Intel's 2.4GHz Pentium 4 Unleashed

EconolineCrush writes: "Intel has released a 2.4GHz version of its Pentium 4 processor, and The Tech Report does an excellent job comparing its performance with previous Pentium 4 processors, and the latest in AMD's Athlon XP stable. There's more to this story than just another notch on the MHz pole, as the review showcases some new benchmarks in an already diverse set of tests, and shows the new P4 leveraging an impressive performance from RDRAM-based platform. Incidentally, the slack demand for RDRAM has it almost as cheap as DDR SDRAM."

Comments

Another Article

[WndrBr3d]

Tomshardware has also posted an article today putting it against the latest Athalon XP.

Some Linux Benches

[TheMatt]

There are some benches on *NIX flavors here: link.

They aren't the most recent, but they effectively show that for us theoretical chemists, nothing beats P4+RDRAM+ifc for Gaussian98 (the timings are in minutes, not the sad seconds on most sites). Of course, more processors help, but the benchmarks looked at single chip+motherboard.

Re:802.11b is on 2.4Ghz.....

[taniwha]

that was back in the days when the signals carrying those frequencies moved on wires between chips - they could radiate - trust me (as a chip designer) no one is even trying to run signals at 2.4GHz between chips in your PCs. Of course a little bit always gets out but at those frequencies lead inductance is going to kill much of it anyway.

The FCC is very carefull about making sure people's hardware doesn't radiate and interfere with various radio services - that's why you have metal cases on boxes rather than cheapo plastic ones -

Time for .13 Athlon

[jmv]

This shows that it's really time for AMD to release Athlon XP's at .13 um before Intel are too much ahead of them. From what I understood, .18 Athlon are stuck at PR 2100+.

More 2.4GHz Pentium 4 Benches here & @ 3GHz!

[FlippedBit]

HotHardware's Benchmarks and 3GHz Overclock!

Re:If a tree falls in the woods and no one's aroun

[Anonymous Coward]

Very incorrect. I do hardware validation for a living. A large percentage of ASIC (application-specific integrated circuit) silicon is designed in a language called Verilog that we use for pre-silicon testing/modelling. I have simulations that take 3 days to run on a P4 2.0ghz. If I can shave that down to 2.5 days by upgrading to a P4 2.4ghz that would be great!

Maybe YOU don't have a need for faster computers, but that doesn't mean that no need exists.

Of course, if I had access to dual 1ghz ultra sparcs, the simulations would go much faster than a dual P4 box. But for the price of those Sun machines I can build 5 intel boxes and run more jobs in parallel. :)

Breakfast

[srichman]

with a big enough heatsink on an Athlon, you can probably already do that.

You think so?

Re:AMD chipsets are fine

[T5]

Unless you're talking about the USB support, which is broken in the latest AMD 760MPX chipset. Most vendors are shipping a USB PCI card to make up for it, but for some that loss of a PCI slot is very painful.

Re:pushing MHz

[Zeinfeld]

But is that a 24 valve V8 or a 24 valve V6, if it's a V6 then the 16 valve V8 (especially if it's a big block =) will kick it's ass!

Not necessarily. The V6 GTI I bought for the wife creates more horsepower than the majority of US made SUVs which are typically based on engines that were originally designed in the 60s. Equally the V8 in my XK8 will easily outperform the V12 engine Jaguar used to use [and still do 20 Mpg arround town rather than 10]

What really matters though is the chasis the engine goes in. For example the GTI will nail any SUV in the street, even if you dropped the Jaguar engine into it. Heck you could drop the engine out of a Ferrari F40 into a Ford Exploder and the Jag would beat it round any track. To go fast arround a circuit you brakes matter as much as your engine.

Its pretty much the same when you get to MHz. A 2.4MHz processor will probably go faster than a 2.0MHz processor all things being equal. However how much faster is pretty variable and all things are usually far from equal.

Unless you have the motherboard and O/S design that will support the beast you will probably notice about as much improvement from a 2.4MHz processor as painting a go faster stripe on the box.

Unfortunately most of the O/S in common use tend to spend a lot of time in unnecessary wait states. They ask a piece of hardware to do something, guess how long it will take and poll for the result. This isn't the way it should be but it only takes one baddly written driver to stonk the whole machine.

Of course back in the days of real operating systems there were these asynchronus service traps...

The bottleneck in UNIX and Windows is the GUI interface in both cases. The Windows GUI has lots of unnecessary blocking states. X-Windows falls foul of the lousy performance of interprocess communications on most modern UNIX boxes.

What about pricee/performance?

