AMD Launches First 45nm Shanghai CPUs

arcticstoat writes "The wait for AMD's next-gen CPUs is finally over, as the company has now officially launched its first 45nm 'Shanghai' Opteron chips for servers and workstations. 'AMD's move to a 45nm process relies on immersion lithography, where a refractive fluid fills the gap between the lens and the wafer, which AMD says will result in 'dramatic performance and performance-per-watt gains.' It's also enabled AMD to increase the maximum clock speed of the Opterons from 2.3GHz with the Barcelona core to 2.7GHz with the Shanghai core. Shanghai chips also feature more cache than their predecessors, with 6MB of Level 3 cache bumping the total up to 8MB, and the chips share the same cache architecture as Barcelona CPUs, with a shared pool of Level 3 cache and an individual allocation of Level 2 cache for each core.'"

Comments

Which to buy now?

[Ed Avis]

Does this mean that AMD chips are now competitive on price-performance with Intel's? I mean for a fairly high-end desktop or server; obviously different considerations apply in the embedded or netbook market.

Re:Which to buy now?

[Anonymous Coward]

Depends on what you were doing, we need both heaps of memory bandwidth and good floating point so Intel wasn't competitive with AMD (though this might change now Intel have released Nehalem, aka iCore7).

But Barcelona's TLB bug certainly blotted their copybook with us. :-(

Re:Which to buy now?

[Anonymous Coward]

This news is about a server/workstation chip, and I don't do any purchasing of those. As far as desktop chips are concerned, AMD was ALWAYS competitive on a price-performance basis. The key word there being price.

Re:Which to buy now?

[MrHanky]

According to Anandtech's review, it's highly competitive for database servers. http://it.anandtech.com/IT/showdoc.aspx?i=3456

Re:Which to buy now?

[Anonymous Coward]

Ed...A fairly high-end desktop isn't anything close to comparable to a high-end server. AMD's have been superior for building high-end servers for some time now, thanks to bandwidth considerations.

Re:Which to buy now?

[Ngarrang]

Does this mean that AMD chips are now competitive on price-performance with Intel's? I mean for a fairly high-end desktop or server; obviously different considerations apply in the embedded or netbook market.

What apps are you running? Do your apps take advantage of multiple cores (regardless of the speed) or risk pushing 17.6GB/sec of memory bandwidth? Then go for the new AMD or Intel CPU, as they are both stupid fast. If not, then you might be better served with a previous gen CPU at a much lower price.

I mean, really, we are talking about very small differences in speed at this point, right? Would the average person actually be able to detect the difference? I am all for ever faster processors, as it allows us to run ever more complex software, but I don't think either processor is a loser, so you would be good to go with either one.

Add in good quality components to match like SATA/300 drives and loads of memory to actually realize all of that speed.

Re:Which to buy now?

[Anonymous Coward]

I would go with AMD for servers because of the lower power/heat for the complete system. FB DIMM's run hot as to Intels southbridges, not to mention Intel understating how much heat their CPU's will generate compared to AMD. Plus, if you go multi CPU, AMD has the more mature way of handling it efficiently.

Re:Which to buy now?

[nizo]

It is always best to do your research when buying a new chip so you don't get shanghaied.

Re:Which to buy now?

[JAlexoi]

Actually, AMD CPUs already are competitive on performance/price with Intel's CPUs. Do you know how much those Core 2 Duos cost compared to AMD Athlons?

Re:Which to buy now?

[MBGMorden]

AMD is still doing OK on price to performance, but what I think is hurting them is that the margins are not the same, because CPU's as a whole are just so cheap now. I remember back when an ENTRY LEVEL off-brand chip like a Cyrix (or, AMD) cost $150. "Intel Inside" cost $350 or more starting out. We'll call that a 50% ratio. The AMD (and certainly not the Cyrix) chips were not quite as fast as their Intel competition, but to a high school student who was making $50 per week part time, I certainly didn't mind that small gap in performance.

