Slashdot Mirror
Glimpses of the Future from the Intel Developer Forum
km790816 writes: "Lots of cool stuff on CNet about the Intel Developer Forum, including the next version of the P4, followed by 3GIO on the desktop (1st version 0.5 gigabytes per second of data-transfer capacity but bumps up to 1, 2, 4 and 8 gigabytes per second with the use of more wires.), and Intel's work to embed wireless in silicon. Quote from the CTO: 'We could get it to the point where radios are built into every product we make, giving every device seamless, roaming connectivity. You will see orders of magnitude of cost (decreases) through integration into silicon.'"
The big question, though:
"DDR or RAMBUS or ????"
Seriously... with Intel recently dumping RAMBUS (see slashdot article a few days ago) what would these new P4s use? I would imagine that this "hyper-threading" along with higher-than-ever clock speeds would make the memory bandwidth even more of an issue than it already is.
Will they go back to using RAMBUS?
Will they continue to choke the P4 with DDR?
Or...do they have something totally different in mind without telling us?
-kwishot
Maybe it's the new XM radio which you can have Snopp Dogg falling to your office. That must be neat.
geek page at KY speaks
It would be really interesting to see how AMD is able to compete with Intel with the 'departure' from the common standard. Of course the new AMD chip ClawHammer, will be able to support current x86 instructions, but it won't really help you much to run it that way.
;-O
Hopefully AMD will really work on a compiler that take advantage of the new chip's strength. Maybe hire those SGI engineers?
I just hope we will still have a choice in the processor market a few years from now.
geek page at KY speaks
I'm wondering what radio spectrum they are going to use to accomplish this. Its a finite resource. If 3G ever gets going every bit of spectum left except the ISM bands is going to be used up. And I have a hard time imagining that products are going to work in a city where there are 10000 other people sharing the band at the same time. Marketing clashes with reality on this one.
One thing I am waiting for with wireless is some decent security functionality. What with the farce that is 802.11 and the proliferation of 'secure' data within companies I work with, wireless has become one of the major security threats.
The number of people I have found using RF keyboards/mice on computers in 'secure' areas, and not even believing that these can be snooped (which is quite trivial), then insisting that we have a 802.11 hub for their flashy new laptop, simply because it has that functionality built in.
I would love to see a standard developed for a plugable security model on top of these transports, so a 'suitable' level of protection can be installed for the situation.
- It's an ultra-low-power x86 chip line.
- It's due in 2003.
- It's not based on the P-4 core, but is a fresh design, possibly related to the P6 core. (Boy, that's a bad pair of abbreviations, but you know what I mean.)
- It probably has a completely redesigned instruction decoder; I found some mention of combining instructions into common bundles.
- It's capable of turning off unused portions of the chip to save power.
- It's intended for laptops and blades (of course).
- It's being designed in Israel.
That's a slim set of factoids; anyone have any more? Or any corrections?I think that it will be hard to get devices without this universal locatability soon.
while this is good for people who want to get their sports scores at 3 am while careening down the highway, it's not good for people who wish to place a private call without some agency knowing what house they are placing it from.
Your extramarital affair, your whistle-blowing telephone call, your obfuscation of whereaboutsa from your parents, etc. could all be negatively impacted by the predominance of radio-implanted devices. That's fine, just allow the option of devices that still allow privacy.
Goat sex free since 2001
I have a strog doubt that we will ever see working DRM. If it were possible to use this kind of technique to control people we would be driving cars that were limited to the speed limit. The situation is the same, consumers will shy away from products in proportion to how limited they are, so the manufacturers will find ways to avoid such a plnalty on their products
A great example of this I was a few days ago was a bin of RCA Lyra MP3 players for sale cheap, and noone was touching them, they are a pain to own simple because of the hoops they make you jump through because of DRM, and there are other players that don't, so they win.
Another example is DVD players here in New Zealand. When DVD first came out we had a crappy region code, and noone was importing suitable disks (as well as the artificial delays imposed by the US, sorry I mean studios), so people 'had' to pay to get them chipped to non-region-coded, these days you have trouble buying a player that is not 'chipped' straight from the retailer, as they know they cann't sell ones without this 'feature'.
I see DRM as a cash grab from the studios, they need to be able to point at something and say 'see - we are trying to save ourselfs, not legistale us more money!', unfortunately that part of the ploy will probably work, makes you proud to pay your tax dollars! (grrr).
A wireless mouse is a good example. Sure it could encrypt its communication with my computer, but how are the keys exchanged? How do I know that my mouse is only talking to my computer? Will this problem be multiplied by 100 when every device has some kind of radio in it?
So far only server/workstation chipsets support dual DDR channels, but VIA has plans to push this down to the desktop.
In the UK it means only one thing, and that is John Prescott the deputy prime minister (roughly the equivalent of Dick Cheyney). By the admitidly low standard of politicians he is quite a character. He single handedly made the last general election interesting by punching a voter who was stupid enough to throw an egg at him (when choosing your egg target, don't choose one who used to earn spare cash by bare knuckle boxing!).
I can't decide if the phrase "Prescott PC" would be good for sales or not. I guess it might suggest a machine with a bit of punch!
Of course the new AMD chip ClawHammer, will be able to support current x86 instructions, but it won't really help you much to run it that way.
Sure it will. When the fastest Pentium 4 is 3GHz, the fastest Athlon is "2600+", and the fastest ClawHammer is "3400+", people will happily buy the ClawHammer and run Windows XP in 32-bit mode.
