New Intel Chipset and Extreme Edition CPU Tested

Steve writes "Today sees the launch of both a new CPU and chipset from Intel. The CPU takes the form of a 3.46Ghz Pentium 4 Extreme Edition, running at 1066FSB, and the chipset is the i925XE, the first Intel chipset to support this new FSB. HEXUS.net have a review of both. It looks like AMD still have the lead when it comes to performance, despite Intel's attempts to counter the Athlon 64 FX-55." Hack Jandy links to more reviews at AnandTech, HardOCP, and ExtremeTech.

Comments

Nobody ever got fired buying

[fembots]

I think AnandTech summed it up nicely "So there you have it folks - the 1066MHz FSB does absolutely nothing for performance, [...], But with the move to the 1066MHz FSB we have a platform launch that, in the spirit of the 925X and 915 launches, does virtually nothing for performance."

However the real question is, how many decision-makers are reading these review/benchmarks, or do they just buy Intel because it's Intel, or that's what xx-business weekly says?

Re:Nobody ever got fired buying

[Wesley Felter]

The Extremely Expensive Edition CPUs aren't targeted at PHBs -- they're targeted at gamers. The PHBs are still buying 915 chipsets like Intel told them.

Re:Nobody ever got fired buying

[Silverlancer]

Interestingly enough many, including myself, were expecting quite a leap with 1066mhz FSB, especially considering the huge leap from 533 to 800. However, it seems as if the P4's bottleneck now isn't bandwidth at all, but latency, like the Athlons always have been. In other words, DDR2 (more bandwidth, crappy timings) is going to do shit all for the P4.

I liked the previous version ...

[Anonymous Coward]

where I clicked it and it said "move along, nothing to see here"

Seemed more accurate.

I wonder if it includes...

[Anonymous Coward]

Intel Extreme Graphics! W00t!

improvement

[Coneasfast]

anandtech shows a 1% increase in speed over 800mhz fsb in most cases, is this really something to get excited about? will this difference open up in the future?

If I wanted the fastest...

[Realistic_Dragon]

...I'd be twice the sucker I actually am.

Thank god for the ponces and their fast stuff obsession making things cheap for me :)

Re:But is it 64bit?

[gl4ss]

no.

and the speed improvement from 800mhz fsb? almost zero.

soo.. what's the point? exxxxtreme.

Price / performance

[FiReaNGeL]

Honestly, who is buying these things? At a price of 999$ US (1000 units lot) and a marginal performance increase compared to other, far less costly solutions (3500+ AMD anyone?), I just don't see a market. Is it just for the performance crown, which they didn't even get to win this time around (or should I say, in the past 2-3 years)?No word on heat, nor power consumption.

AMD all the way. Intel is alive just because of Dell (among others) and a large reserve of cash. They cost more, do less, and heat your bedroom to boot. But it has 'Intel Inside', so I guess it must count for something...

Another, older review of 925XE

[DrMrLordX]

Right here. Though, I must admit that I found some of the results to be a little wonky, along with the test bed. How'd they get a FX-51 running on a socket 939 board? Underlocked a FX-53?

Re:Another, older review of 925XE

[MightyPez]

Because Tom's hardware is notoriously biased towards Intel. Photochopped P4 cores, funky timings on AMD rigs, and of course; editorials like this from staff writers which say the following:

There is nothing finer than raising the hackles of delusional AMD lovers. However, today I do so with a heavy heart. This is no time to take aim at the pompous, self-righteous head-in-the-sand-ostriches of the alternative chip lifestyle. One must embrace them, hug them and wipe away their tears.

Re:Another, older review of 925XE

[DrMrLordX]

I've heard this complaint many times, but I'm beginning to think that Tom's Hardware may simply feature insane reviewers and article authors. The results in the benchmark suite to which I linked are a bit odd(in particular are the rather stunning Doom III results which are wildly different than the ones features not long ago in Anandtech's Doom III CPU shootout), but what is most noteworthy is that the author is not at all enthusiastic about the 3.46 ghz P4EE in his conclusion. If he truly was biased towards Intel, or if everyone at Tom's was thusly biased, you'd think he'd try to spin the benchmark results as a major victory for the EE.

Ouch

[heli0]

http://hardocp.com/article.html?art=NjgyLDY=

I think Intel knows that they are not going to get a lot of kind press today.

I think Intel has put this launch where they think it can do the least amount of damage by actually being noticed.

Intel's new Pentium 4 3.46 Extreme Edition processor touting its 1066MHz FSB and supporting 925XE chipset bring nothing new to the table in terms of real-world performance.

"Is this a paper launch?" Quite frankly, I don't know, and I don't see any real reason to care.

http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2261&p=17

So there you have it folks - the 1066MHz FSB does absolutely nothing for performance.

We can only wonder what Intel is thinking, releasing an entirely new chipset just four months after they released the original. Either the 1066MHz FSB is going to make its way to CPUs faster than we have anticipated, or Intel has just introduced the world's first useless FSB improvement for the next 9 months.

But given that Intel isn't planning on ramping clock speed up too high anytime soon, we'd say that the 1066MHz FSB is best left for late next year, when more useful implementations of it will appear.

Will they ever beat intel

[kff322]

no matter how fast the clock speed the pipline is sooooo long for P4 chips causing a sever preformance degradement. This is why G5 and AMD chips are faster at lower clock speeds because of there shorter pipline. Intels high clock speeds just look good

Re:Will they ever beat intel

[evn]

If only it were that simple! Cache sizes, prediction facilities, execution units, register count, etc. all play a significant part in CPU performance and to reduce this to an argument about who's pipeline is bigger ignores many of the important issues.

