Search results

Relax, I'm just going by what you said yourself, that you've fallen for the hype before. That rarely happens to me, because I always check stuff out for myself... What technology do their marketing buzzwords actually specify, and how will this technology actually be an advantage in the real...

AM2 engineering samples are 'around'. Ofcourse AMD won't want to draw too much attention to them, given their poor performance, but I'm sure that AMD would be promoting their engineering samples just as much as Intel does with Conroe, if AMD had something to promote. They were promoting their...

So you're saying that you're not knowledgeable enough to tell hype from fact. Well, I am usually right about hype or fact, and I'm 100% sure that Intel's Conroe is fact in this case.

Isn't that obvious? The Engineering Samples are basically their prototypes, so the first working examples of their new design. That's still a long way from having them mass-produced. They're still busy setting up the production lines to produce the new Conroe, and to tweak the CPU and production...

Yes, but as mentioned, the truth will come out soon enough, which would mean that stock prices would go back down, and Intel's reputation would be tarnished. Intel doesn't seem like the company that should risk such a short-term success, especially not with those consequences. But why this...

Yes, except Windows and most benchmarking programs also measure the clockspeed by using the RDTSC-instruction. It's a lot harder to fake that, and it could cause undefined behaviour in many applications that depend on proper functioning of RDTSC. So I don't think Intel would take the chance and...

Like I said, why would you not trust the current benchmarks then? Since you're not answering that question, it appears you don't have a good reason -> sticking your head in the sand.

Why not? What would Intel possibly have to gain from fake benchmarks? They aren't selling the CPUs yet, so they won't boost sales in any way. By the time the retail CPUs come out, independent benchmarks will be all over the place, so if they faked anything, it would be found out soon enough...

But that's basically a blanket statement. Which is a good excuse for sticking your head in the sand as long as you can. I believe in case-based reasoning, so to me the question would be: in what ways would the Intel Conroe samples and tests not be trustworthy? I don't see why both brands...

There are many highlevel languages, C++ is just one of them, so they're not equal. You're just not thinking about it the right way. Gates are deep inside the processor, and aren't exposed to the 'outside world'. Think of it like cylinders in a car engine. You won't ever see them, because...

Exactly, which means there are two problems: 1) We didn't know, because we couldn't verify it. So we had to reason, and believe that our reasoning was right. 2) Reasoning wasn't done in a very structured and logical way, and sometimes mistakes would be made, and not noticed for ages. In the...

DOS is 16-bit, you can't write a 32-bit or 64-bit program for DOS. The only thing you can do is write a 'DOS-extender', which is basically a 32-bit or 64-bit replacement kernel which handles switching to 16-bit and calling BIOS/DOS handlers. But that's all rather pointless when there are real...

Gates are the most basic kind of electric circuits performing some kind of binary operation. A programmer doesn't deal directly with gates. Gates are the building-blocks of processors. A processor offers the programmer instructions so he doesn't have to deal with the underlying gates and other...

Yes, although in such cases, you would be bandwidth-limited. And since the x86 has had a 64-bit databus since the first Pentium, you will gain very little today by such 64-bit code (assuming the compiler will roll out the loop and basically eliminate the loop-overhead, making it purely a...

No, 64-bit binary code is no more complicated than 32-bit binary code. It's still x86, and just like the 32-bit and 16-bit variations, the differences are small and insignificant from the point of coding complexity. And yes, I can and have written compilers. Point is just that most coders...

Apparently you don't get it. There are instructions on how to use AMD64/EM64T. Just look for the programmers manuals on either the AMD or Intel website, and you'll find some PDFs with the info you need. They've been there even before the actual hardware was on the market. I'm not sure what...

AMD might not go out of business, but a 'bad' CPU design would mean they'd be out of the CPU-market for a few years, until they can finance a new design again. Or perhaps they'd be out forever, because even now, when CPUs are their biggest product, and their CPUs are very competitive, they are...

I just don't think you know how a CPU works, and what the significance or insignificance of 64-bit computing is. My professional opinion as a software engineer is that 64-bit is quite insignificant in itself, but the extra registers and tweaks in 64-bit mode on the x86 make for some nice...

But programmers don't work at that level. They work at a more abstract level. Programs are much like maths... c = a + b. Does it matter how big a and b are? No it doesn't, the CPU does the addition, so for the human it's as simple as putting a and b in, and getting the result from c, no matter...