Now today, the ratio has changed. AMD still beats Intel in price to performance, but not by the same ratios, and the margins are much different. If a $40 AMD chip is slightly slower than a $65 Intel chip, then that's great, but the difference is only $25. A lot of people are going to be pretty quick to drop the couple of extra $$ for the Intel chip. Particularly now that I've noticed that, quite often, when you go over to the motherboards, Intel compatible motherboards are often coming in just a bit cheaper than AMD motherboards.

Now personally, when I can, I still buy AMD at least 50% of the time, but the only reason I do that is because I remember the days when Intel's competition wasn't as tough, and I remember those days of $350 chips from them. I only support their competitors to ensure that that situation doesn't repeat itself. For people without such a goal though, Intel is certainly tempting.

Re:Which to buy now?

[blair1q]

No, they weren't. For the past year Intel has boxed AMD in with chips at the same performance and lower price, or the same price and higher performance, or both.

And Intel has had performance segments (QX*) stretching well above AMD's, and pricing segments (Atom) well below AMD's.

AMD's short-lived price/performance superiority in the desktop sweet-spot in 2004 and 2005 has left many people thinking they're still in that position. That hasn't been true since Core 2 came out. HyperTransport gave them a slight edge in very-high-end servers for certain applications, but Intel stayed near them with reliably higher clock speeds, and is coming out with QuickPath in four days, wiping out those few use cases where AMD can make easy sales today.

What I'm saying is, right now you are likely to choose Intel in almost all situations, if you are objective.

Re:Which to buy now?

[rgviza]

cycles aren't everything in all cases... AMD still has more system bandwidth, which speeds up everything when talking about IO bound applications. FSB speeds up every aspect of the computer.

The applications where AMD is superior are IO bound applications like database servers, and music production.

Intel is better for video because you are dealing with a limited number of streams and it's computationally expensive, so is CPU bound.

With audio you can have hundreds of streams (often 4-6 per fader on the mixer), and at 24/96, will quickly overwhelm any intel based system. Since a lot of us use DSP cards ( think of it as GPU for sound) the data path capacity, especially to the DSP processors (PCI/PCI-e) is very important, and Intel simply can't touch AMD in this respect.

AMD architecture simply has untouchable plumbing. If you will notice, Apple is looking for a new chip vendor. This probably has a lot to do with it since most audio professionals use Apple gear.

If Jobs and Co. were smart, they'd offer both intel and amd architectures, depending on the job being done. Intel is fantastic for video and a lot of pro video peeps use Apple gear too. Those are two market segments that couldn't be more different in their requirements. To be the best of the best for multimedia, Apple needs to either build a new architecture or offer both AMD and Intel.

-Viz

Re:Which to buy now?

[afidel]

AMD had ruled on price/performance since the original Athlon shipped. It wasn't until Core2 that Intel was competitive again. For anything memory bound AMD still had the lead until Intel shipped Nehalem a few weeks ago and they won't be shipping the MP chips until Q2-09 so they don't matter to me yet.

Re:Which to buy now?

[RicktheBrick]

If one is building the fastest computer in the world, would not one want the best microprocessor? Why do the two top computers(IBM's roadrunner and Cray's jaguar) both use amd microprocessors? Intel is now producing a chip with 6 cores and a quad with hyperthreading so it is equivalent to 8 cores, so why aren't these chips being used in the fastest computer?

Re:Which to buy now?

[0123456]

"AMD is still doing OK on price to performance, but what I think is hurting them is that the margins are not the same, because CPU's as a whole are just so cheap now"

$1500 for an Intel Quad-Core 'Extreme' CPU is 'just so cheap'? You must earn a lot more than I do.

The only reason AMD CPUs look cheap is because AMD don't have anything at all to compete with Intel's desktop CPUs at the high end; they'd love to be able to charge $1500 for their fastest desktop CPUs, but no-one would pay it.

Re:Which to buy now?

[Chandon Seldon]

For the past year Intel has boxed AMD in with chips at the same performance and lower price, or the same price and higher performance, or both.

That's been true in some price ranges, but Intel hasn't trumped AMD across the board any time recently. There's always been a couple price ranges - and usually the relevant ones like $120 to $150 - where AMD has a better product.

pricing segments (Atom) well below AMD's.

Geode?

