wait to see that model and how it compare to the t42 you got before

[aamsel]

As far as pictures, it looks like any other T series 14.1" that I have ever seen. I still have the 2378FVU here, so I will note any obvious differences. First impression, however, you couldn't tell them apart unless someone told you which was which.

(Now, this may sound a little "nutty" (and it is off-topic) but I still kinda miss the X31 that I had a day to play with. It was the most solid notebook I have ever seen, I liked it better than any X40 I have ever seen. Even with one speaker, and no optical, it may be my favorite notebook ever.)

Andrew
Austin, TX

nice, make sure you post Lots of PICs....do some review...etc...can't wait to see that model and how it compare to the t42 you got before

[Tsuioku]

I should have noted that I have a T43P 2668G1U on order, so this is a moot point, unless other notebook manufactuers were including something significantly better.

I gathered from a previous post that the graphics adapter is soldered to the MB? I really doubt I would ever replace it, but one of the significant advantages I saw to the PCIe technology was the ability to update the graphics adapter on the laptop if something new came along.

Don't confuse PCIe as being able to swap chips... What you are thinking of is MXM technology and I haven't heard of a laptop that offers this yet...
PCIe is simply a faster standard (16x) compared to AGP slots (up to 8x?)

Regarding 9600 vs X300.. I believe the 9600 is better since it has faster memory and a higher bandwidth (assuming that they are similar to their desktop counter parts) SEE EDIT BELOW
A rundown of ATI vs nVidia can be found here [T2 = 9600, V3100 = X300]
http://graphics.tomshardware.com/graphi ... ie-10.html

I also remember reading some reviews saying that the 9600 has a higher framerate.. can't find it atm.. This posting might be useful though when comparing ATI cards..
http://www.notebookreview.com/forums/to ... C_ID=12185


EDIT: After a little research, turns out that the desktop counterparts are quite different from the mobile solutions. I'll still leave the link up for those that want some sort of reference to compare FireGL cards. My new conclusion is that the 9600 ~ X300

Here's is a "better" link for comparing mobile cards
http://www.rojakpot.com/default.aspx?lo ... =98&var2=0

I say "better" because the specs listed for the X300 seem weird.
Desktop specs are: 0.11 micron, 1300 MTexel/s, 6.4 GB/s...
Mobile specs (from Rojak): 0.13 micron, 1400 MTexel/s, 9.6 GB/s
ie The Mobile ver is better than the Desktop ver??
(I know the manufacturing process is not completely relevent to performance, but I just find it weird as to why it would be different)

[craigg]

I received the scores in the low to mid 700 range with the stock 512mb of memory. I will test it again on Monday with 1GB of memory. Judging by the tests the X300 has slightly better performance over the older ATI 9600.

[aamsel]

I spent more than several hours comparing the 266846U to the 2378FVU today.
I ran a variety of benchmarks and did some “eyeball comparison” of the two.
I am not about to post all of the benchmark results, but I will post the most important data, and my conclusions.

I tested them both right out of the box, except that I upgraded the 2378 to 512MB. So, each had 512MB. I did not put my 7200rpm drive in either of them, so they had the factory pre-load. The 2378 has a Hitachi 5K80 40GB 5400rpm, and the 2668 has a Hitachi 5K100 40GB 5400rpm. For whatever reason, the older 5K80 benched a hair faster, but I am not going to be using it anyhow, so that is irrelevant.

I set power to be full-power on A/C. The 2668 ran at 1.86GHz, and the 2378 at 1.7GHz.

To finish up with the question of the 3DMark05 scores:
2668 with 64MB X300 got a score of 710.
2378 with 64MB 9600 got a score of 688.
(no real difference).

Then I ran Aquamark version 3 benchmark:
2378 got a score of 19,530.
2668 got a score of 17,269.
(not sure why, or how accurate Aquamark is, although it is supposed to be “real world” based.

