The new conroe cpu?

I read that the new conroe cpu is coming soon. Should I wait or stick with amd? deciding on a new comp is so hard. I mean I dont want the best ones on the market cause I can't offord it but I want something that will last for a few years.

Intel's Conroe will be coming in 6ish months. If you can wait, then that means you don't need a new computer, so wait. If you can't wait, then get an AMD.

Longer you wait, the better the computer will be in the end.

Conroe will have the new architecture based off of the M series. You can look at the benchmarks in tomshardware when they ran an M series as a desktop processor. That will give you some kind of idea of what to expect. It won't be exact but it'll be like a ballpark idea.

"The Pentium M architecture has no trouble reaching clock speeds of about 2.5 GHz. At this frequency, "the little chip that could" even outperforms Intel's P4 Extreme Edition and AMD's flagship Athlon 64 FX in 3D games. In other applications, the Pentium M also makes a surprisingly strong showing, despite its lack of the SSE3 instruction set and HT. Increasing the chip's core voltage could probably push it even closer to the 3 GHz mark without incurring the kind of thermal dissipation that has caused the Pentium 4 to fall from grace in the eyes of many readers and hardware testers alike."
http://www.tomshardware.com/2005/0 [...] page3.html
Kind of an old article, but an interesting read.

There are numerous other changes as well in Conroe, so until reviewers actually get their hands on it, no one really knows.

P.S. I'm waiting for a dual-core Conroe. (I currently have an Intel and an AMD system right now).

Its hard to say what it might turn out like since its been worked on for the past years (in other words its not just another pentium m) and conroe is going to be better then yonah, and yonah is close enough to an A64 to be compeditive and this is intels 'next generation' so they might have addressed the issues the the old P6 design failed in (FPU).

The last time the two core designs officially competed was the days of Intel Coppermine and AMD Thinderbird (was it thunderbird?) - Intels CPUs were cold and quick for a slow SDR FSB where as the AMD was a touch quicker and had a faster fsb (DDR) and was on average 60% hotter, and the AMD managed to scale further too (180nm Intels 1ghz max, AMD - 1400) - perhaps we will see the same war again except ~5 years later?

Iv seen benchmarks for a Pentium M @ 3+ghz and 1+ghz fsb (similar to conroe's final clock speeds but not dual core and 1/2 the cache) and it rips everything on the market right now, but conroe will have a longer pipeline (~10 vs ~14) length so not exactly the same.

Either way im waiting until conroe and a few months after to see what AMD has before i buy a new system.

This is all pure speculation.

Conroe should be able to outperform what AMD will have out in H2 2006 in most applications. Yonah can already perform similarly to the X2 at the same clock speed besides the X2s integrated memory controller. While the move to M2 and DDR2 will help increase the X2s performance, Conroe will likewise increase bandwidth from a 667MHz FSB to a 1066MHz FSB. Conroe's doubled cache, higher clock speeds, redesigned execution units, and expanded micro ops and macro ops fusion will likewise allow it to increase its performance over both Yonah and Dothan. The Inquirer notes that if both AMD and Intel execute perfectly, AMD with 65nm and Intel with new core, then 2006 will end with Intel outperforming AMD.

http://www.theinquirer.net/?article=28602

It should be noted that both Merom and Conroe (likely Woodcrest as well since they're the same family) are ahead of schedule. They were roadmapped for Q3 2006, but the latest reports indicates a July launch. As early in Q3 as you can get. The next generation platform will also launch at that time with support for DDR2 800 (which is available unofficially now on some motherboards) or possibly DDR2 1066.

http://www.xbitlabs.com/news/cpu/d [...] 55201.html

Intel isn't aways behind schedule. The current 65nm processors are ahead of schedule or on time. For instance the 955EE Presler was scheduled for launch Q1 2006, which became January 2006, which became December 27, 2005. As well, the rest of Presler, Cedar Mill, and Yonah have all launched today which is as early in Q1 as you can get. Granted they aren't shipping yet, but they should be in the next week or two. The older models will all have their prices cut for clearance by the end of the month.

http://www.theregister.com/2006/01 [...] cedar_mill
http://www.theregister.com/2006/01 [...] _up_yonah/
http://www.theregister.com/2006/01 [...] s_xeon_dp/

The Merom name is used both for the mobile chip and the entire family in general. As such news reports citing Merom are somewhat confusing as to which they are referring to. Supposedly, Conroe and Woodcrest are based on Merom. More likely that evolution is just marketing to show that Conroe, etc. is derived from the mobile architecture of the Pentium M. The entire family would be designed from the ground up otherwise Conroe and Woodcrest would be constrained if they were based too strictly on the power limiting requirements of Merom.

Merom has long been reported to be ahead of schedule as it has already been taped out and in production since June. Only recently has Conroe specifically been reported to be ahead of schedule as well. There haven't been benchmarks released as far as I know. It wouldn't be wise for Intel to release them anyways as it would further canabolize sales of their current processor generation.