I'm not trying to say that Intel hasn't been "the winner" for the past year or so, but it certainly hasn't been as one-sided as you're claiming. AMD has been selling chips, based on being the best choice for individual consumers, the whole time.

Re:Which to buy now?

[MBGMorden]

$1500 for an Intel Quad-Core 'Extreme' CPU is 'just so cheap'? You must earn a lot more than I do.

High end has always, and will always, be expensive. That's not the area of the market where most of the chips are sold anyways.

The low end market sees far higher volume, and in that arena, AMD's introductory processors start at around $25 (with a heatsink & fan - I remember paying more than that for JUST a heatsink - no CPU or fan included), and Intel starts at around $40. That's for a 2.0Ghz and a 1.8Ghz chip, respectively, and that's quite capable of doing most of what your average user would want to do these days. If you have to have dual core, Intel retail starts at $50 whereas AMD Retail starts at $60.

Both companies have quad core offerings under $200. So yeah, that's pretty danged cheap in my book.

Re:Which to buy now?

[Bengie]

other than my cousin's p100 managed almost 2xs the fps as my cyrix 233+.. cyrix had no fpu.. :*(

other than that, your argument holds mostly true to now-a-days and simular performing chips.

Re:Which to buy now?

[Anonymous Coward]

AMD's short-lived price/performance superiority in the desktop sweet-spot in 2004 and 2005

Price/Performance is a matter of what application you use. Before the Athlon, AMD had a better price/performance for office applications, but it was worse for everything else. With the first Athlon, AMD was better with Office and Games, but was behind in multimedia. The sweet-spot you are referring is when AMD was superior in every way, both from price/performance point of view and from a pure performance point of view. Even now, AMD is still better for low end computer.

Personally, I still recommend AMD and an Athlon X2 for offices, particularly because I can use ECC memory (without having to buy a $250 motherboard).

Re:Which to buy now?

[0xABADC0DA]

First, Intel didn't have better price/performance until recently. Certainly well into core 2 duo period the AMD Athlon 64, X2, etc were much better price/performance.

Second, on the low end:

AMD Sempron 1150 2ghz: $22
Intel Celeron 430 1.8ghz: $39

That Sempron is much faster than that Celeron. Atom is cheap for the processor, but the other parts cost more and use a lot of power (%50 of total power, 15 watts for chipset on Intel's Atom mITX board). Why do you think netbooks only get ~3hrs with 4 watt processor? Because the rest of the intel chipsets suck.

Combos:

AMD Sempron 3000: $50
AMD Sempron LE1150: $60
Intel Atom: $70

Prices from newegg.

Now lets think about this, AMD has matched Intel in price/performance despite a 2x larger process. What does that say about the respective merits for AMD and Intel designs? If Intel stumbles in advancing the process they are going to fail, because AMD has clearly better designs and engineers.

Re:Which to buy now?

[Anonymous Coward]

You need to stop worrying about the clock speed so much. Looking at two different processors and basing the comparison on clock speed is foolish. Go look up some prices, and then go look at some benchmarks and then come back and we'll discuss. Oh, and while you're at it, look at the motherboard prices as well. Let's not forget that a PC can't run on a CPU alone... it needs other components.

Factor in the cost of motherboards, and right now, you are likely to choose AMD in almost all situations, if YOU are objective.

Re:Which to buy now?

[MBGMorden]

A lot of those are designed for heavily parallel operations, and as such other concerns come into play than pure chip speed. One VERY major issue is heat dissipation. Another, even for super computers, is cost. If it takes 400 AMD processors to do what 350 Intel's could do, it might still be worth it if the price difference is such that the AMD's would still come out cheaper.

Particularly when comparing on a desktop level, I don't think you'll find ANYONE who will claim that AMD is making faster chips than Intel these days. They may be more cost effective though, and that will sway lots of people. From my gathering, just like me, most homebuilders still seem to be leaning towards AMD over Intel, though the OEM market still favors Intel more like always (though AMD IS at least available, particularly from HP which seems to use their chips very heavily).

Re:Which to buy now?

[Chandon Seldon]

You seem to be thinking of Intel as the "good brand" and AMD as the "off brand". That made sense in 1997, but now it's like Toyota vs Honda - both companies make top quality products.

