64bit processors a waste of time

"THERE WILL BE a good few people who'll say it's all just a conspiracy but there is no getting away from the fact that the whole industry needs to move to 64bit soon. Very soon. "

Why?

What is the average user doing that requires 64 bit?

64 bit is there and has been for quite some time for those that need it. AMD is providing an option that can meet some users' needs.

The question is WHY do we need to move 64 bit anytime soon? Even most power users are barely using 1GB of RAM in their systems now. It surely isn't the >4GB of addressing space. It can't be because we need 64 bit versions of Word and Netscape.

And it isn't necessarily true that we need to move because "64 bits must be better because it's twice as big as 32." For most users, most OEM boxes are more than what they'll need (and have needed) for quite sometime.

I'm not seeing how you can justify an industrywide move just because AMD released a new processor.

Kris Courter

"Buying DVD copying software is like buying a lock pick, crowbar, gun, striped jumper, balaclava, gloves, a bag with 'swag' written on it and a book called "Beginner's Guide to Burglary" all in one neat little package. It doesn't guarantee that you're intending to be a thief but there aren't many other reasons for buying it."

I think Aaron will find 'real' burglars no longer have striped jumpers and swag bags, "Idiots Guide to burglary" is a good idea though. Banning the tools will not stop the crime, even I have a crowbar, but I have not been tempted to 'jimmie' any windows open recently. People will always find a way around anti-piracy stuff (ie Macrovision), they seem to view it as a challenge.

+

Wojciech E......opera comments, excellent, I assume you will be soon selling framed versions of this email?
A proper rant.

Michael Batt

Arron replies

Strangely enough, a few years ago a friend of mine disturbed a burglar who fled the scene sharpish. The burglar dropped a rucksack that he had filled as a swag bag. When asked by the police to describe what the guy was wearing, it was trainers, jeans and a football shirt. A striped football shirt...

As I said in the original answer, professional thieves will still find tools to copy stuff but that isn't a reason to let piracy tools be sold in your local supermarket. Unless maybe you think hacking tools and spamming kits should be sold there too?

[wish you would spend as much time writing articles as you do replying to email, Arron. Ed]

Subject: Gallileo

Dear Mike:

Nice article! Don't get me wrong; I'm all in favor of an alternative to GPS.

However, the discussion about Gallileo allowing Great Circle navigation is suggesting (by omission) that GPS doesn't allow it. In fact, GPS does. The issue is the Air Traffic controllers, who want the planes arriving and leaving in nice, orderly queues (are they all English?) so that they are in known locations, and not everywhere. This topic has been well-discussed among the pilots of the world, who would like to be able to fly direct routes when practicable.

Regards,
George Byrkit

Just how is a Gallileo receiver chip going to spell doom for GPS? GPS is operational and presently being used by surveyors, cartographers, vehicle tracking companies, the military and airlines. Airlines are ALREADY able to navigate to within +/-5m world wide with GPS so how is this big news? Gallileo is vapour ware, the only thing its missing are the satellites....which depend on the European Union. And, if how the EU handled the Bosnian crisis is any indication of how things will go, I'll be collecting my pension before the Europrats have exhausted their 450->infinity committee sessions discussing it (all with the standard Europrat fat meal allowance to finance the lovely Beaujolais to wash down the duck pate). And the satellites are the cheap part, right? Oh yah.. France, Italy and Germany, the 'big boys of the EU' are on the verge of or have already broken the EU maximum government deficit ceiling and unemployment is still rising. Just where is the 10-15 billion US$ (or more) going to come from to design, build, launch and maintain these satellites? And for what? Gallileo is suppose to be a system that does exactly the same thing that GPS does....but GPS is already up, working and, most importantly, paid for by the yanks! How will the Europrats rationalize spending that sum of money for something that already exists?

All I can surmise from this article is that Tony (the author) is a magazine boffin who is quoting PR bollix from the Siemens team. 'Pretty sure it wasn't suppose to leak out?'. Yah right.