I then ran all of the SiSoft Sandra Lite 2005 SR1 benchmarks:
I won’t post all of the scores (too many, too boring).
In general, the 2668 beat the 2378 by about 10% (not surprising.)
This included CPU Arithmetic, CPU Multimedia.
However, the 2668 TROUNCED the 2378 in all memory-related testing.
It beat the 2378 BY ABOUT A THIRD, in Memory Bandwidth, Cache and Memory Bandwidth !!!
Everything else was either a wash or a 10% victory to the 2668.

I also ran some older versions of 3DMark. I realize that some of the techniques used for these tests is obsolete but some is still valid, I believe.

Anyhow:

3DMark03 v.3.6.0
--------------------
2668 got 1909 3DMarks
2378 got 2508 3DMarks !!!!
(the 2378 put the 2668 to bed!)

This is a DirectX 9.0 enabled benchmark


3DMark01 Second Edition Build 330
----------------------------------------
2668 got 7021 3DMarks
2378 got 9779 3DMarks

This older version of 3DMark uses DirectX 8.1
However, the 2668 got smoked by the 2378 on this as well.

General:

I looked at color purity of the screens.
Each has a Samsung panel. The 2378 has FRU # 92P6760 and the 2668 has FRU # 11P8348. Not sure what the difference is. Neither have any dead pixels. Both have the little light leakage on the right-side of the LCD frame, but you have to put your nose up to it to see it! Both are darker on the top than the bottom, due to the backlighting method.
Color clarity was about the same, the pure green looked a hair better on the 2378. Pure green seems to be the hardest to get nice on an LCD panel from what I have seen.

Haven’t compared the sound yet, but I expect they are identical.

Each had the same DVD/CD-RW.

The 2668 came with SONY 6 cell FRU # 92P1089 and a free Panasonic 6 cell promotion. Neither of them latches in as solidly as the SONY in the 2378 does. Need to do the heat-gun trick. I was surprised that even the SONY was not snug, but no big deal there.

Quality of fit, finish and assembly appear to also be identical. There are no substantial gaps in either.

The fan in the 2668 is louder on startup, but then quiets down a few moments after Windows boots. It is a faster, hotter CPU after all.

NOW...I REALLY don't know what to do !!!
But...I have to return (at least) one of them.

I think I would have favored the 2668 until the older 3DMarks and Aquamark all favored the 2378.

CPU is going to favor the 2668 since it is, of course a faster CPU.

Memory also favors the 2668.

Your opinions, comments, and snickering will all be more than welcomed!

(also, I ran more benchmarks and posted more benchmark scores than I ever intended to when I woke up today...SORRY.)

Andrew
Austin, TX

[craigg]

Nice post Andrew. Your results seem to duplicate my results only I was able to get the numbers a little higher with some tweaks.

I have a spare 60gb 7200rpm drive here and I expect to receive my additional memory on Monday so I will run the test again with this enhanced configuration.

The final score should pretty much mirror what one should expect from a new T43p. I am curious if 1gb will make a big difference. It is possible that the 3dmark is memory sensitive since it was posted that it crashed when run with only 256mb.

[trikster2]

Aamsel


Just want to thank you for the work! Those comparisons must have taken a good chunk of your evening. Thanks for giving us some more data to make T43/T43 buying decisions on.

You should post it a s a main topic, T43/T42 comparision for all to enjoy!

Now we just need the same for the T43P when it comes out and and we will have all the info we need.

Thanks again!

[aamsel]

The T43p is going to have much better video than the X300.

Andrew
Austin, TX

...The final score should pretty much mirror what one should expect from a new T43p...

[baraider]

aamsel,

it's interesting to know that you got the free pana 6-cell promotion. i think it applies for edu and other programs only, not applicable for epp. oh well...

All things considered, your t43 has only 512mb ram while the t42 has 256m more....if you take the 256mb out, then i think you should see a bigger difference.