Re:Which to buy now?

[i.of.the.storm]

I think you're at least partly correct, but there are certain AMD processors which are as fast as an equivalently priced Intel, although I wouldn't put my money on any being faster. Also one thing you forget is that AMD motherboards tend to be cheaper, which knocks a bit off the overall price. Of course, I don't have any numbers to back myself up with so I might be completely wrong, but this was the situation a year ago when I built my desktop, although I will admit that I was going to buy AMD anyway and was just happy when I found that the motherboards were cheaper.

Re:Which to buy now?

[MBGMorden]

other than my cousin's p100 managed almost 2xs the fps as my cyrix 233+.. cyrix had no fpu.. :*(

Cyrix did have an FPU - it just wasn't a really good one :).

A big problem with Cyrix was their P-rating system. The integer performance on the Cyrix was a bit faster than that of a Pentium, so they'd take 150Mhz chip for example, and call it a P-200, insinuating that it performed as well as a Pentium 200 (which in integer operations, it was close). However, the floating point unit on the Cyrix was worse than a Pentium, so it actually worked backwards. That 150Mhz Cyrix, despite being touted as a Pentium 200 equivalent, was more like a 100 Mhz Pentium in floating point operations.

They also started doing stupid things with it later in the M-II series. The FPU still had the same issues, so Cyrix continue to rate on integer performance. Their own benchmarks would show a Cyrix M-II of a similar rating being close to, but just a bit slower than a Pentium II. Problem was, the "close to it" gap was actually the difference between the actual Pentium II models in some cases.

For example, when you looked at the M-II 300 (233Mhz) compared to the Pentium II 300, benchmarks (even on CYRIX's site, so any bias would be on their side), it was close, but still lagged slightly behind. If you pulled up the comparison between the M-II 266 (200 Mhz) and the Pentium II 266mhz, then the same thing. HOWEVER, if you actually looked, you'd see that the M-II 300's score was EXACTLY that of the Pentium II 266mhz . . . so why was that chip P-rated at the level of a Pentium 300mhz and not a 266Mhz?

It was a frustrating time to be sure, but my primary computing interest at the time was emulating consoles like the SNES and Genesis and such, and heavy FPU wasn't needed anyways. It turned many against the whole idea of a "performance rating" though, and when AMD started doing it with the Athlons it was a dissapointment. These days though even Intel's chips don't compare right to their other chips so everybody seems to have adopted obscure model numbers instead.

Re:Which to buy now?

[Bengie]

too bad you're not comparing AMD's new chip against Intel's new chip. Intel's IO bandwidth is benchmarked at 15BG/s and AMD claims ~17GB/s. These numbers are pretty close for there to be a 'huge' difference. On top of that, if you compare core for core, AMDs Shanghai is showing to be about 20-30% faster per clock than it's previous generaton while Intel is about 100% faster. and AMD is bragging about a minor % drop is power, while Intel about cut it in half (cpu power, not total platform power). overall Intel is packing less heat/electricity, more raw CPU power, and simular IO bandwidth. Lets see how this pans out

Re:Which to buy now?

[CAIMLAS]

Er, huh?

AMD dominated the price/performance war with Intel from the time they released their K6 chips - that'd be 1997 (hello, remember the "sub-$1000 PC"? that's thanks to AMD). This was the case until just recently when things started to go multi-core - and even then, AMD had a bit of resurgence while playing leapfrog with Intel.

From about 1999 to 2003 AMD was way, way ahead of Intel; Intel didn't pull ahead of AMD in terms of simple performance (without spending close to a grand for a processor) until the release of their Core based processors. Their performance started to improve quite a bit with the M based processors, but your common desktop price/performance was still dominated by AMD.

Arguably, AMD's memory management is still better. We'll see how this generation hashes out.

Re:Which to buy now?

[slaker]

I certainly agree that Sempron is a generally better CPU than Celeron, but the thing that keeps me from looking at them is frustration with AMD motherboards and chipsets.

