Intel Pentium 4 2.6c 800MHz FSB - Conclusion
(Page 4 of 4 )
(Benchmarks Continued...)
3dMark2001 (Build330)
While Futuremark in in the middle of quite a controversy over it's newer 3dmark03, 3dMark01 is still a standard for current reviews. It is very dependant on the CPU and bus speed, making it a prime candidate for today's review.

Oddly here the HT did not offer us even a marginal gain in 3dMark01 after we've seen other sites show the opposite. We actually reran the bench several times at 2.6Ghz in both modes to see if we were skewed but came up almost dead even every time. The other noteable here is the hefty increase of 600 points by the 2.6 overclocked with it's higher FSB. This is a synthetic bench and theoretically means absolutely nothing, but with 3dMark01 being the current "braggin' rights" application of choice, it's nice to see.
BENCHMARK COMMENTS
The benchmarks pretty much told a tale that surprised me a bit. I honestly expected the 2.6 CPU running at the higher FSB overclocked to beat the overclocked 3.0c CPU by a larger margin than it did. While there were increases generally, especially in memory benchmarks, they were pretty close. The good thing about that is the 2.6GHz was roughly half the price of the 3.0c, so how can I complain about it only being "slightly better in performance".
One thing I'd like to point out is that the "no-HT" numbers may be a bit misleading. HT was designed to shine in a multitasking environment. Running benchmarks is not the battleground it was meant to fight upon. Running a minimized Freelancer game while downloading from an FTP, while typing an email, and launching a word doc, THAT is the environment that you will truly notice HT shining in. I can tell you that at least to my perception, I can truly tell a difference between running it and not running it. My standard tour on the PC has me editing reviews, replying to emails, and having about 4-6 IE sessions launched. For me it is a benefit. was it worth the extra money of the 3.06 P4 when it was the only game in town sporting HT? I don't think so. Is it worth it now that you can get it in the below USD$200 range via the 2.4c? I'd have to say yes.
OVERCLOCKING
Well I guess I've already let the cat out of the bag a bit in the benchmarks huh? Yes, you've already seen what our max "stable" clock is with this 2.6c CPU, right around the 3514Mhz point. We've run it up to just over 3.6GHz (can you say a FULL 1GHz OVERCLOCK!!??) but it was not a rock at that speed so we backed it back down.

(Default speeds above)