Again, compare the price of the t43 and t42, i say keep the t43...( the 42 is at its 30 day end while you can keep the 43 for another 30 days)

[aamsel]

I may hang onto the T43, but my guess is that the T42 is faster at 3D.

I am going to drop the 2378 to 512MB tomorrow, and repost the results.
I doubt that it will have much effect.

My opinion right now is that the 64MB 9600 is superior to the 64MB X300.

Here is what is really bothering me:
1.) IBM has a faster CPU and a faster memory controller which they have put with a very low-end video card (some of the other T43's even have Intel graphics).
2.) I don't usually put a lot of weight on benchmarks, but Aquamark being a higher score for the 2378 really proves that the 9600 is faster in 3D. Aquamark is one of the few benchmarks that graphics chip manufacturers can't mess with. Why and what do I mean?? It has been a known fact that graphics card manufacturers are not dumb. They are well aware of the common benchmark programs that are used by review sites and magazines, and they tweak their hardware, firmware and drivers to get the best benchmark results. Look at the benchmarks from the earlier programs...3dmark03 and 3dmark01. If these tests are still valid, which I think they are, the 9600 blows the X300 away.

Am I getting a Thinkpad for gaming...NO, but a better video card could make a difference down the road, even short-term.

Andrew
Austin, TX

aamsel,

it's interesting to know that you got the free pana 6-cell promotion. i think it applies for edu and other programs only, not applicable for epp. oh well...

All things considered, your t43 has only 512mb ram while the t42 has 256m more....if you take the 256mb out, then i think you should see a bigger difference.

Again, compare the price of the t43 and t42, i say keep the t43...( the 42 is at its 30 day end while you can keep the 43 for another 30 days)

[baraider]

do you pay the same price for your t42 and 43?

True that the x300 is in the lower end of t43 where the 9600 is the one of the top cards in the t42 .

Does it mean anything in everyday task? No because you are not a gamer. Gamers don't buy TP. We buy TP to get our work done and because of the high quality of TP.

Does it feel amiss when we only get x300 in the 43 where we can get 9600 from the 42? Yes, but that's part of life when you are early adopter.

Remember that the resell value of the t43 is higher than the 42 while you paid the same price when you buy both of it.

[danny_isr]

do you pay the same price for your t42 and 43?

True that the x300 is in the lower end of t43 where the 9600 is the one of the top cards in the t42 .

Does it mean anything in everyday task? No because you are not a gamer. Gamers don't buy TP. We buy TP to get our work done and because of the high quality of TP.

Does it feel amiss when we only get x300 in the 43 where we can get 9600 from the 42? Yes, but that's part of life when you are early adopter.

Remember that the resell value of the t43 is higher than the 42 while you paid the same price when you buy both of it.

you talking like the 9600 is much superior to the x300.
from reading this review i'm not convinced at all that it's the case.

another thing , on the T43 you get other things like PCI-E , DDR2 , 533FSB better WiFI/Blue , newer model . lets say that the x300 is slight slower (not sure yet at all might be faster). you still got other things. i woudn't keep the 42. just my opinon.

And another thing , i'm really not sure about it at all , but how both will run underlonghorn ? will the T43 get bigger advantage there (PCI-X) ??

Danny

[aamsel]

That is all (pretty much) true.

Andrew
Austin, TX

do you pay the same price for your t42 and 43?

True that the x300 is in the lower end of t43 where the 9600 is the one of the top cards in the t42 .

Does it mean anything in everyday task? No because you are not a gamer. Gamers don't buy TP. We buy TP to get our work done and because of the high quality of TP.

Does it feel amiss when we only get x300 in the 43 where we can get 9600 from the 42? Yes, but that's part of life when you are early adopter.

Remember that the resell value of the t43 is higher than the 42 while you paid the same price when you buy both of it.

[aamsel]

The requirements for Longhorn (as we know it today) are having DirectX 9.0, so both should work. The T43 model I got has the same Intel b/g card that my T42 had. (mainly because the darned salesman told me that I couldn't upgrade it, which was wrong).