Like I said before, now they just have to do the easy stuff like build the 24 satellites (another 10 for short term spares), launch the satellites, build the earth station control, trouble shoot the system and we're away!!!! But hey, from the article, this is the cheap and easy part. As if...

Cheers,
John

Subject: HP and AMD meet in secret

Hmmmmm

Compaq buys Dec, get's alpha. HP buys Compaq, gets alpha. Intel and HP/Compaq in bed. Intel get's alpha. Itanium gets faster.

HP buys AMD. Intel/HP still in love. Intel gets x86-64???? Intel takes best of x86-64 and integrates into 'other' chip. Intanium runs 32bit code better.

Stabs in the dark....

Hmmmmm

Tim

Subject: RIAA stories and the term "theft"

Seems to be a pretty regular whine in the RIAA stories that file swapping and music ripping are "theft". What, are you guys on the RIAA payroll? If not, why are you supporting their theories?

Everything the RIAA is complaining about with regard to file swapping amounts to copyright infringement, not theft. *Possibly* one could make an argument for 'theft of services', but real theft requires that the offender deprive the rightful owner of his or her real or personal property. Copying music does not deprive anyone of access to that music.

By bandying the "theft" term about as you do, you're lending legitimacy to their cockeyed arguments and playing into their plan to maintain control of music distribution and the accompanying revenue stream.

I'll leave aside the question of whether music file swapping is ethical. I don't do it myself, so I don't have much of an opinion on the ethics. But ethical or not, downloading music is not theft.

Roy Silvernail

Subject: Re: Financial Times goofs on Windows Server launch

Now let's see: Linux was started in 1990/91, NT, as an off-shoot of OS/2, was started in 1992 as a result of the IBM/Microsoft split-up.

Linux was designed from the start to be multi-user, ergo multi-tasking; Multi-user has never been a high priority in the NT scheme of things.

Linus Torvalds despised DOS as insufficiently interesting, and I doubt he gave OS/2, such as it was back then, much thought. How he in 1991 managed to clone NT 4.0, which came out in 1997, is one of those things which must remain forever beyond our limited, mortal understandings! :)

You are right - MS = Vole. This time it seems they are eating their own tails. With any luck they'll be right up their own arses soon - we should be so lucky!

Wesley Parish

Subject: PS2 on a chip

8w consumption... Single Chip... Could this be the making of a handheld PS2? Sony is going to have to increase their revenue stream some how if the next CPU is delayed 'til 2007.....

Thanks!
Thomas

Subject: Re: Not much Wi-fi action

I've also been misled by a large Centrino ad into thinking there would be wi-fi where there was actually none. A few weeks ago I had a layover at DEN (Denver International Airport) and was greeted by a giant Centrino banner over the gate area that claimed "Wireless Internet Access Here." A wave of relief came over me--as I had feared that I would have nothing to do during my layover--and I excitedly opened up my notebook. When my laptop couldn't find any networks, however, I asked the gate staff how to get on the network, thinking there might be a weird SSID and password I needed to enter first. But I knew I was out of luck when I asked the question and the staff started looking at me like I was a crazy person.

"Wireless what?" "Wireless internet. That banner says I can get wireless internet here." "I have no idea. They just came by and put those things up and didn't tell us shit."

So it seems Intel's banner deployment team has gotten ahead of its base station deployment team. We'll just have to be patient.

Sincerely,
John O'Leary

While there might not be hot spots in all the terminals, according to Intel there are some.

BAA Heathrow Airport Limited
Address: Terminal 1 International Departure Lounge
Hayes, Middlesex
Protocol: 802.11b | Provider: BT Openzone

JFK International Airport
Address: Admirals Club
Concourse D, above ticket counter (Domestic)
Protocol: 802.11b | Provider: T-Mobile HotSpot

JFK International Airport
Address: Admirals Club
Concourse B, above ticket lobby (International)
Protocol: 802.11b | Provider: T-Mobile HotSpot

Intel hotspots finder

Bruce

I sat in LHR for a few hours on Tuesday with no WLAN (quite depressing, as I work for Intel. :-)). United's frequent flyer lounge had it for it's internet PCs, but the stations were not broadcasting the ESSID. They locked the machines down, so I couldn't pull it off their either. The SAS lounge is under renovation, but when it comes back up, it will have WLAN (or so the poster in the temporary/Air Canada lounge) said. It was definitely very frustrating.

