Intel's Core 2 Extreme & Core 2 Duo: The Empire Strikes Back
by Anand Lal Shimpi on July 14, 2006 12:00 AM EST- Posted in
- CPUs
FSB Bottlenecks: Is 1333MHz Necessary?
Although all desktop Core 2 processors currently feature a 1066MHz FSB, Intel's first Woodcrest processors (the server version of Conroe) offer 1333MHz FSB support. Intel doesn't currently have a desktop chipset with support for the 1333MHz FSB, but the question we wanted answered was whether or not the faster FSB made a difference.
We took our unlocked Core 2 Extreme X6800 and ran it at 2.66GHz using two different settings: 266MHz x 10 and 333MHz x 8; the former corresponds to a 1066MHz FSB and is the same setting that the E6700 runs at, while the latter uses a 1333MHz FSB. The 1333MHz setting used a slightly faster memory bus (DDR2-811 vs. DDR2-800) but given that the processor is not memory bandwidth limited even at DDR2-667 the difference between memory speeds is negligible.
With Intel pulling in the embargo date of all Core 2 benchmarks we had to cut our investigation a bit short, so we're not able to bring you the full suite of benchmarks here to investigate the impact of FSB frequency. That being said, we chose those that would be most representative of the rest.
Why does this 1333MHz vs. 1066MHz debate even matter? For starters, Core 2 Extreme owners will have the option of choosing since they can always just drop their multiplier and run at a higher FSB without overclocking their CPUs (if they so desire). There's also rumor that Apple's first Core 2 based desktops may end up using Woodcrest and not Conroe, which would mean that the 1333MHz FSB would see the light of day on some desktops sooner rather than later.
The final reason this comparison matters is because in reality, Intel's Core architecture is more data hungry than any previous Intel desktop architecture and thus should, in theory, be dependent on a nice and fast FSB. At the same time, thanks to a well engineered shared L2 cache, FSB traffic has been reduced on Core 2 processors. So which wins the battle: the data hungry 4-issue core or the efficient shared L2 cache? Let's find out.
On average at 2.66GHz, the 1333MHz FSB increases performance by 2.4%, but some applications can see an even larger increase in performance. Under DivX, the performance boost was almost as high as going from a 2MB L2 to a 4MB L2. Also remember that as clock speed goes up, the dependence on a faster FSB will also go up.
Thanks to the shared L2 cache, core to core traffic is no longer benefitted by a faster FSB so the improvements we're seeing here are simply due to how data hungry the new architecture is. With its wider front end and more aggressive pre-fetchers, it's no surprise that the Core 2 processors benefit from the 1333MHz FSB. The benefit will increase even more as the first quad core desktop CPUs are introduced. The only question that remains is how long before we see CPUs and motherboards with official 1333MHz FSB support?
If Apple does indeed use a 1333MHz Woodcrest for its new line of Intel based Macs, running Windows it may be the first time that an Apple system will be faster out of the box than an equivalently configured, non-overclocked PC. There's an interesting marketing angle.
Facebook
Twitter
RSS
202 Comments
View All Comments
soydios - Friday, July 14, 2006 - link
Awesome article, as always from AT. Between the article and the 100+ comments above this one, almost all my Conroe questions have been answered.;)
johnsonx - Friday, July 14, 2006 - link
Wow is all I can say. I don't think there's ever been a performance jump quite like this. AMD will still do ok (not win mind you) on the low-end, where most people live, but the mid-range and high-end is so overwhelmingly in Intel's camp now there's just no comparison. AMD would have to go to 3.4Ghz+ to even be competitive (and still not win mind you).My office computers will continue to be AMD Sempron for the forseeable future (AM2 from now on of course), but next time someone wants a CAD box I don't see how I can quote anything but Core 2 Duo.