(Default speeds above)
All results you see in this review were done with STOCK COOLING! Our reasoning for this is simply. Sure, I could strap a water cooler on the chip but in all reality that would not paint a clear picture of this chip the way the majority of you are going to run it. The point is, with this chip anyway, what you've seen here should be the worst case scenario. There is plenty of head room left in this CPU and I can tell you for a fact that heat was definitely our enemy when overclocking this chip. How much potential is left in it I can't tell you yet as it'll be at least next week before I get to put a water block on this chip and see what it's made of. The bottom line is that the results we've shown you are VERY attainable with nominal cooling, with even more potential with either a high end air cooler or water.
I have to admit, the first time I ran at 1100Mhz FSB I had to sit there a full 10minutes and just enjoy the view. Sure it crashed about 20min later, but it was fun while it lasted. Overclocking these C revision P4's is the most fun OCing I've had since I messed with my Celeron 366 and my Pentium3 500e. Ah, those were the days. The potential of these chips has me that excited all over again, and that can be seen as nothing but a good thing.
The only real challenging part of overclocking this 2.6 was locating the "sweet spot". With the bus as high as it is you are forced to run the memory at a divider. That is fine as the additional bandwidth made available by the CPU's excessive bus is made apparent by even the memory scores run at a 4/5 DDR divider. I could run the memory 1/1 at semi-relaxed settings all the way up to 230Mhz, but the performance there was no where near what I gained by dropping to 4/5 and cranking it up. Let's suffice to say that there is plenty of headroom in these C revision P4's and it is a pretty safe bet that if you have the memory and the motherboard to push them, they are the current CPU to own for overclockers/enthusiasts.
FINAL THOUGHTS/CONCLUSION
It's been over two years since Intel's last "must have" CPU, the Pentium4 Northwood 1.6a, I think it's safe to say that Intel has found it's next "must have". The only question remaining is whether the chip to have is the 2.4c or the 2.6c. With the high bus speeds of these chips it leaves us in somewhat of a bind when it comes to memory simply not being able to hang with the higher speeds. Fortunately, as we've seen in this review, memory bandwidth does not appear to be throttled back too greatly by running the 2/3 and 4/5 memory dividers. Even when running a 4/5 divider at 272Mhz front side bus we were able to achieve memory bandwidth in excess of 5500MB/s, and that my friends, is impressive. So do you go with a 2.4GHz C revision and aim for 300Mhz FSB, or nestle down with a 2.6Ghz C revision and settle for a measly (note the sarcasm) 270+Mhz FSB? That folks, is an outstanding problem to have.
From the limited experience we've had with the C revision Pentium4 in addition to what we've been seeing around the net, 3.3GHz appears to be close to a lock, 3.5Ghz+ a very real possibility, and insane lottery winner type results of the 4GHz range on the outside. These speeds seem to be achieved using any of the four different C revision P4's, the only difference it appears is the bus speed at which you're running those speeds at. Placing heavy emphasis on the "more is better" ideology, I'd have to say the 2.4GHz chip is the better bet based on the fact that it will be at the highest front side bus with it's lower 12x multiplier. Another added bonus is that it is running approximately USD$45 or so cheaper than the 2.6GHz part. Either way, you "should" be getting a stellar performer.
Like every other overclocked product there are no guarantees. Intel ships a chip at a rated speed and that is all it's guaranteed to run, so there is no legitimate complaint if your newly acquired chip doesn't hit the speeds discussed in this review or higher, but it's a safe bet that these cores will far exceed their rated speeds.
Having been on the AMD train for the past year it was an easier transition to make than one would think. Judging by the discussions in the forums there are more and more "AMD whores" (their word, not mine!) making the switch as well. Is a shiny new C revision P4 right for you? Only you know the answer to that one.
Personally, I'm happy with the 2.6GHz part we reviewed today. I've been running this CPU in my main rig for everyday use at 3.25GHz (13x250) using default voltage (1.525v) and a 4/5 memory divider (memory sits at 200Mhz) without a single hitch. With a 2.4GHz you will likely gain slightly better results when going for the gusto in benchmarking, but for everyday use I prefer the slightly higher multiplier allowing me to remain at a sane FSB. (I can NOT believe I just referred to 250MHz as a sane FSB!). Yes, these chips are all of that and then some by all experienced accounts. My Barton 2500+ is now nestled soundly into my backup box and has been replaced for the foreseeable future by the 2.6GHz Pentium4c.

HIGHS
Insane front side bus thru overclocking
Hyper-Threading enabled on all C Revision P4's
Outperforms the 3.0c which is twice as much $$$
Intel's heat spreader rocks compared to AMD's bare core.
Arguably, the best CPU available for your money
LOWS:
The 2.4GHz MAY be a better choice, depends on user preference.
Now let me comment quickly on the "Lows." I've never before written a review on a product with no lows. The bottom line on this P4 is that it's priced competitively compared to the 2.4c and 3.0c, it performs better than the 3.0c by utilizing it's higher front side bus potential, it offers a heat spreader over AMD's bare core, it's overclocking potential is downright insane. We're talking about pretty much a full GHz overclock! I'd like to have been able to find more fault in this product to show we're unbiased, but plain and simple I'd have to create one. The only thing to consider in my opinion is whether you want the 2.4c or 2.6c, you honestly, judging by initial performance and yields, can't go wrong with either. For these reasons, we believe the Intel P4 2.6c easily deserves the Dev Hardware Recommended Award.

That's it, that wraps it up from here. If you're looking to buy a C revision P4 you can check out stellar prices of the 3.0c, 2.8c (which I would not recommend), the 2.6c and the 2.4c by clicking on whichever you'd like. Feel free to head into our forums to discuss this review or back to the Dev Hardware front page for more GooSH!™.
| DISCLAIMER: The content provided in this article is not warranted or guaranteed by Developer Shed, Inc. The content provided is intended for entertainment and/or educational purposes in order to introduce to the reader key ideas, concepts, and/or product reviews. As such it is incumbent upon the reader to employ real-world tactics for security and implementation of best practices. We are not liable for any negative consequences that may result from implementing any information covered in our articles or tutorials. If this is a hardware review, it is not recommended to open and/or modify your hardware. |