Intel Quad Q9000 or Intel T9800

I am planning on buying a Sager notebook and I noticed that the T9800 is actually more expensive than the Q9000. I am good with computers but I haven't checked the specs of each one, why would the T9800 be more expensive? Does it actually perform better than the Q9000?

 

In terms of raw processing power, the Q9000 should be the more powerful chip, since it has four cores. Nonetheless in most applications, the T9800 will have the edge, since many programs still aren't optimized for four and more cores.

Core 2 Quad Q9000 2000 2/3072 1066
Core 2 Duo T9800 2933 6144 1066

 

so at this point it will be better to get the T9800? But as time goes on and applications get optimized for four cores, wouldn't the Q9000 be better on the long run?

 

You will not notice differences in clockspeed in everyday applications. In terms of raw performance as Commander Wolf noted, the Q9000 will obliterate the T9800.

What will you be doing with this machine?

 

If developers get off their butts and actually get this whole multi-core optimization business going, then yes, the Q9000 should be better in the long run. As of now, the T9800 is the more practical choice, unless you've got some heavily-used application that's known to make use of all four cores.

 

well, I am mainly going to use it for gaming (Fallout 3, Crysis), for CAD, and probably video editing too

 

gaming will be more benefited by the T9800, although to be honest both processors will be able to game fine.

for CAD and video editing, depending on what app you are talking about the quad core could help on your CAD work though

 

Even in fully threaded apps the Q9000 has a lot of ground to make up. You should look up some quad vs duo benchmarks.

I haven't been able to find a test where a Core 2 Quad outpaces a Core 2 Duo that nearly a full 1Ghz advantage, gaming or otherwise.

 

I would get the q9000.

 

I have Intel® Centrino® 2; Core™2 Duo T9600 1066 MHz 2.80 GHz 6 MB, on my notebook. It is great but, I think Im gonig to upgrade it to Quad 9000 becouse on exsample GTA IV my cpu ussage is %97
I play the game at highest level with NVIDIA® GeForce® 9800M GTX 1,024 MB with TurboCache

 

remember there is overheard when it comes to running a Quad(the Duo is faster because some CPU time is used to maintain the Quad). while this overhead is small it combined with the 1Ghz advantage of the Duo, the Duo will be faster likely in everyway, if the Quad is faster in anything, it will be a very small margin. this includes video editing. as for your 97% usage is because you are using a mobile CPU, which is far less powerful than a desktop cpu. a Q9000 isn't going to help at all. now an i7/P2 Desktop will.

 

Get the quad core. They'll start rolling out more applications soon. Dual core will be the thing of the past in a year.

 

A QC cpu is a definitive advantage in programs that require plenty of CPU power.
CAD would benefit most indeed along with video editing programs.

The downside is that T9800 is clocked at 2.96 Ghz, while Q9000 at 2.0 Ghz.
For the immediate time frame, T9800 would likely be a better choice given it's speed ... but long term, and IF games become optimized to make better use of multi-core cpu's, then a quad core would be a better choice.

Given the fact how CAD and video editing is in your plans, then I'd go with the quad core option.
Incidentally, it's also a cheaper solution.

On a side note: I'm using a P7350 c2d cpu (2.0 Ghz Centrino2) and it's most definitely not a bottleneck in most games I play.

So performance/price wise ... it's a win-win situation for you if you go with the Quad Core.

 

Quad , I have a X-9000 , which is blazing fast , but if a quad were available for my chipset I would have gotten it . Quad is futurproof

 

I still think 2.93x2 is a better option.

 

For gaming now yes x2 is most probably better, video = Quad CAD x2 might have a lead now , but in 1 or 2 years ?
Your right , get the x2 , worry about quad in a year or 2 .

 

even in 2years I can't see the Quad out pacing that x2 chip by a significant margin. simply it's a 2.0Ghz Quad, if it was a 2.4 or 2.53 I would have said get the quad.

 

Exactly. This is a 2Ghz Quad vs a ~3Ghz Dual Core.

Anytime someone with a Q9000 wants to run some benches against my T9800, I'll be around.

 

I'm as big a proponent as any of quad core processors, I love my OC'd Q6600, but lets put this in perspective. On the desktop side, there are valid arguments to be made that at stock an E8400 (3.0 Ghz) is better for gaming and light audio/video/3D editing than a Q6600 (2.4 GHz).

If 600 MHz is a big difference, 1 GHz is a huge difference. The only reason the lower clocked quad vs higher clocked dual debate is cooled off on the desktop side is that every one of Intel's quads can hit at least 3.2 Ghz on air which completely closes the gap between the E8xxx series.