Also, loved the dinner table analogy on page 13.
crimson117 - Friday, July 14, 2006 - link
http://www.tgdaily.com/2006/07/13/dell_xps700_to_f...">http://www.tgdaily.com/2006/07/13/dell_xps700_to_f..."Grapevine (TX) - On the day Intel announces its next generation of Conroe desktop processors - which is expected within days - Dell Computer will upgrade its top-of-the-line XPS 700 desktop computer model to offer not only an overclocked Core 2 Extreme CPU, but also the option of two Nvidia GeForce 7900 GTX cards in SLI mode. These will apparently replace the Pentium Extreme processor and GeForce 7900 GS options currently available, and will be in addition to the Aegia PhysX accelerator already offered."
epsilonparadox - Friday, July 14, 2006 - link
Probably an Nf4 chipset since there are motherboards with conroe support using Nf4.Avalon - Friday, July 14, 2006 - link
This is one fine architecture. It wound up performing as well as I had suspected it would. Definitely wow.cgrecu77 - Friday, July 14, 2006 - link
I gambled by purchasing a A64 3700+ a few weeks ago when it dropped to under 150 ...If I were to upgrade to the cheapest solo I would probably have to pay 500 dollars or more (new cpu, mb and memory) and I will probably get ~ performance in games. I stopped overclokcing a long time ago (except for maybe 10%) because I am tired of my new system becoming unstable after 4 months (plus I don't really need it anymore since nothing I do is really cpu limited).
98% of people actually buy the cheapest CPUs in a range (A64 2800-3200, x2 3800, C6300). That's where the real competition will be and frankly I'm not impressed at all, A64 is an ~3years old processor and intel's newest product can only beat it by 20%? As far as I can remember the Athlon 3200+ was even worse in comparison to P4 3,2 Ghz and most people still bought the 2500+ Bartons ... So I wouldn't worry too much if I were an AMD stock holder, especially that it looks like Opterons still have no competition in the multi cpu servers (more than 2 ...).
After all, if Dell decided to use AMD NOW of all times it must mean something, they probably had access to these benchmarks a little earlier than Anand had ... :)
One thing I would have expected to see for such a new CPU that is supposed to carry the Intel flag for the next 2 years is 64 bit performance. This is probably the last year when 32bit OSes have the upper hand, starting with Vista most people will take advantage of their 64bit cpus. Once the majors concentrate on 64 bit drivers you can be sure that they will gradually take resource from the 32 bit development and the balance will shift very quickly ... If A64 has even a 10% advantage over the Cores combined with the .65 switch it will probably balance the equation to the point where core and a64 will have similar performance per clock ...
dargaard - Saturday, July 15, 2006 - link
http://babelfish.altavista.com/babelfish/trurl_pag...">Some Conroe problems (??) with 64 bitdev0lution - Friday, July 14, 2006 - link
First off, why mention Core Solo when it's totally unrelated to this article and any of the comments?
Second, socket 939's days are already numbered. You'd have to pay the same $500 to upgrade to the AM2 processors benched in the article, so mentioning this as a diss to Core 2 Duo is pretty weak.
So you'd have the option of spending $500 to upgrade to a E6600 setup that would beat out a $1200+ AM2 setup. Sounds like a no-brainer. And if you think AMD's going to slash prices below Intel's parts I don't think you can hold your breath that long.
And 98% of people don't even build their own computers, they buy tier-1... who aren't making nearly that percentage of their systems with the lowest end parts.
Looks like I'll have a 3200, 3700 and 4000 (939 parts) for sale in the near future...
MrKaz - Monday, July 17, 2006 - link
OK.But if you dont want to spent more than 120$ for the processor, the Conroe looks expensive.
aznskickass - Friday, July 14, 2006 - link
Dude, 20% is a HUGE margin when you are talking competitive benchmarks.When A64 was released back in 2003 it had about a 10% edge on the P4 and people were lauding it for it's leap in performance, and rightly so, considering the AXPs were getting beaten by 10%, so it was a 20% turnaround altogether.
This time around, it's even more impressive from Intel as they have turned a 10 - 20% performance deficit into a 20% advantage, you really can't expect much more than that, can you?
What did you expect Intel to do, double A64 performance? I'm sure if they ran their chips @ 4GHz/400FSB they might be able to get close to that, but what is the point when you have beaten your competition so convincingly already?