Pentium 4 3.46 Extreme Edition and 925XE: 1066MHz FSB Support is Here
by Anand Lal Shimpi on October 31, 2004 3:00 PM EST- Posted in
- CPUs
Does the 1066MHz FSB Improve Memory Performance?
Quite possibly the biggest feature of the 1066MHz FSB today is the fact that it runs at a clock multiple of DDR2-533's frequency. Why is that such a big feature? It's analogy time:
If two people are having a conversation and they can both talk and listen at the same rate, then the conversation will flow as smoothly as possible. If person A talks and listens slower than person B, then person B will always be waiting for person A instead of communicating as fast as possible - a frustrating situation for those that have been here before. Running a FSB and memory bus asynchronously is just as frustrating to the CPU; if the two frequencies aren't synchronous then there is an additional latency penalty incurred while transferring data between the two buses and because of that additional latency penalty, there is a reduction in usable bandwidth.
With the original 925X chipset we were a bit unhappy to see that the Pentium 4's 800MHz FSB was paired with DDR2-533, creating one of those frustrating asynchronous situations. But with a 1066MHz FSB (266MHz x 4), the 925XE can communicate synchronously with DDR2-533 (266MHz x 2), thus reducing memory latency and increasing memory bandwidth in theory. What do we see in practice? To answer this question we look to two trusted measures of memory bandwidth and latency: CacheMem and ScienceMark 2.0.
First looking at latency we see that with the 1066MHz FSB, memory latency with DDR2-533 looks like it's hardly improved. Running the FSB at 1066MHz manages to shave off a just a few clock cycles.
| DDR2-533 3-3-3-12 Latency Comparison - Cachemem | ||||
1066MHz
FSB
|
800MHz
FSB
|
Performance
Improvement
|
||
| 512-byte stride - 32MB block | 226
cycles
|
227
cycles
|
0.4%
|
|
| 1k stride - 32MB block | 239
cycles
|
241
cycles
|
0.8%
|
|
| 2k stride - 32MB block | 266
cycles
|
266
cycles
|
0%
|
|
| 4k stride - 32MB block | 311
cycles
|
311
cycles
|
0%
|
|
We look at ScienceMark and see the same basic situation but with slightly improved performance; looking at the absolute latency values in nanoseconds we see that the 1066MHz FSB manages to reduce memory latency by around 2 - 6%.
| DDR2-533 3-3-3-12 Latency Comparison - ScienceMark 2.0 | ||||
1066MHz
FSB
|
800MHz
FSB
|
Performance
Improvement
|
||
| 16-byte stride | 4.69 ns
|
5 ns
|
6.2%
|
|
| 64-byte stride | 17.5 ns
|
18.12 ns
|
3.4%
|
|
| 256-byte stride | 73.43 ns
|
75.93 ns
|
3.3%
|
|
| 512-byte stride | 75.93
ns
|
77.81
ns
|
2.4%
|
|
The reduction in latency isn't insignificant under ScienceMark, but what about its affects on memory bandwidth?
Looking at Cachemem once again we see an increase in memory bandwidth of just under 4%.
| DDR2-533 3-3-3-12 Bandwidth Comparison | ||||
1066MHz
FSB
|
800MHz
FSB
|
Performance
Improvement
|
||
| ScienceMark 2.0 | 4742.02 MB/s
|
4347.63 MB/s
|
9.1%
|
|
| CacheMem | 3455.3 MB/s
|
3324.7 MB/s
|
3.9%
|
|
ScienceMark appears more considerate of Intel's hard work and shows a 9% increase in memory bandwidth. The thing to keep in mind here is that the improvement in memory bandwidth will depend on the types of applications run, but the ScienceMark and Cachemem results should give you an indication of the range of improvements to be expected in applications that are memory bandwidth limited. In those applications that aren't currently bound by memory bandwidth, the impact will be much less.
What is important to keep in mind here is that DDR2 is still running at relatively high latencies. Even while running synchronously with the FSB, the 925XE and DDR2 combination still posts higher memory access latencies than 875/DDR400 platforms thanks to DDR2's high memory timings. We were able to run 3-3-3-12 timings on our DDR2 test platform by keeping memory voltage at a safe, but overclocked, 2.0V, but anything faster than that was too much for today's DDR2-533. It will take even lower latency DDR2 in order for even the 925XE platform to show some further performance advantages.
Given the relatively small increases in memory bandwidth and decreases in latency, the extra 66MHz of the 3.46EE will have to go a long way in order to gain any more ground for Intel. Let's see how things shape up in some real world tests.
Facebook
Twitter
RSS
63 Comments
View All Comments
T8000 - Thursday, November 4, 2004 - link
The most important part of this release is the Intel 925XE chipset, that will allow much higher overclocks because of its 1066 bus support.This is because the 925XE will have the right divider to reach 1066 without any PCI-E overclock.