Andrew
Austin, TX

...another thing , on the T43 you get other things like PCI-E , DDR2 , 533FSB better WiFI/Blue , newer model . lets say that the x300 is slight slower (not sure yet at all might be faster). you still got other things. i woudn't keep the 42. just my opinon.

And another thing , i'm really not sure about it at all , but how both will run underlonghorn ? will the T43 get bigger advantage there (PCI-X) ??

Danny

[baraider]

you talking like the 9600 is much superior to the x300.
from reading this review i'm not convinced at all that it's the case.

Danny

Danny boy,

You are right, the feeling around here for the last couple of days is that the x300 is inferior to the 9600....i have no idea why, how and if it is true or not.

the x300 is just so new and not mainstream yet.....we need to wait couple weeks or months to have a fully credible report.

I guess because 9600 is better than 7500,9000,9200...while the x300 is the first card so it tends to discredit the x300

But i have the felling that the x300 is at least equal to 9600 if not much better...

My next machine won't be a t42...that's for sure... because i have a wider range of discount.

[aamsel]

I just took the extra 256MB out of the 2378, so both systems had 512MB.
I updated the results above. There was virtually no difference in scores.
The 3DMark05 score was identical for the 2378 with 512MB and 768MB.

Apparently, you need 512MB to run these benchmarks properly, but additional memory is not helpful.

For those thinking that 1GB will make their benchmark scores higher, it won't.

Andrew
Austin, TX

[Conmee]

Folks,

Based on the scores I've generated on a T42 and T42p, and the scores I've seen posted for the X300, I wouldn't trade up or exchange a T42/T42p for a comparable T43/T43p with X300. While the X300 hardware specs are, on paper, superior to the MR9600 (same core for the FireGL T2), there is always a huge difference in theoretical v. real-world numbers. In fact, IBM and other OEMs tend to down-clock both core and memory in many implementations. The specs ATI gives are what OEMs would implement, assuming no power or heat constraints. And as we all know, that is never the case. And even ATI's marketing literature indicates that the X300 is the entry level thin and light solution. From what I've seen, the X300 is slightly below to slightly above comparable MR9600-based solutions in terms of benchmark results based on various clockings/implementations. Only an X600/X700/X800 solution would compel me to upgrade from my current FireGL T2, and even then I think I'd only be swayed by the X700 (I've heard and seen very good things about this GPU).

Not sure why your scores for the MR9600 w/64MB are so low. Here are my results from my tests (also have my T42p results in my sig) and the MR9600 compares favorable with the FireGL T2, which makes sense, given the same GPU cores, only different drivers and amount of VRAM.

2373-CYU (1.8GHz, 1GB RAM-Upgraded, 80GB 5400RPM, MR9600/64MB):

3DMark01 (3.3.0) - 9609
3DMark03 (3.5.0) - 2617*
3DMark05 - N/A (didn't run this when I had the T42 at the time-last summer)

2373-GRU (1.8GHz, 1GB RAM, 60GB 7200RPM, FireGL T2/128MB):

3DMark01 (3.3.0) - 9629
3DMark03 (3.5.0) - 2629
3DMark05 (1.1.0) - 971


I conducted these tests on the IBM Preload, and had system restore, indexing services, and networking interfaces all turned off/disabled. Power management was set to High Performance.

From all indications, my FireGL 'smokes' the X300 64MB variety. Wouldn't even bother with the X300 since it only comes in discrete, 32MB and 64MB variations. Anyone with an MR9600/FireGL T2 would probably be best served waiting, especially since the amount of VRAM is still more important (for games and high texture CAD stuff) than a modest boost to core/mem speeds and a couple extra pipelines. And as for the PCIe v. AGP throughput, most games and apps don't tax even 4X AGP at this point, so the bus throughput for the time being isn't a necessity. Tom's Hardware did a good article on why PCIe is being pushed so hard... it's really a solution looking for a problem... hence, the big marketing buzz about digital/HD video, etc, needing the PCIe throughput (sorry to digress).