Personally overkill for the voltage regulators is preferable as it allows great overclocking potential. As well, multi-phase voltage regulators will further increase power efficiency when not used for overclocking. The ASUS 8-phase voltage regulators increase power efficiency by 10% over 4-phase voltage regulators and allow the 840EE to run 5C cooler. Overperforming voltage regulators would be great for the ViiV computers Intel is now marketing.

You guys are making it extremely hard for me to decide on what to get. I mean I been hearing get AMD and I have been leaning toward AMD. I plan on getting x2 4400. But I kinda want to wait for the M2 socket. Now I'm hearing wait for conroe it will be better than AMD. What the hell i'm suppose to do. Besides how much do you guys think it will cost for the conroe and the m2 socket for amd??? I want a new comp so bad but at the sametime I do want to get one right away. I kinda want to know what's ahead in the next few months. But of course that isn't really easy because even though i'm leaning toward AMD I still have a hard time decideing. I'm really a picky ass person so, if i get something I end up regreting, i'll be pist and shit. But so far, for those people on here that have been answering my questions thanks. It does help and I do learn a couple of things or 2.

Yes, Cedar Mill and Presler are stop gap processors, but its unfair to say that they are completely worthless. While Cedar Mill is unimpressive despite the power reduction, Presler does show promise. The entire Presler line will receive a speed boost from the doubled cache to relieve some pressure from the bandwidth starved FSB. This time Intel doubled the cache with no increase in latency so the lack of improvement from the 5xx to the 6xx series isn't repeated. Multipliers are now unlocked down to 12x, so all Preslers can save power by downclocking to 2.4GHz in addition to savings from the 65nm process. The 920D remains the cheapest dual core and the 840D which was criticize for underperforming despite being overpriced is now replaced by the 940D priced $100 lower.

As well, despite doubts Intel is releasing a 3.6GHz 960D Presler on April 23rd. It is timed to coincide with AMD's Socket AM2 launch. While the performance improvement of the 960D won't likely be breathtaking, especially compared to the new AM2s, what is significant is that the 960D will be priced at the level of the 950D causing a cascade in prices. The 920D probably being the first dual core for under $200 and other processors underpricing AMD. What's interesting is that AMD seems to be very confident in its AM2 platform and is going to relabel the X2 3800+ the X2 4200+, thereby assuming a 400MHz equivalent boost from DDR2. The new X2 4200+ will be priced as the X2 3800+ comparatively to the price cut 940D, and AMD appears to have no plans to release X2s at the 930D's or 920D's new price points.

http://www.theinquirer.net/?article=28721
http://www.theinquirer.net/?article=28367

As far as I know, Yonah has the same number of stages as Dothan which is estimated at 12. The decrease in performance is due to the latency of the L2 cache being increased from 10 cycles to 14 cycles due to additional algorithms for allocation. Latency in the cache and in general was also increased because the entire processor starts asleep and must wake up for processing. The performance hit is worth it as it allows average battery life to be close to that of Dothan despite having two cores. As well, I don't think the performance hit was actually that high in most cases since because of enhancements to the FPU, redesign to SSE2 execution and addition of SSE3. Even when comparing a Core Solo to an equivalent Dothan, the Solo has the advantage of a wider 667MHz FSB and RAM to help mitigate performance drops. Gaming performance is still lower than Dothan though due to the additional cache latency. Yonah's performance overtakes Dothan's when the second core is taken into account.

http://www.matbe.com/articles/lire [...] /page6.php

Conroe will have 2 more stages than both Yonah and Dothan but performance will increase despite this. The cache will be doubled, likely without increasing latency as the transistion from Banias to Dothan, to correct the reduction in performance from Yonah's added latency. The doubled cache will also bring cache allocation in line with Dothan by insuring both cores can have 2MB if needed. The cache allocation routines themselves will be updated and the sleep by default mode will likely be changed to sleep when not in use as its a desktop chip, both of which will control latency. While Yonah can keep up with the X2 on a 667MHz FSB, Conroe will be widened to 1066MHz and likely up to 1333MHz for the Extreme Edition. The nearly doubled bandwidth ensures Conroe can stretch its legs. The execution units will be redesigned and their number likely doubled which will be especially beneficial to FPU and vector performance. Micro-ops fusion was expanded to SSE and SSE2 instructions and will be expanded further likely including SSE3 among others, with Macro-ops fusion being added. As well, Yonah like Dothan was designed to maximize power usage so data paths that operated faster than rated clock speed were limited. Conroe being for desktop will have no such power limitation allowing its paths to speed to their full potential regardless of clock speed. Data prediction will also be expanded to the level of Prescott and beyond. All these features mitigate the pipeline increase of 2 stages and ensures that the 20% or more of clock speed increase potential goes toward increasing performance not making it up.