Slashdot Mirror
Intel Pushes Pentium 4 Past 3 GHz
denisbergeron writes "Yahoo has the news about the new P4 who will run at nothing less than 3.06 GHz. But the great avance will be the hyperthreading technology (already present in Xeon) that allows multiple software threads to run more efficiently on a single processor."
What's nice about that is that the new 1.7 P4's will easily overclock to 2.6Ghz+, so you are getting almost a Gig free for just knowing which switches to push.
Or you can build a dual processor Athlon system for less money. No need for HypedThreading.
It has been reported on various sites that Athlon XP 2400+ chips (2GHz, new Thoroughbred Revision B core) are trivial to mod for dual CPU operation and easily overclock to 2.25GHz (150MHz FSB, aka 300MHz DDR, which is the most my ASUS A7M266-D will allow) with proper cooling (Thermalright SLK800 being my favorite). The chips are under $200 apiece. Imagine a Beowulf cluster of those...
Proper Athlon MP 2400+'s are due shortly I'd assume.
Hyperthreading works well for certain types of software, and awful for others.
Here's an article from Ars Technica on HT/SMT.
A couple of sites have benchmarked Xeons with HT enabled already (Anandtech and Aces Hardware spring to mind.) It provides a boost in some applications but can actually decrease performance in others. It's rumored that Intel has improved their implementation of hyperthreading but I wouldn't expect the 20-25% performance gains in most applications.
-- Jim
Granted, it will be good for people who are still running those 200MHz clunkers but what's the incentive if you're already running in the GHz range?
Unfortunately where I work the secretaries for division heads get these 3GHz machines and run Word on them while the scientists and technicians get to keep working on their Pentium 200MHz system. Maybe if they're lucky they get a hand-me-down from a secretary like a nice PIII-1GHz box.
I was torn between building another dual-CPU box (currently on twin 533Mhz Celerons with an ABit BP6 board), or going the small form-factor route. Now I can do both.
More at Shuttle's site.
Cheers,
Ian
Processors gain self consciousness at about 3 MHz, therefore we shall refer to them as "who", not "which".
The key to fast 3d is not how fast your procesor is, but how little it has to do. the better[1] your gfx card the less there is for the processor to do and the faster it will run. A better gfx card will almost always be a bigger win than a faster cpu for these types of games.
[1] better meaning one which implements in hardware more of the functionality used by the game.
It changed the CPU clock frequency - generally you left the switch on since that was the faster mode of operation. If you had an application/game/whatever that didn't handle the higher clock rate nicely (and a lot didn't) then you clicked the switch and the CPU core dropped from 10 MHz down to around 5.
They eventually became disused because instead of dropping down to 4.77 MHz (the orginal XT speed) they'd just drop some fraction of the regular CPU speed - down to maybe 7 or 8 MHz, which was way too fast still. Plus applications stopped doing stupid things like presuming the CPU frequency and using it for timing loops.
Fooey. It has to do with *everything*. Yeah, your graphics card is sooper-fast. But what has to feed that card data? Your CPU. Your memory bus. Your AGP bus. UT2003 happens to be CPU-limited even with the latest-and-greatest video cards, all the way up to the fastest Athlon chips available.
Take a look at this UT2003 benchmark chart:
http://www.anandtech.com/showdoc.html?i=1650&p =3
You can see that the GeForce 4 Ti cards are ALL still getting faster the faster the CPU gets, right up to the bitter end.
That's not to say that a couple of years from now that 3D cards won't handle physics and AI onboard-- but they don't exist now, so it's hardly fair to say "A better gfx card will almost always be a bigger win than a faster CPU."
It depends on the game, and the newer they are, the more CPU they'll eat. (See Battlefield 1942)
yeah, buying mail order's a real chore. I mean, apart from saving a ton of cash, you get a bigger choice too! What a drag.
That was classic intercourse!
Spreadsheets were the killer app that caused the PC to take off, and Lotus 123 came with a super-annoying floppy-based copy protection scheme. They intentionally misformatted the floppy, then the program verified that it was an original by doing low-level tricks with the floppy controller.
The most ridiculous and shortsighted part was that they used CPU-based timing loops to do the timing for their stupid floppy tricks. Of course, these were calibrated to the only CPU speed available at the time, 4.77MHz. As a consequence, if a PC was going to run Lotus 123, it needed to be able to slow down to the original 4.77MHz speed while it read the Lotus floppy. IIRC, Compaq had a nifty patent that automatically slowed the PC whenever the floppy controller was in use. Others had to make do with a manual switch.
The cost to society for this DRM fiasco, hundreds of millions of useless bezel switches, undoubtedly was far greater than any revenue that Lotus made by thwarting piracy. (In fact, their revenue from DRM might be negative, because they were eventually displaced by non copy-protected comptetiters.)
Sorry, but your post reeks of "armchair CPU designer" : It's all so clear and so obvious. I mean, it's not like Intel and AMD have a lot of extremely clever people who seek the best balance between all of the systems...is it?
Yes, they are slowly improving, but modern PCs are still behind where workstations were years ago, and a modern Intel based server is well behind a SPARC based machine.
Intel and AMD will spend their money on whatever generates the most ROI. They have collectively spent literally billions of dollars convincing Joe Public that CPU Mhz is the best way to measure the speed of a system - they aren't going to throw that away. A competent manager with R&D dollars to spend will therefore spend them on increasing Mhz.
Oh, and your post reeks of being underexposed to any architecture other than x86.
though the cost/benefit is out of whack. A P2 2.4Ghz with 2MB of L2 would get trounced by a 2.6Mhz with 512MB of L2 cache, disputing your claims that CPU speed doesn't matter. Large cache chips only make sense if you can't get a faster CPU:
Yes, assuming the code to run is 512k in size. If the code is ~2M, so it fits into L2 on the slower processor, then it will have the advantage, because the faster one will have to waste cycles moving the cache back and forth to main memory. Cache size is related to CPU speed only in terms of memory bandwidth: if your CPU cannot get data from main memory fast enough to keep it occupied, then you need faster memory closer to the CPU, which is what a cache is. If you are context switching, then you will have to keep dumping the cache and reloading it, which puts larger caches at a disadvantage.
Ultimately, caches are a hack; an elastoplast solution to the fundamental problem, which is the mismatch between the rate at which a modern CPU can process data, and the rate at which memory can supply it. In an ideal system, there would be no CPU caches at all, because the CPU could get data from main memory fast enough to keep it fully occupied. Systems used to be built like this, before the current obsession with clock speeds.
Unless your system is locked down, click on "Change my preferences" in the Search Pane and choose "Without an animated character." I did that so long ago that I hardly remember that the dog was there in the first place.
You can only drink 30 or 40 glasses of beer a day, no matter how rich you are.
-- Colonel Adolphus Busch
I went ahead and got you a link to another page in the SAME ARTICLE I linked that shows a chart of CPU usage with Radeon cards:
http://www.anandtech.com/showdoc.html?i=1650&p =6
Note that other pages in the article include the Kyro II, Matrox Parhelia, and the older GeForce 2 and 3 lines, as well as the GeForce 4. Keep in mind that the faster the card gets, the faster the CPU must be to keep it fed with data. You may not see CPU saturation with a slow card, because the card is maxed long before 100% CPU usage. In the Radeon chart, you can see that the faster the Radeon, the more CPU constrained it is. Just like with Nvidia.
The Radeon 9700 isn't there because it didn't exist when the article is written. It will be even more CPU constrained than the GF4.
I suspect you are a troll, but I'd hate to see the issue confused any further.