In Intel-land, I buy a made-by-Intel DG31PR or 945GCPE motherboard for about $60 and in either case I've got a board that is basically bulletproof, and even if it's not, I know that I'll be able to obtain a replacement and that Intel will seriously and consistently support that product.

In AMD-land, I can buy nVidia-based board with a bunch of idiotic hobbyist features that is always a tiny bit flakey, or I can buy one of a dozen AMD-based chipsets that are generally undifferentiated from one another and all available at approximately the same price. In either case, in a year and a half, when one of them breaks, I find that the manufacturer can't provide an exact replacement, won't replace unless I pay shipping both ways, has an utterly broken RMA process (coughAsuscough) or doesn't even have an office in the USA. The CPU might be $15 cheaper and the motherboard might be $20 cheaper, but I'm still not going through that through another generation of the machines I have to support.

Re:Which to buy now?

[0xABADC0DA]

"Hey, if you want me to take a dump in a box and mark it guaranteed, I will. I got spare time. But for now, for your customer's sake, for your daughter's sake, ya might wanna think about buying a quality product"

There are dozens of AMD boards that are much cheaper than the Intel boards, with more variety of chipsets and designs. And the companies that make them have to prove themselves. They can't just rely on their good name while shipping a POS (such as Intel's Atom mb).

in a year and a half, when one of them breaks, I find that the manufacturer can't provide an exact replacement, won't replace unless I pay shipping both ways, has an utterly broken RMA process (coughAsuscough) or doesn't even have an office in the USA. The CPU might be $15 cheaper and the motherboard might be $20 cheaper, but I'm still not going through that through another generation of the machines I have to support.

While you say $20 cheaper, you could also say $20 instead of $40, or 'half the price'. You could buy twice as many as you need for the same cost as buying Intel. Do you really expect half of the AMD mb's to fail? That seems outrageous to me.

By arguing for exact replacement parts, warranties, "I have to support", etc, I take it you are in an organization and have some kind of specific requirements. Basically your reasons are the same as why people bought $200 Sun keyboards... For most people, it's cheaper and easier to just buy a new, better mb 1.5 years later, should the old one die, than it is to get an exact replacement.

Re:Which to buy now?

[compro01]

Simple : hypertransport. When you're dealing with LOTS of processors in parallel, it kicks ass. Intel has had nothing to compete against that advantage until now with their quickpath thing.

But dealing with a small (5) number of processors, Intel's approach of going through the FSB has been Fast Enough(TM), though I doubt it would work well for 8 or even 6 cores, so they're ditching that and doing as AMD did.

Re:Which to buy now?

[blair1q]

No, Intel held the performance crown until the early 00's. AMD was the cheap chips, but didn't really have a chance in performance. Intel let that happen by focussing on Itanium and ignoring x86 improvements. That snapped back when they realized the mobile M-core could be enhanced for desktop operation, and then modularized for multi-core desktops.

It's like AMD took the lead while Intel hobbled into the pits to fix a flat, which took 12.8 seconds, and then Intel came right up and passed them in the next lap. AMD's a lap down now, and is realizing it just didn't bring enough horsepower to the race. This thing is all over except the turning of the donuts, the spraying of the milk, and the back-flip onto the tarmac.

Re:Which to buy now?

[tomz16]

Why exactly is the sempron better? AFAIK, this Celeron is based on core architecture, and should actually be appreciable faster than the sempron (even with a 200MHz clock disadvantage)!

Re:Which to buy now?

[sgt scrub]

I agree with you 100%. AMD has been seriously injured because of their inability to shrink die size on demand. If they had Intel's cash things would be different. What is the saying? If frogs had wings they wouldn't bump their ass.

Re:Which to buy now?

[Kjella]

Now lets think about this, AMD has matched Intel in price/performance despite a 2x larger process. What does that say about the respective merits for AMD and Intel designs?

Nothing. But it says a lot about their respective profit margins.

Re:Which to buy now?

[Gr8Apes]

Intel makes lots of claims. I'd wait and see the real hardware performance before believing Intel claiming such a large jump in I/O.

Re:Which to buy now?