(Those numbers are made up; I'm too lazy to check the latest leaked roadmaps.)
Well I found the background for the codename 'Banias' which ties into the mention that it's being developed in Israel. But I couldn't find a damn thing about the codename 'Prescott'. Anyone? Could it be really referring to the British deputy prime minister?
http://news.zdnet.co.uk/story/0,,t269-s2085308,00. html ...
Banias, formerly Caesarea Philippi, is the Arabic name for the Hellenistic city of Paneas whose name derives from Pan, the Greek god of herds and shepherds. His cult was observed in a large cave at the foot of Mount Hermon, where a source of the River Jordan emerges.
Pepperdine University has conducted digs in the area that have unearthed parts of a palace from Herod Agrippa II. Modern-day Banias is located in Israel, where the Intel design team for the new chip is based. The company typically codenames its chips after geographical features.
Why do you assume that ClawHammer will cost more than Athlon? The roadmaps I've seen show that Athlon will be discontinued soon after ClawHammer is released.
Also, Hammer is supposed to give the same performance in 32-bit mode as in 64-bit mode.
How much simpler can installation get? You take out the plate, lightly shove the card into the slot and screw it in, then reboot windoze. /hour fee to install said pci card.
OTOH, there are plenty of users in this world who pay the teenager down the street $20 to install a new NIC or modem. Or, even worse, pay the local FutureShop (insert US equiv. here) $40 + a $50
One innovation I would like to see would be a total overhaul of the concept of installing a driver (mainly for Windows, although *nix would be nice too).
Basically, I envision a ROM chip on every card, containing a copy of the device driver. User A puts card B in slot C, boots Windows, driver D is auto installed from the chip without the user clicking a button.
This would be especially useful for those situations where Windows PnP loads the wrong driver for the device, or asks the user for a file located 7 directories deep on a floppy.
Of course, both of the above are of little to no use for the average /. user, but could be a godsend for Joe Newbie.
My other sig is funny!
I hope you were kidding. Your neighborhood software hacker may also be a ham. I have to admit that since the advent of the internet I've done less ham hardware hacking and more programming, mainly python. As for spectrum space, the entire amateur radio allocation would only be enough for a few CDMA channels. Its really pretty small. I wish we could get more ham radio youngsters interested in the hobby, as I've been involved in a number of disaster relief efforts and our work was deeply appreciated. I'd hate to see that public service die.
Hammer-based CPU's will not be running 32-bit code in an emulation mode. Hammer is basically another 32-bit native x86 architecture, with some additional 64-bit registers and other 64-bit specific items included, which can be used by a programmer if they wish. It WILL run 32-bit quite well, very much unlike the Itanium.
First things first -- the Apple chips are *NOT* RISC. Reduced Instruction Set Computing means that you have a processor with vary few instructions, and you let the compiler and ASM coder do the work. This leads to larger code and simpler, cooler running processors. The Apple CPU has a huge number of instructions, but most of those are designed to finish in one clock cycle. This makes that processor a SISC(Simple Intruction Set Computing) processor. Any RISC vs. CISC arguement is (and should be) lost in the light that both are merely models, and implementation is rarely, if ever pure.
As for your assertion that the quality of Apple computers, the speed of the G5, and the quality of MacOS 10 somehow mean that they will thrive in the marketplace, you are both ignorant and naieve. Welcome to the MS dominated marketplace. Innovation, and the idea that quality breeds sales, died years ago, along with such quality OSs as BeOS, OS/2 Warp, DR-DOS, CP/M, and even earlier incarnations of MacOS. This industry has proven that the one dominant product, once entrenched, will remain the dominant product, no matter how innovative, high quality, or impressive the competition, no matter how shoddy the dominant product. In this case, I see the rise of Linux as a sign of the destructive force of MS. It takes an OS whose structure is as cancerous as a starfish, where you can cut off an arm, and that arm will grow into a whole new creature, whose economic model is as painful, where companies literally give away their products *and* the blueprints to those products, and whose longevity has already been proven, where Linux has been in development far longer than Windows 9x(I believe it was first released in '91, and I used a UMSDOS version of slackware in '95, and X was well on it's way by then) to even exist on the same platform, let alone thrive. Apple is a special case -- they were able to get a large portion of their user-base a long time ago, when Apple was popular. Many users stick with apple because it is indeed a superior platform for many applications, but new users are few and far between.
Basically, you sound like I did about a year ago, but I was talking excitedly about BeOS R5 for the x86 platform. Be was a reality check -- no single company can compete in this marketplace. A painful lesson, but very true. Learn this quickly, and you may not be as hurt when the next Apple flops commercially, like every non-MS operating product before it.
It's been a long time.
Perhaps if Eastern Germany decides to adopt Linux, they will insist on using DDR memory ;-)
--
The Cap is nigh. Time to get a fresh new account.
Actually, Intel is still pushing IB. They sponsered a multicompany IB demo that is being taken around from show to show (IDF, SNW, NetInterop, etc). They discontinued their 1X products because everyone wanted 4X and other companies started to deliver 4X while all Intel had was plain old 1X. The ironic thing is that they hit the market so early that almost every IB company has some Intel IB equipment, especially HCAs.
The party isn't over either. A lot of big companies (Dell, IBM, Microsoft, HP, and Sun) still have IB on their mind. My company is enjoying lots of interest in our IB products. 3GIO solves some problems but it doesn't provide for host to host communication, and I/O device sharing. These are two really big plusses for InfiniBand.
-- soldack