That's nonsense actually, especially in the case of EM64T. The instructionset is pretty much identical to the 32-bit one, you just get some extra registers and your pointers are now 64-bit instead of 32-bit. Pretty trivial changes. And that only affects assembly programmers. Most programmers...

The concept is simple to understand, yet quite far-fetched. The idea of a transistor is really not that hard. Just someone who discovered that you could create 'holes' when you 'tainted' a piece of germanium or silicon or whatever they used first. And these holes could be used to conduct or...

That is out of the question, since we had computers before the transistor, namely with vacuum tubes. And that is a very ancient technology, as far as we know not linked to any aliens. I also don't see why we would need aliens for this technology, since it's all just based on chemistry and...

Thing is, FP code is dead in Windows x64. No 64-bit application will, or even CAN use FP code. They have to use SSE/SSE2/SSE3. So only legacy 32-bit applications will be using FP to start with. And it's nothing new, Intel already sacrificied x87 for more efficient SSE/SSE2/SSE3 in the Pentium4...

Who cares? As we all know, germanium sounds way better than silicon, so hold on to your vintage stompboxes, kids :)

Sciencemark... Well I happen to know some of the guys who worked on it, I think... Sean 'Vulture' Stanek, and another one... his name escapes me. Thing is, they were pretty much the most AMD-biased coders I know, and they did everything in their power to make the benchmark run as fast as...

Afaik the cross-licensing agreement came out of the lawsuit in the 486 days, just like the Pentium name. AMD used to be one of the companies that produced chips for Intel. AMD then figured they could make their own copies of the chip, and use the same name, so they started selling AMD...

Yes, my brother has bought a new PC (P4 640) a few months ago, and he only got Windows x64 with it. Never really had big problems. Just small problems, like having to get the latest update for your defrag program, because that was the only one that was x64-compatible. He uses an X1800XT card...

Yes, but MS has developed NT since 1992 or so, and that was completely 32 bit. Since about 95, pretty much all Windows applications were developed for 32 bit. Windows 9x was just partly 16 bit because that allowed for better backward compatibility with DOS and 16-bit Windows than what Windows...

I read some news today that said that some of the Am2 CPUs may get some more cache aswell. But it isn't certain when this is going to happen, may not be at introduction, because they may want to first get yields up on the new 65 nm process.

It's fine that you don't agree, but don't call me biased. We'll see who's right in a few months.

That is a very flawed view of reality. The fastest 130 nm CPU was 2.6 GHz, I believe. And they are now at 2.8 GHz if I'm not mistaken. So the difference between 130 nm and 90 nm was only 200 MHz. I expect the move from 90 to 65 nm to be even smaller, because leakage becomes an ever larger...

Exactly, I couldn't find it either. I wonder why nobody has responded yet.

All I can say is that these AMD guys seem to be VERY confident in the company. I've even seen someone state that he believes the 65 nm process will bring AMD to about 3.6 GHz. I'm not trying to be anti-AMD, and I don't want to be sceptic or pessimistic... I like to see myself as a realist, but...

The 840D EE doesn't stand a chance against Conroe. Then again, it's up to you whether you think the difference is big enough for an upgrade. Your old system will still be exactly as fast as it is today. If that's still sufficient, then there's no reason to upgrade, is there? Unless you want...

Indeed. If gflops were all that mattered, we wouldn't be using CPUs anymore, because GPUs have a lot more gflops than CPUs, so let's run all our applications on them, and make them go much faster? Gflops just give an indication of the number of operations that can be performed in a given...

First of all the PhysX card is way faster than any videocard at physics. Secondly, the PhysX card allows for actual interactive physics, while ATi and NVIDIA can only apply physics to particle-effects and such. So the physics are far more limited, and you still need the CPU for a lot of stuff in...

Yes, and ATi uses one device for each display, so you get two videocard devices, a primary and a secondary display.

Thing is, the US doesn't really have anything better to offer than the Viper, in terms of supercars, and I use that term loosely :) But even the cheapest Porsches, Ferraris and whatever other European sportscars are still far more 'thoroughbred' than a Viper.

That's not what I understood from your opening post, apparently. You literally said "Eventually, a hardware PPU won't be needed at all". Which I interpret as "no dedicated hardware, so software emulation over hardware designed for other purposes". Else it'd be dedicated hardware, wouldn't it...