[Gr8Apes]

Untrue. You'd choose Intel if clock cycles over power consumption are important. If you're footing the power bill and AMD CPUs are adequate, you'd quite likely buy anything but Intel. If you need more power than 2 Intel chips can provide, you are also only left with AMD or non x86 chips. Intel is wholly outclassed in the 4+ CPU market.

I have both an X2 and a Q6600. The X2 runs far cooler, although it is less powerful. However, I can get a Phenom in the same TPD range that comes near the Q6600 in performance and costs less in the bargain.

The choice is not nearly as clear as you make it seem.

Re:Which to buy now?

[slaker]

I'm normally maintaining computers by the fleet, by myself. I want identical machines that all perform the same way, use the same drivers and need a minimum of technician involvement.

My experience is that AMD-based motherboards, seemingly regardless of manufacturer, are much less reliable than Intel-made Intel motherboards, particularly after the first year of operation. I can provide exact numbers if you'd like, but generally what I see with vanilla Asus, Gigabyte and MSI board for AMD products are failure rates hovering about 20% per year after the first year. USB will stop working. Caps will leak. The onboard video dies.

In a given year, I might be responsible for building 300 new systems. Three years ago I started to transition from AMD to Intel. This is my first all-Intel year, but in three years I've RMA'd four Intel brand desktop motherboards, compared to dozens from Gigabyte/MSI/Asus over the same time period.

Re:Which to buy now?

[xrobertcmx]

Asus might not have a completely reliable RMA process, but how often does it have to be used? They have one of the lowest failure rates in the industry. I have used them almost exclusively on every build, with one exception being an Abit, and while it is completely subjective, still have customers out there with '01 machines still humming along.

Re:Which to buy now?

[Penguin Follower]

Actually, from what I've heard, Intel designs their mainboards, but farms out manufacture of them to Foxconn.

Re:Which to buy now?

[slaker]

A couple vendors I use actually notate certain motherboards as being Intel by... but they arrive in the same packaging as all the others.

I don't care. I very strongly suspect that there's a difference in the quality of components that Intel specifies vs. what AMD and nVidia do for their designs. I further suspect that at least in some cases, the people making $40 AMD boards are cutting corners in order to be profitable at that price point.

'm sure I could buy 3 AMD boards for every two Intel, but there's a cost associated with putting a new one in, and frankly I'd rather that my machines just not have those kinds of hardware issues in the first place. The end point of motherboard reliability is really the deciding factor for me.

Re:Which to buy now?

[Anonymous Coward]

The fastest AMD 386 was clocked at 40MHz, significantly faster and cheaper than Intel's fastest 33MHz 386.

AMD's 486 series were much less expensive than Intel's but offered nearly identical performance.

AMD's 486DX4 CPUs were much cheaper than those from Intel and they also offered higher performance than the Pentiums of the time.

The K6 CPUs walked all over the P2/P3 CPUs when it came to integer performance or gaming (thanks to 3DNow!).

The Athlon absolutely destroyed the P3 and P4 in all aspects.

The only times that AMD really fell behind Intel was with the K5 and their current batch of CPUs. They'll surpass Intel again, you can be certain of that.

Re:Which to buy now?

[turing_m]

It's like AMD took the lead while Intel hobbled into the pits to fix a flat, which took 12.8 seconds, and then Intel came right up and passed them in the next lap. AMD's a lap down now, and is realizing it just didn't bring enough horsepower to the race. This thing is all over except the turning of the donuts, the spraying of the milk, and the back-flip onto the tarmac.

I think I've just witnessed the Ford Edsel of car analogies. Thank you.

Re:Which to buy now?

[aliquis]

As long as you don't overclock, care about a stable motherboard (not a flamebait, you need a 790FX one for the best phenom or your system is very likely to get unstable) and temperatures AMD do offer decent alternatives at the low end.

Comparing a Phenom 9950 with the E8500 or Q6600 seems reasonable for me (I may have get some number wrong.) Q9550 or whatever it's called is more expensive. Of these I think I would go with the quad Q6600 even though games may get more benefits from the faster cores in the E8500. But in terms of speed there isn't that much difference in them (Q6600 and Phenom 9950 that is) is there? But the 9950 is a 130 or 135 watt chip, is probably pushed quite hard since AMD had to increase the voltages for each modell and don't overclock much. The Q6600 got plenty of juice left if you want to trust anyone which overclock their chips, so it seems like a much safer bet anyway.