So with a 925XE mainboard, you can run an Intel 530 CPU at 4Ghz with any PCI-E GPU you choose, because only the CPU will be overclocked and Prescott has excellent chances of reaching 4Ghz with modest water cooling or good air cooling.
Odeen - Wednesday, November 3, 2004 - link
Realtek codec on an Intel board... and here I thought Intel actually made quality motherboards, which entails Sigmatel or Soundmax onboard audio chips.Sigh :(
johnsonx - Wednesday, November 3, 2004 - link
Slim: You're right... my bad. I didn't read every single page. I read the couple of introductory pages, then skipped to the test configuration page, perused a few benches, and then skipped to the conclusion.The measured results of course are no different than I thought they would be...
bob661 - Tuesday, November 2, 2004 - link
We need to have our own review website called www.dontreleasesh!tunlessitsactuallyabetterproductthan theonebeforeit.com.SLIM - Tuesday, November 2, 2004 - link
johnsonx,Anand did isolate the fsb as the sole variable when he DOWNclocked both chips to 3.2ghz (266 x 12 and 200 x 16) on page 3. There was a slight caveat that faster chips would benefit more from a fsb boost. And yes the faster bus increased performance by almost 1% in some tests woohoo!!!
SLIM
johnsonx - Tuesday, November 2, 2004 - link
One thing that might've been interesting to see:Overclock the 3.4EE to 3.46Ghz by OC'ing the FSB to 203Mhz or 204Mhz (812 & 816 respectively). This would completely isolate the effect if the increased clock speed of the 3.46EE, showing only the increased FSB performance... at that point I suspect that the tiny performance gains would completely evaporate.
Mind you, I'm not suggesting this would change the conclusion much, but it would put a big exclamation point to it...
BTW, one does have to wonder why Intel bothered with this. If the 3.46EE/925XE combo is no faster than the 3.4EE/925X combo (I'm assuming the 925X=925XE @800FSB), then why go through all the trouble? Indeed, isn't it true that an 'old' 3.4EE/875 combo is faster still?
Good grief, at least when AMD releases a new top-end chip it is actually measurably faster. Regardless of whether the rating is 'earned' or not, no one can argue that the 4000 isn't (generally) faster than the 3800, nor that the FX-55 isn't faster than any other A64.
Tides - Tuesday, November 2, 2004 - link
Some benchmarks? Hardly. AMD owns in actual games, workstation apps, and half of the other stuff. Not to mention AMD doesn't make you upgrade to ddr2, and AMD cpus are 64bit. Intel's new chips have low shelf lives while the current AMD 64's you buy will last you a lot longer.Performance, realiability, and long lasting.
danidentity - Tuesday, November 2, 2004 - link
IntelUser2000, you couldn't possibly be any more wrong. I will be the first to admit that AMD chips excel above Intel chips in many benchmarks.However:
1. Intel is no where near dead. Calling them so is ridiculous. In Q3 of this year Intel posted revenue of 8.5 billion compared to AMD's 1.2 billion, or SEVEN times as much.
2. AMD is NOT closing "very rapidly" in marketshare. It would appear that way from reading sites and forums like these, but it gives you a false impression. Keep in mind that the largest supplier of PCs on the planet puts Intel chips in every machine. AMD's mobile chips can't compete with the Pentium M in terms of performance and functionality.
3. Intel is not stupid, they have some of the best engineers on the planet. If they seriously thought that AMD was going to topple them as the market leader, or even if they are predicting it, you can GUARANTEE they have something in the works to strike back. They have the means and the money.
4. While many people don't know exactly what clockspeed is, everyone thinks it is the ultimate measure of performance. That mindset will take a LONG time to change, and by then, Intel will have something new.
Most people out there don't even know AMD exists. Just because AMD chips beat Intel chips in some benchmarks posted on technical computer sites, don't mean they're going to topple Intel.
JonahStone - Tuesday, November 2, 2004 - link
Performance is not the only reason why somebody buys a CPU. Although 64 bit might not be available now, does not make it unimportant. Many who buy a computer will keep it for a long time. I do not want to buy a new PC in a year's time to run 64 bit apps. All reviews keep on comparing 32 bit performance and do not even mention the advantage 64 bit will bring. It does matter!!!!!!!IntelUser2000 - Monday, November 1, 2004 - link
Intel is not doing bad. They are doing terrible. So terrible that you might as well call them dead. Probably will last till 2009 before they fill bankruptcy.To those people who say people in forums don't know anything and that there are other people stupid enough to buy Intel chips(I mean all Intel chips): Uhh, yeah, get your head straight, since AMD is closing with Intel very rapidly in marketshare, in server, desktop, and laptop, and that means that gamers actually do make a difference(albeit slowly) making other people buy computers. You think other people will buy P4's because of high clock speed? That's BS, since people who is stupid enough to buy Intel chips don't even know what clock speeds does. There are only a very few that knows computers JUST enough to say clock speed is good.