[Crag]

This may be redundant since I browse at 4, but I saw no mention in the entire article of the prices of the CPUs and their support hardware.

Pricewatch doesn't list 2.4Ghz P4s yet, but a P4 2.2 mb/cpu combo is $570, and the Athlon 2100 combo is under $300. The fastest Intel mb/cpu combo under $300 listed is 1.9Ghz, which can NOT keep up with an Athlon 2100 setup.

There's certainly more to a purchasing decision than price and performance, and I don't expect every article to cover every angle, but the disparity in price/performance ratios between the companies seems VERY signifiant to me.

Perhaps this article is too targeted for gamers. Business and home users will be more concerned with economy, and professional high-performance users (server/workstation/research) will probably spring for dual processors if raw throughput is so important.

In any case, I look forward to AMD's next moves.

Rambus for a reason

[jbischof]

We are all finally learning why Intel chose Rambus, they just maybe should have supported DDR first and weened us off of DDR, and not the other way around. I remember when Rambus first came out and everyone was preaching DDR and saying that RAMBUS completely sucked. It is frustrating now to see the big effect it has at the higher Ghz and watch Intel abandon it because of marketing and ignorance of the general populus.

Well lets add another technology to the long list of products that were better than many commonly used products, yet never got significant market share. (BeOs, Alpha Processors, etc. etc.)

$600+ on Pricewatch...

[HiyaPower]

For that kind of dough, I can roll a dual 2100+ system and run rings around it in most real life tasks that would require this sort of speed processor (like video encoding).

For the moment, Intel may even have the highest preformance, lower priced processor (so as to exclude the Alphas, Itanics, etc.), but on a total price performance basis, the AMD chips beat them hands down.

Re:802.11b is on 2.4Ghz.....

[Skuld-Chan]

The question is bandwidth - how much bandwidth of electronic interference does the chip radiate? Probably not a lot.

Read up on the effects of narrow band transmitters on spread spectrum recievers and visa versa. Typically the frequency hopping mechanism can avoid interference with narrow band trasmitters and narrow band transmitters typically recieve low background noise when adjecent to spread spectrum transcievers. In summary the two devices can co-exist on the same frequencies and pretty much not interfere with the two.

This is probably why the USAF claims their awacs network is unjamable...

Re:pushing MHz

[Anonymous Coward]

1-nader is a fuckhead
2-britney has a nice rack
3-orange halls taste yummy
4-same reson devolution belongs there
5-prozac is a pill, no second hand nut-casing

Re: PCI 802.11b cards

[Raetsel]

First, I never said the pins were carrying 2.4 GHz signals. I said they were "waveguide-like". They will likely facilitate the radiation of some of the ~75 watts dissipated inside the chip package. Simple physics: energy goes from source to sink -- there is less similar radiation outside the package, thus there will be leakage. Fact of life. Need to reduce / prevent interference? That's what the grounded metal case is for.

Second, at 2.4 GHz a signal doesn't follow a wire (or a circuit board trace) like it does at 60 Hz. At 2.4 GHz a wire is more of a 'suggestion' than a 'command'. This is why (radar | microwave ovens | certain satellite communication systems) use waveguides instead of wires. It's also one of the reasons everything isn't running at the same clock speed.

Third, one of the Ten Commandments of /. -- Thou shalt query Google.

• Arcowave AWL-1100P
• D-Link DWL-520
• Intel PRO/Wireless 2011B LAN PCI Adapter
• LG Wireless PCI Card
• Linksys WMP11
• Proxim Harmony 802.11b PCI Card
• Samsung SWL-2000P or SWL-2100P

None of these are PCMCIA > PCI adapters, though some of them look like they're using the same innards. I'm not even going to include all the 'Mini-PCI' cards being used in laptops these days. Yes, they all have some shielding. No, it's not as complete as a PCMCIA card -- if I even dare call that complete.

PCI Cards are installed with the PCB facing in the general direction of the processor (in the ATX spec). I don't know the shielding capabilities of circuit board material, but it sure isn't a solid conductor -- and... many of your traces are exposed to the radiation inside the case. This is where I expect problems and performance degradation to have their roots.

Perhaps you remember a few years ago when it was trendy to install shielding around your audio card for a greater Signal/Noise Ratio? I saw people use copper flashing (the stuff you use to keep your roof from leaking) to construct a box, doing a very nice soldering job, use stand-offs for installation... all to remove a little static. The whole trick was to construct a Faraday cage that would allow the ISA connector (remember those?) as little clearance as possible, without actually shorting it.

We may see a resurgence of that technique.