On the mobile side, where heat and power is an issue, you simply can't overcome 1 GHz of clock speed. The T9800 will perform better in games and likely almost identical in 4+ core optimized applications. Do a search for E8400 vs Q6600 benchmarks and you'll see what I mean, at stock even that matchup is a tossup - T9800 might be faster in everything than the Q9000.

Let me do some statistics here in closing. The Q9000 is 66% as fast in dual threaded apps as the T9800. If we assume non-perfect scaling, lets just say the T9800 is 30% faster real world for the 34% clock advantage. This is virtually guaranteed, and most games are dual optimized, not quad. If we assume perfect scaling, a quad app will make the Q9000 50% faster, at the same clock speeds. But it's having to overcome 34% lower clocks already. If we extrapolate real world performance, where quad apps maybe net you 20-40% boost, we can deduce that the Q9000 will split half and half with the T9800 even on quad optimized applications. Quite simply, the Q9000 vs T9800 is no contest, the T9800 is better all around, even in quad optimized applications since you never will see perfect quad scaling.

 

Well it is kind of a tough choice, I wish i could have an extra $500 so I can order the NP5797 with a Q9100 or XQ9100. Anyways, I got another question too, What does the TurboMemory Module does? is it only for shorter boot time? or does it also works as RAM?

 

If Q9100 vs T9800, i choose Q9100.

 

screw gaming, CAD takes all cpu power. Go for the Q9000, it offers more power, and will be much faster at rendering than the T9800

K-TRON

 

t9800 is better for games, whilst Q9000 would be more beneficial for heavy CAD works. Just decide what's more important for you.

 

I do a lot of audio/video editing and quad makes a huge (2x) difference since I'm converting 4 files simultaneously instead of 2.

 

You guys don't get it.

Assuming perfect scaling of course in applications that support 4 cores. I realize you can't quantify cores as core times megaherz, but we're assuming perfect scaling here for the illustration.

Q9000 = 8 GHz theoretical
T9800 = 5.92 theoretical

Scaling is not perfect, lets assume an 80% drop off per-added core.

Q9000: 6.6 GHz
T9800: 5.3 GHz

In this scenario, you're looking at best at a 20% performance increase by having a very low clocked quad core versus a high clocked dual core, in a highly multithreaded application.

So you either choose a 30-35% boost in single to dual threaded applications (The vast majority of everything) or a theoretical boost of 15-25% in very few fully multithreaded applications.

I like quads, but the Q9000 is just abysmally clocked.

 

Yeah but you're @ 2.66Ghz stock.

 

I dont think that formula is fair. Scaling it worse as number of cores increase. So if dual cores scale only 80%, quad cores should scale only 80% of 80%, which translates to 64%.

I would go with T9800 as it operates cooler than Q9000.

I would wait for clarksfield(Nehalem) before migrating to quad core or if intel releases 35W quad core.

 

who said scaling is bad with one cpu?
This is not true in apps which support multiple cores

On multiple cpu systems scaling becomes an issue.
For example my ASUS workstation board is rated at 80& efficiency between both quad core processors. Meanwhile Supermicro rates theirs at 85% and Tyan rates theirs as 95%.
When I am rendering my system takes almost exactly a quarter of the time as my core 2 duo.

K-TRON

 

yes but for those that run their q6600 at 3.2 i run my e8500 at 4.4 every day 24/7 and temps are perfectly fine. so then we are right back to basically the quad at 2.4 and the e8400 at 3.0 scenario you described..

imo the t9800 will be plenty even for a while. no way the q9000 is going to outpace it except in a true app that actually fully uses all 4 cores and even then at a 1.0 advantage to the t9800 i still dont think the margin will be anything significant.

quad yes is more "futureproof" if you want to go that route for sure

 

....I'm not arguing with you, I'm agreeing with you. At least with desktops you can overclock a quad to 600 MHz over Core 2 Duo stock, you can't do that with a notebook.

Regarding scaling K-TRON, it has nothing to do with ability of cores to function simultaneously and nothing to do with the motherboard. Programs themselves are not written to run with perfect scaling for number of cores. With the exception of wPrime and some synthetic benchmarks, and maybe some media encoding programs like dBpoweramp, you won't see over 80%.

 

Hi to all.