From an Intel email address

Subject: Re: Intel makes Itanium an x86 emulator

As I understand it, Itanium _doesn't_ have an x86 processor on chip. That would increase the size of the core dramatically. What it does have is an instruction translator that translates x86 instructions one-by-one into Itanium instructions for feeding the core.

This part of the processor is extremely simple - it translates the registers one-to-one from an x86 register to an Itanium register, and translates the instructions from a single x86 instruction to a single IA64 instruction where possible, or a canned sequence of IA64 instructions where more complicated, but essentially with an architectural stop directly after each x86 instruction.

As far as I am aware, no out-of-order execution occurs and no branch prediction occurs. It's hardly surprising that an Itanium executing x86/x87 code performs about the same as a 1.5Ghz 486 would. I doubt it is able to take advantage of the new architectural features of the IA64 (e.g. predicated execution, automatic loop unrolling, register renaming) which should, in theory, allow the Itanium to outperform classic RISC or CISC/RISC designs.

The processor simply wasn't designed to run 32-bit code at speed. The intention was just to allow applications that couldn't be ported - or hadn't yet been ported - to run.

The software component is basically going to be a just-in-time compiler from x86 object code to IA64 object code. This likely will be able to analyse the program in more depth to take advantage of the features named above.

I seem to recall reading somewhere that there already was an x86/PC emulation available for Itanium systems which outperformed the 'native' support, but I can't find a link for it. That would be along the VMWare lines of an OS within an OS, rather than having applications running side-by-side on the same desktop as the native support does.

Mike Dimmick

To Mihael in yesterday's letters:

Sorry to read you disagree with my comments regarding "Abit and Asus fluffing the Nforce 2 ball", in yesterday's letters. We're the "small reseller/assembler/consultancy in Australia".

1. Firstly, you should never invite a pissing contest when you don't understand your opponent. You gleaned all of what you said from what I wrote?

You have no understanding of our business, you have no understanding of my skills or my staff, and nor can you know of our many satisfied customers. I have long and sorry experience dealing with people who criticise without having the slightest understanding or respect for other the party.

Respect, that is what I will offer you.

2. "The man says they always try the products their distributor suggests".

We are large enough that our distributors come to us to say "this is a new product, here's one for nothin'. Give it a try and let us know what you think."

So sure, we do that. We test it. We test it some more, and then we test it again.

If we think the feature set is good and stability 100%, we quote it to our clients, and eventually we sell it. Sometimes, we select a product purely on spec but we test that too. That's how we arrive at a product offering we think will work well for certain clients.

For our own selfish reasons, we like to see product work with a fairly standard set of components we keep. Inwin cases, Matrox/ATI graphics, Adaptec SCSI, Intel NICs, Maxtor/Western Digital drives and Sony storage/displays.

As I said, our clients are very happy, and we strongly believe we have about as low a hardware incident rate that you could possibly have. Our servers are high-end and mission critical, and our clients love them.

3. "The point is that the 761 chipset is the best solution for the old Athlons, and Abit KG7 is the most stable mainboard of that time suggested for small servers and alike."

When we were looking at the KG7, we thought it would be a great small server machine. We tested it. We had problems. Abit did not reply to our emails like they usually don't. We got some answers from the link I provided in my previous email to Arron.

I note to you that on that very website's forums are hundreds of tales of the crappy stability of this motherboard. This has been our experience and we have removed all of them to the BIN.

We use Tyan's excellent 761 chipset based S2460 family and have next-to zero problems with those.

Tyan respond to emails. Abit/Asus don't.

4. "I have been using it (KG7) from it's launch... shows no problems at all, never did..."

You appear to be basing your assertion on a sample of one?

We thought one was working too....... that's how we ended up selling them.