Anyhow, with the prices that T42's are going for with Employee/Shareholder discounts, I wouldn't even think twice about purchasing a T42. If you really want the next-generation notebook with enough future-proofing technology to justify the IBM premium, wait for the dual-core Pentium-M ThinkPads with SATA drives, and minimum 128MB VRAM GPU implementations, late this year and/or same time next year.

And one final final final point... lol.... The T43 doesn't offer all that much in terms of 'other' technologies... a few hundred more Mgz for the CPU, faster FSB, DDR2, PCIe (only the GPU and the ExpressCard slots utilize the PCIe throughput), no SATA (chipset supports it, but there's no way to connect a SATA drive unless it's an external USB/Firewire variety lol), no Azalia advanced audio (supported but not implemented), and what do you have to do to get all this? Pay about a 20-25% premium over similarly spec'd T42's and give up anywhere from 20-40 minutes of battery life. Oh, yes, you will give up battery life for those precious few extra seconds the CPU/FSB/DDR2/PCIe shave off Windows/Word/Doom3 loading. lol And as for Longhorn, the T42 is plenty powerful enough to run the GUI. DirectX 9 and 64MB VRAM are minimum requirements for the 3D environment, but the standard GUI will be available for legacy/slow machines.

Having said all that... I love my ThinkPad and I support IBM and understand the trade-offs they make in terms of battery life, heat dissapation, performance, stability, made for 'serious' work not games, yadda yadda. But when it comes right down to it, IBM has always made incremental updates to the T Series. And this one is no different. There are technologies included in the T43 that are worth getting excited about, just not for the price you'll pay... The T41's were a great deal when the T42's first came out, and the T42's are now a great deal as the T43's ramp up. Wait until October if you want a great price on a T43... of course, the next must-have technologies will be in the T44/T50/T100 or whatever Lenovo/IBM chooses to call the next ThinkPad.



Daniel


* The scores for 3DMark03 (3.5.0 and 3.6.0) and 3DMark05 (1.1.0 and 1.2.0) are the same (see the Futuremark site for details). So I'm not sure why the score for the 2668 and 2378 are so much lower than my scores.

[Kenn]

Just for reference, last time I ran 3DMark05 on the default IBM installation on my t42p, I got just under 1000 as well. The MR9600/FGLT2 is no graphics monster, but it'll do until my next upgrade (which I'm sure will be a sub-4lb 15" with 8-12hour battery life )

[aamsel]

Well, you make some very good points.
However, I don't see where the T42's are less money than the T43's. The 2378 I have with a 3 year warranty added on is within $25 of the price I paid for the 2668. I purchased from the Eduational site.
Can you tell me which 9600 based models have great deals on them now?
(I am not going to pay the extra for a 128MB 9600 based T42p.)

I do agree that the faster memory and slightly faster CPU don't amount to a "hill of beans". When you are on battery, you aren't running max speed anyhow.

Andrew
Austin, TX

...Anyhow, with the prices that T42's are going for with Employee/Shareholder discounts, I wouldn't even think twice about purchasing a T42...Pay about a 20-25% premium over similarly spec'd T42's...the T42's are now a great deal as the T43's ramp up...

[aamsel]

Not sure, unless your 1GB of RAM made a difference??
Also, perhaps our test settings were different. I was on 32bit color depth.
Also, I used the pre-load (as you did) but I didn't have all of the stuff turned off like you did. I was on max power.

What REALLY confuses me is why did the 2668 do just as well (or actually a hair better) than the 2378 on 3dMark05 but way worse in just about everything else?? Why better in 3dMark05 but worse in 3dMark03 and 3dMark01??? Just makes no sense.
(Of course, I didn't write those programs, so I have no idea.)
Also 3dMark01 is just DirectX 8.1.