I want to say that my HDX18 1180el with a Q9000 2ghz its much faster than a Q6600 desktop.
(HDX18 has DDR3 memory)

Its simplistic to compare this data (because of different configuration and OS) but in general, in rendering a scene with 600.000 polygons :
Desktop Q6600 its 24seconds (32bit OS)
and HDX19 Q9000 is 20seconds (64bit OS)

Hope its useful.

cya

 

it doesn't have ddr3 memory. and it's slower than a stock q6600, fullstop.

 

In some SSE4.1 optimised tasks, I think it might be possible for later generation Q9000 to take on Q6600.

 

The Q6600 can easily be OCed to 3+ GHz. It clearly destroys the Q9000 even at stock speeds.

 

oced or not, an extra instruction set is not going to make up for 400mhz and 2mb cache difference.

 

What are we comparing? I do't see any point of bringing on desktop CPU when the title concerns two mobile chip.

First off, Q9000 is mobility chip, its meant to stay clocked at boring 2Ghz in a laptop; likewise, to be fair Q6600 would stays at 2.4 ghz in one and only Clevo lappy due to locked bios (unless you pin mod), thus no matter how good G0 stepping is for overclocking, its not relevant here.

Secondly, Q9000 mobile chip draws less than halve of power of Desktop Q6600 at maximum, according to TDP (45w vs 95w); they are not meant to be compared clock per clock directly, its like apple and orange.

 

sorry but it is DDR3 memory.

DIMM1 and 2: Micron 16JSF25664HY-1G1D1
Type of Memory: DDR3 SDRAM
Velocity DDR3-1066 (533 MHz)
http://www.micron.com/products/partdetail?part=MT16JSF25664HY-1G1


My pc works 24/7 and I never OC a computer because of stability and safety.
I compare computer "as is".
I want to tell you that my HD18 is far fast than my desktop in any circumstance; the only limitation its the hard disk as 5400rpm but I wanto to substitute.

Anyway at full render (3 days) the HDX18 have a temperature max 66-67° and its really silent than my old HP8460ea Pavillion (P4 3.2ghz).
Its much silent than my Q6600.

Hope is good.
bye.

 

Quads "future proofing" isn't worth the currently inferior performance and extra power draw.

just my 2 cents
_Nate

 

Yeah in three years when 4 cores is useful in all applications you'll be getting a new laptop with an i7mobile chip anyway...

 

Definitely the Q for me. Multiple tasks at once, and future gaming.

 

hi guys
i just ordered t9800 and it's ES version
what do u think about engeneering sample ?

 

Its a gamble. It might be an excellent performer, average or absolute rubbish. I hope it will turn out to be a good one

 

It's no gamble, if you buy from a reputable seller.

As long it's QS with E0 stepping, you'll have no problems. My T9800 is an ES, and I've had zero problems.

 

yeah it says E0
is it worth to overclock it on 3000 ?

 

How does it perform in real GTA IV gaming?

The 97% CPU usage you mention I assume is from the in game GTA IV benchmark. If you upgrade to the Q9000 you may get better performance in GTA IV because it is optimised for Quads but wouldn't the vast majority of your other games suffer performance loss since they can't take advantage of the Quad?

To your other games, it will be like having a 2.0 Ghz Core 2 Duo and might be worse overall.

I wonder how much performance improvement you would get if you instead upgraded your 9800M to the GTX260M (im assuming you can do this). I know GTA IV is more CPU intensive though.

 

First time poster here and hope you all can continue to provide valuable input! This thread closely resembles my dilema right now as I'm in the market for a new notebook. Would your opinions change at all if the q9000 processor is configured with 6GB of RAM instead of 4GB RAM for the T9800? I would be using the notebook for video editing and very little gaming (yes, all work and no play does make me a dull boy ).

 

The Q9000 will undoubtedly be the better processor for video editing. More memory might make a difference if you're working with big files.

 

Video editing is composed of 2 things.

1.) The editing usually in a non linear editor (I use Sony Vegas)

2.) The encoding of the final product.

In part 1 RAM does have a function, in the case of Vegas the more ram you have the longer your preview will be for the "ram preview" function. You can get away with 4gb of ram easy, but if you have 6 you can have a larger preview.

Running filters and things is mostly going to use the hdd and cpu though.

Part 2 the encoding is all CPU and HDD. If you use a high quality H264 AAC encode like I do the HDD wont be an issue because even with a 3.6ghz Q6600 on my desktop the file encoding process is so intense that it can take over 30 minutes to do a 10 minute video encode and it will max out all 4 cores on the cpu for the entire time.

This is where a quad core really shines, it will literally cut the time in half from a dual core of the same clock speed.