5720 now... or dual core PM... when?

Im in the same boat as you. I wanna pick up the 5720 or something comparable to it. But Im worried that if I dont get a 64 bit processor then I will be shit out of luck in a year or so. Fortunatley I wont be buying until mid december, and by then I should be able to see what the 9750 with the AMD 64 bit processor and 7800gtx looks like.

I'm waiting. Apparently the newest (Q3 2006) Merom mobile chips should be pin for pin compatable with Yonah notebooks (albeit with a bios update). I'm buying sometime around January-March, and my notebook needs to last 4 years, so I figure no rush for now, my desktop and Latitude C600 should last me another few months.

Up to you, though.

Dual core is NOT going to benefit you at all if your just looking to game. If you want to Multi-Task, now that is another story. Dual Core will allow you to MT amazingly.

If your a gamer the GPU is doing almost all of the work and there is currently no sense in coding a game in dual core. I see no sense in coding games this way. I am not on the forefront of gaming technology and there maybe something I am missing. However, Dual Core is a multi-tasking technology for the foreseeable future and as a consumer, I would make a purchase based on what your going to do with your machine most...

Just my 2 cents...

meh, im a mild gamer, mostly going to be using internet and stuff for school. im not waiting, im buying my 5320 within the week.

Dual - Core 64 Bit AMD Turions in Q1 '06, thats what im waiting for =P

it would be possible to have one cpu do AI while the other does physics, or something like that. right now, one core is being totally wasted, so there is a definite advantage ot creating dual core games.

also, yonah will have a faster FSB, so taht might help with games. merom will have a 4mb l2 cache and be 64bit, which will also help.

and i too am waiting for those turion benchmarks...

hmmm, i feel sure in about a year y'all will have machines that will make me weep....but for today the 5720 kicks behind all the way down the street........

review up

its about 86.78% done but has some good info. Check it out under reviews and let me know what you think! or or or or or:

IF dual core becomes popular, you'll see games take advantage of it... and it will be popular.

As games become more and more complex this could be a possibility, however the GPU is still doing most of the work. This is gradually changing with games like BF2. I think it's MUCH more likely that DUAL CORE GPU's will become the norm before dual core CPU's with regard to gaming. Its really the next evolution to SLI...

L2 Cache allows a processor to be more efficient, period. The more instructions that can be stored in the processor, the more the processor can aleviate the need to access mainboard memory. This is why a 3.6 GHz P4 w/1Mb L2 Cache is no faster for gaming then a Pentium M @ 2.13 GHz w/2Mb of L2 cache. Now combine this massive on-board L2 Cache with the beautiful short pipeline architecture in the PM and your data gets processed even faster. This is the real secret behind the PM, the P4 HT does very well with multi-tasking, which is to be expected with the incorporation of HT. However, the Prescott's long pipelines and excessive heat kill it in the end. So I do agree with you, a Larger L2 cache really is the future for CPU's as the more GHz's we want to atain will require transistors to continually shrink to continue with the GHz maddness we see today with 4GHz machines. So higher GHz machines = More and smaller transistors to achieve continually higher speeds, smaller transistors produce less heat due to less voltage required.

So the only way to increase GHz is to increase the number of transistors on a CPU, so in order to keep increasing GHz, we must continually make transistors smaller to produce less heat. This is how 3+ GHz CPU's use the relatively small voltage they consume. Problem: We are reaching a point where producing chips with extremely small transistors is getting too costly. Bottom line, why use GHz and molecular (j/k) sized transistors to achieve performance when chip designers can do the same thing with a lower clocked chip that is more efficiently designed (PM, shorter pipelines, huge L2). Now combine this with Dual Core, larger L2, shorter pipelines, and 64 bit technology = Get more info per clock cycle processed = Efficiency

Come on 4Mb L2!!!

great post codename!

One question for ya: It was my understanding that as wth Mh we are also reaching the point of diminishing returns on the L2 cache. Larger and larger chunks o' data can be stored (and heres the question) but can they accessed by the cpu in a manner that uses it all?

totally!

the current pm's have 2mb of l2 cache for a single core, merom will have 4mb of cache for 2 cores. this means that each core will still only have 2mb of cache. however, if you are gaming and only one core is being used, then it will have all 4mb to itself. combined with meroms higher clockspeed(2.6ghz ive heard, could be waaay off) using 4mb of cahce between two cores should hardly be a problem, especially in multithreaded apps.

in games, 4mb of cache on one core can still improve it over 2mb. look at the pentium 4 extreme edition, which has 3mb of l3(slower than l2) cache. it is quite a bit faster than other pentium 4's, although it is way overpriced.

I'm curious to see what kind of performance the single-core Yonah's will bring to the table

The answer to your question is "Yes", the real limitation has been the size of the L2 cache that can be placed on a chip. Remember, more L2 means less room for the processor (The modern processor needs many transistors for the complex amount of logic gates we use today, more CPU = more transistors, more logic gates). So in essence, you have to have enough room for the processor's design. L2 is cache, there must also be a highway for that cache to get processed. Chip makers only place a cache on a chip that can be completely utilized, the real limitation is logic to storage on a single chip...

crap, because from what ive seen, the highest clocked one will be 1.66ghz. this could be an option though, considering that many yonah purchasers will want to upgrade to merom later on, so this could minimize the amound ot money lost in the upgrade.

4Mb of L2 is an absolutely amazing feat of technology. I am ASSuming here, but I am betting that the L2 cache of 4Mb you're speaking of is (1) large 4Mb cache and NOT (2) seperate caches for each core. This would explain why the CPU could access the Aggregate cache value (4Mb) when not running a multi-threaded app as you mentioned. Have not kept up with Dual Core to be honest.

You're dead right about P4 EE...

drizek and codename:

good stuff, there! I guess the real point of my question is the (and this has been mostly answered) pipeline on the chip big enough to use all that L2. One to many an Intel marketing scheme has made me leery. My fabulous home Prescott core chip is getting beat down by my P3 core P4M, which is partly due to the L2 cache. I just have this picture in my head of a dual core pentium with a tumurous growth of 1 gig L2 sticking out