Andrew
Austin, TX

...The scores for 3DMark03 (3.5.0 and 3.6.0) and 3DMark05 (1.1.0 and 1.2.0) are the same (see the Futuremark site for details). So I'm not sure why the score for the 2668 and 2378 are so much lower than my scores...

[danny_isr]

Folks,

Based on the scores I've generated on a T42 and T42p, and the scores I've seen posted for the X300, I wouldn't trade up or exchange a T42/T42p for a comparable T43/T43p with X300. While the X300 hardware specs are, on paper, superior to the MR9600 (same core for the FireGL T2), there is always a huge difference in theoretical v. real-world numbers. In fact, IBM and other OEMs tend to down-clock both core and memory in many implementations. The specs ATI gives are what OEMs would implement, assuming no power or heat constraints. And as we all know, that is never the case. And even ATI's marketing literature indicates that the X300 is the entry level thin and light solution. From what I've seen, the X300 is slightly below to slightly above comparable MR9600-based solutions in terms of benchmark results based on various clockings/implementations. Only an X600/X700/X800 solution would compel me to upgrade from my current FireGL T2, and even then I think I'd only be swayed by the X700 (I've heard and seen very good things about this GPU).

Not sure why your scores for the MR9600 w/64MB are so low. Here are my results from my tests (also have my T42p results in my sig) and the MR9600 compares favorable with the FireGL T2, which makes sense, given the same GPU cores, only different drivers and amount of VRAM.


From all indications, my FireGL 'smokes' the X300 64MB variety. Wouldn't even bother with the X300 since it only comes in discrete, 32MB and 64MB variations. Anyone with an MR9600/FireGL T2 would probably be best served waiting, especially since the amount of VRAM is still more important (for games and high texture CAD stuff) than a modest boost to core/mem speeds and a couple extra pipelines. And as for the PCIe v. AGP throughput, most games and apps don't tax even 4X AGP at this point, so the bus throughput for the time being isn't a necessity. Tom's Hardware did a good article on why PCIe is being pushed so hard... it's really a solution looking for a problem... hence, the big marketing buzz about digital/HD video, etc, needing the PCIe throughput (sorry to digress).
s.

your tests smoked the 9600 as well correct ?
but in the 9600 you said "Not sure why your scores for the MR9600 w/64MB are so low. Here are my results"

and when you compared to the x300 you said "From all indications, my FireGL 'smokes' the X300 64MB variety."

and second to that ,Fire GL not suppose to be compared to the x300.

T42 guys , you probably happy with your laptops and you have a good reason for that. But why bashing the T43 here on every thread ?
maybe for you it's not good enough upgrade , fine..keep your 42 nothing wrong with that.
but for guys that are looking for a new laptop T43 is excellent.

if i had a 42 now i wouldn’t upgrade. i wouldn’t upgrade even from a 41 actually. what for ??


Danny

[aamsel]

Nobody here is "T43 bashing" in any way, shape or form.
Certainly, I guarantee that I am not. No desire to.

Here is the summary of what I was trying to ascertain:
Right now, both series are available, T42 or T43.
T42 is FULLY available, T43 is available in limited models.

I am someone that has a choice of what I want right now, so it make sense that I am trying to see if the T43 with X300 has features that make it advantageous over a T42 with 9600.

I would be happy to know that either were better.

Right now, it appears that the T43 does not have sufficient advantage over the T42. (note: I am not talking about T42p at all like some people here are).

As far as I can see, IBM released a slightly superior T43 with an inferior video card.

If you are never going to be concerned with 3D performance, then the T43 would still be the way to go.

If the T43 was the only series available new from IBM, then this discussion would be moot.

For me, the 2378FVU is probably still the best overall deal.

Also, another thing I failed to mention:
I tested both models doing DVD playback, and I did find the 64MB 9600 to produce smoother overall video than the 64MB X300.
This is just a personal opinion, and is based on my own observations, but that is the way I saw it.

Andrew
Austin, TX

[baraider]

danny, shorten your quote....