Pipeline length has some impact on performance and until recently Intel has been able to perform well by jacking up the clock speed. Sure it ate tons of power, and heated your room but it didn't really matter provided Intel's chips could perform as well as the AMD, IBM, Motorola, etc. competition. Think of a trip to the drag strip: if my 5.7L corvette runs the quarter mile in 12.5 seconds and your 1.6L civic does it in 13 seconds I still win the race. In a race to be the fastest you can't lean out the window and yell "You won, but I was almost as quick and I did it with 75% less motor!": you'll look like a fool. The performance crown is about being the fastest. period.

For the last 9 months or so Intels small-block Corvettes have not only been losing the races, they're getting beaten by Subarus that produce more power, get twice the gas mileage, and cost less.

You might want to read some of the ARS Technica articles that cover CPU design and illustrate some of the differences between the various architectures:

• Pentium 4 and the G4e
• The future of Prescot
• An Architectural overview of the worlds most famous desktop processor
• An introduction to pipelining

Catching up

[thedogcow]

Intel is slowing catching up. But the fact is that my DP 2.5GHz G5 is at 1.25GHz Frontside bus - per processor.

Hey, slow down!

[dsanfte]

Do you guys think we could, like, halt the march of progress for a year or two? 1066FSB already?

You know, some of us over here have Athlon XPs with 333FSBs, and we're crying our eyes out. Please, think of us. Sometime? Maybe?

Fuck.

Not really...

[cbreaker]

Athlons have always been fairly LOW latency chips, and the memory used (fast DDR memory) is low latency too.

The P4's on the other hand have used Rambus memory for awhile, although that's not really the case anymore. But when they did, they always excelled at memory THROUGHPUT because Rambus runs at high frequencies. Rambus memory however is fairly latent - it's the trade-off.

DDR2 RAM won't be "fast" until we see it in much higher speeds - DDR2-800 most likely. Of course, it will always have more latency then DDR because it uses four banks of DRAM instead of two.. I'm sure you can research all this via google if you're interested in learning more.

Think of the drivers

[poptones]

I use an AMD box now - I have for some time. It's not even a powerhouse, it's an old XP1800. I've used this CPU on three different motherboards now, and so far I'm still looking for a reason to consider AMD when I finally replace it. I've had an S3 motherboard, a via motherboard, and TWO Nvidia motherboards. The closest I came to having decent chipset support was with the S3 and that's only because the guy who wrote the 3D drivers was able to basically con S3 out of the information he needed in order to do it (ie if you get a mainstream distro his drivers won't be in it due to potential legal issues).

The first Nvidia I bought to try out, then decided I wanted that great whajamacallit sound support so I spent weeks looking for a miniATX motherboard that had this feature. When I finally got it I discovered it has TERRIBLE sound - I mean atrocious, like the crap you would expect from a five year old emachine. Overtones, quantization noise - just horrid. And this is using THEIR drivers, which I cannot use along with THEIR 3D supporting video drivers because of random lockups the two together cause on my mandrake system.

If I get an intel system I at least get decent drivers. So here we have an intel motherboard that offers basically the same performance as the top of the line AMD, meaning "it can be done" and a lesser system (as I would buy) will also be proportionately less expensive. So for a premium of just a few bucks I can get similar performance AND I get open drivers that will work with my linux system?

Where do I sign up?

Dear god...

[Tyler Eaves]

The FSB on that thing is clocked faster than my CPU....

What??!!?

[JamesTheBard]

C'mon. This has gotta be useful for all the bachelors out there. Nothing like surfing the net at home while warming the house at the same time. You know that Intel is just waiting for the right time to unleash the hot-plate add on so I don't even have to leave my computer to cook the vast stores of Top Ramen.

Re:64 vs. 32

[Sivar]

I noticed that the summary points out how the AMD64 is faster. Well, what were you expecting, its 64bit vs. 32bit.

If you want a fair comparison then compare the AMD 32bit processors against Intel 32bit processors.

And compare AMD64s against Intel's 64 bit chips.

Someone needs a gentle tap with a cluestick.

1) Being 64-bit does not necessarily improve performance and, in fact, can degrade performance when used on the VAST majority of applications that primarily use integer numbers of less than 4.3 billion (2^32 unsigned). Take a look at Solaris/SPARC64 for an example.

2) Even in applications that can make use of 64-bit integers, the AMD64 specification defines an "integer" as 32-bits. Software has to expressly use a "long" (or similar) to make use of the other half of the register size, and because on 95% of computers out there (read: vanilla x86 systems) a "long" is the same thing as an "int", this is done rarely at best.

3) Even if all software in the universe could get a staggering performance boost from 64-bit registers AND were instantly tuned to use them, it wouldn't matter because all of the software used to compare the Athlon64 to the Pentium IV is 32-bit software running on a 32-bit operating system, except in the occasional tests that are designed specifically to test the benefit of the Athlon64's 64-bit mode.

4) Even if every one of the professional review sites were manned by biased or clueless authors (generally true of Tom's Hardware and GamePC (and any review website run by your average l33t w4r3z d00d or non-technical game enthusiast), though the former appears to be improving), the 10% average gain when compiling software to use the 64-bit extentions of the Athlon64 is nowhere near the actual performance gain, in 32-bit software, that the Athlon64 has over the Pentium IV in most games and a number of other applications.

5) Even if the performance gain of 64-bit mode was greater by far than it is now, the bulk of the performance improvement in most software is from a: the integrated memory controller (which is also used in 32-bit mode), and b: the fact that the number of general-purpose registers has doubled from 8 to 16, greatly reducing the amount of register variable swapping needed. Again, most apps simply do not care if they can fit huge numbers in a register, because they do not need them.

So as you can see, your assertion is flawed.