5. "He points out that they used the "high quality" Kingmax DDR".

Actually we use Kingston, which is what I wrote. Read it again. We use Kingston RAM in everything we sell and support. HP/Compaq/Apple etc.

6. "its YOUR job to know which parts work best when paired together, not just buying and assembling whatever comes your way."

We know what parts work best because we usually start with the manufacturers' recommendations. Abit have nearly zero recommendations or compatibility testing, and you can't get them to reply to emails asking for a recommendation!

7. "The other thing that threw me of the chair is that they didn't take the time to see what's wrong in "years" (talk about higher-value), but just e-mailed the hell out of mbo makers. With all the info on those (old) problems on the net I'm not surprised they didn't reply to all their e-mails."

I can't make sense of the above. What I was referring to was that over a period of years which started with us using Abit's LX6 (early 1997 if I recall) to the last KG7 we have had an appalling level of simple customer service from the original manufacturer. I was contrasting that with Tyan's performance. (Intel even too if you want - they respond, Abit/Asus don't).

8. "The other two board I won't comment but this... it's like saying that the Intel BX chipset is crap... and when a reseller tells you that you have to wonder"

Why? You've made un-informed comment on other issues?

Absolutely all of the Abit LX6s, and Abit VP6s we installed have failed completely since install. Other Intel and Tyan product we have used in GREATER numbers in that same timeframe are pretty well all operational.

Why did you generalise what I was saying by reference to a product I hadn't even mentioned? I wrote the Abit LX6 (Intel LX), Abit VP6 (VIA 694X) and Abit KG7s (AMD 761) are crap and so is their customer service. We stand by that.

AMD better take notice because they will fail with Opteron letting companies like Abit/VIA make the running.

Nice "speaking" with you.

Regards,
Connor

Subject: Has Intel spannered AMD Fujitsu deal

No, Mikey - AMD has spannered the AMD-Fujitsu deal.

Why?

Poor Performance

Low Clock Speed

HUMONGOUS DIE SIZE - Funny you never mention the 193 sq. mm GODZILLA SLopteron Die Size that Puts pout Chimpziilla performance.

And..the HUMUNGOUS 85 WATTS POWER Dissipation at a measly 1.8 GHz.

Fujitsu just saw the facts - and chose Intel.

Of course, the $1,900,000,000 DEBT of AMD and their string of EIGHT consecutive Quarterly losses and $200 Million/Quarter Negative Cash Flow may have had a small influence on Fujitsu.

Funny...none of these oh so important facts seemed to have crossed your mind.

You have become the new subject of Sanders - Blame Intel for all you faults and failures.

Paul R. Engel

Subject: AMD drops older Athlon XPs

You know what, I really don't have very many options open to me when it comes to discussing this article.

The first conclusion that I come to is that you don't know anything at all about this subject. But I would much rather asume that your not ignorant, and that you do understand that it is a completely normal process for a company to phase out older products when it comes time to launch newer ones, otherwise I guess you would have to expect AMD to still be producing the K6-2 processor that was launched some 8 years ago.

So this leaves me in a slight quandary. You see sir, if I do give you the benefit of the doubt, and I take it as a given that you actually have a clue about how companies produce and market their products, then there remains only one conclusion about why you would write this article in the manner that you have. It seems that you are simply trying to smear AMD for some unknown reason. That you are attacking them by generating completely biased, unethical, and misleading articles about the company.

Now I don't favor one over the other and I have purchased both Intel and AMD produsts in the past. I base my purchasing descissions on a value / cost basis, as any smart shopper would. But just like the old "Ford vs Chevy" confrontations that are so common amoung the less intelligent auto enthusiasts crowd, it seems that their are also a good number of fools who, for some unknown reason, tend to associate some element of their ego with a companies product. As if buying that companies product in some way imparts power and prestige upon the buyer.

Of course it is also possible that you are motivated in a much more base manner. Maybe you just own a good bit of stock in Intel.

So Mike, which is it, are you ignorant, stupid, or greedy ?

Lloyd Price

µ