[danny_isr]

danny, shorten your quote....

done


aamsel , you should compare apples to apples . T42P to T43P.
if it's not here yet , then wait with your review.

like i said , i wouldn’t upgrade from the T42 either.

"
If you are never going to be concerned with 3D performance, then the T43 would still be the way to go.
"

you means that x300 will be bad at 3d. But if the performance is similar to the 9600 that was ok/good. how come the x300 is not ?
just because it’s “entry level” now means nothing. Entry level today can be top of the line of yesterday and that will be still very good.

i did play games with it , run really good with everything turned on . i still didn’t try Doom3 yet but i will.

all those benchmarks are just benchmarks . they are not real world for me.
Play a game , see what the frame rate . that will be the test.
so bottom line , no one knows yet for sure how it performs .
but opinions we got here plenty.

Price wise , with epp T43 was VERY close to T42 , for me it was easy decision.

Just my opinion.

Danny

[baraider]

i agree, with epp, the price is close or better than t42. For people without epp (99% here), their price structure is difference so it won't make sense to get a more expensive 43, the 42 makes so much sense now because it's getting cheaper now ...

Again, depends on what you can get, either 43 or 42 will make more sense to you.

[aamsel]

I DID compare "apples to apples".
I compared T42 to T43 (2378 to 2668).

I never mentioned any T42p or T43p whatsoever.

It was someone else above that tossed in T42p data.

I have never had an interest in paying the premium for a "p" model in any series.

I am simply trying to find my best "bang for the buck".

If resale value was my only concern I would go for the T43.
If slight CPU and memory speed increase mattered to me, I would also give the nod to the T43.

However, the T42 may "go down in history" as being the last "pre-Lenovo" T series Thinkpad, and the T43 may be thought of as the start of the Lenovo products. We don't know yet how this is going to play out.

I am simply "going by my gut" saying that I like the graphics better on the T42 with 64MB 9600.

I don't give any real weight to benchmarks either, and I am kind of stunned that I took the time to run them. It was just "curious" that the X300 did not do any better than it did.

As for DDR-2 memory, I have read the reviews of it on the web, and the speed differences (real world) are 5 to 15 percent.

The 9600 card right now is a "tried and true" middle-of-the-road AGP card.
The X300 is the entry-level PCI Express.

If IBM had gone with the X600 to replace the 9600's, then there would be no contest, and I would have undoubtedly preferred it to the 9600. But, by using the X300, they are kind of "cheaping out" like they have done for so long with the 7500 cards.

Again, theses are just gut-reaction personal opinions.

I am not telling ANYONE to base their decisions on my thought processes of findings.

Andrew
Austin, TX

...aamsel , you should compare apples to apples . T42P to T43P...if it's not here yet , then wait with your review...
like i said , i wouldn’t upgrade from the T42 either.

you means that x300 will be bad at 3d. But if the performance is similar to the 9600 that was ok/good. how come the x300 is not ?
just because it’s “entry level” now means nothing. Entry level today can be top of the line of yesterday and that will be still very good.

[aamsel]

Correct, I am buying off the Educational site, and the T43 is not more attractive. Also, there are no higher-end T43's available to me.
(no, I have not checked with Bill, but he is not set up for returns or credit cards)

If the T43 had an X600 (not an X300) then it would be a "done deal", but that is not the case right now.

I had read MANY reviews saying that the X300 replaced the 9200, and that the X600 replaced the 9600 (and was on a par with the 9700). There was the one website that claimed higher fill-rate and bandwidth for the X300 than the 9600, but that was just ATI marketing blurb that was posted.
EVERY OTHER REVIEW I have read says that the 9600 is superior to the X300, period. Every one.

Andrew
Austin, TX

i agree, with epp, the price is close or better than t42. For people without epp (99% here), their price structure is difference so it won't make sense to get a more expensive 43, the 42 makes so much sense now because it's getting cheaper now ...

Again, depends on what you can get, either 43 or 42 will make more sense to you.