But un-clocked the difference isn't that huge. Just more advantages with the Intel chip.

Re:Which to buy now?

[aliquis]

But isn't Intel supposed to catch up when it comes to integrated memory controllers and maybe FSB speeds to when it comes to this next generation?

Re:Which to buy now?

[pxlmusic]

this is something that i was really struggling with.

i have started to buy the parts for a audio workstation centered around Ableton Live 7 and a whole slew of VST's -- many of which are very CPU-intensive.

i never really got a very clear answer on which would benefit me the most CPU-wise. i'm willing to go either way. i'm a fan of AMD for many reasons, but i'm not sure which would deliver more performance with those VSTs: AMD or Intel (going quad core either way)

Re:Which to buy now?

[blair1q]

The only times that AMD really fell behind Intel was with the K5 and their current batch of CPUs. They'll surpass Intel again, you can be certain of that.

No, I've seen their roadmaps.

AMD is 1-2 years behind and its marketing department is paddling away from the Titanic as fast as it can.

Intel has changed the way it does business completely since Itanium balled them up. They only develop for a working fab process, and they only start fabs with known-good chip designs. They have tight integration of their design and fabs (AMD is pushing their fabs away) and much more significant design advances in the pipe.

AMD just pushed Fusion back by two years; and it wasn't going to impress anyone running a Larrabee anyway.

Unfortunate name

[FlyByPC]

"AMD Shanghai -- the perfect CPU for your newly-acquired botnet!"

Re:Unfortunate name

[Loibisch]

Intel Nehalem, the "perfect CPU for your native american tribe."?

Don't read too much sense into it...

Re:Unfortunate name

[Yvan256]

I still find it funny that Apple computers now use intel Core 2 Duo processors.

Re:Unfortunate name

[poetmatt]

Shh. Remember apple runs "fast" and is "glorious for multimedia!" somehow we skipped Linux and AMD, but hey, want to pay 2x as much for half the performance?

Re:Unfortunate name

[Anonymous Coward]

ahem.... Woooooooosh!

Re:Unfortunate name

[raddan]

You seem to have forgotten that software matters too. CoreGraphics is both extremely fast and easy to use.

Re:Unfortunate name

[poetmatt]

Right, because compiz/xgl could never compare to an apple fanboy. Is that what you're saying? Whoops.

Re:Unfortunate name

[broken_ms_windows]

hell im still using 3 ppc mac's and there fine for
  what i ask of them.no need for me to get an intel mac just yet

Re:Unfortunate name

[Freultwah]

Somebody is so obviously humour-impaired that it is not even funny anymore. In short: whooooosh!

Re:Unfortunate name

[poetmatt]

Oh crap, did I? I blame Thursdays and the remaining hour of work.

Re:Unfortunate name

[Niten]

Shh. Remember apple runs "fast" and is "glorious for multimedia!" somehow we skipped Linux and AMD, but hey, want to pay 2x as much for half the performance?

What? Are you actually suggesting that AMD's processors perform better than Intel's offerings? As a company AMD may have its strengths, but they cannot even remotely compete against Intel's Core 2, let alone Core i7.

As for OS X versus Linux, the majority of multimedia encoding and decoding operations occur in userspace, so the Linux kernel has no speed advantage to offer there. Where the two operating systems do differentiate themselves is with their multimedia APIs and codecs; as a user of both Linux and OS X, it saddens me that OS X is so far ahead in this regard.

Re:Unfortunate name

[professorfalcon]

I still find it funny that Intel Core 2 Extreme processors have 4 cores.

AMD needs more advances...

[Anonymous Coward]

They need to push for smaller fabrication.. intel released 45nm a while back..

Re:AMD needs more advances...

[MikeDirnt69]

That's why girls prefer AMD.

Please tell me AMD is not betting it all on shrink

[Anonymous Coward]

If AMD is betting the company on an improved production process, they've already lost. More of the same hasn't even worked for Intel, but unlike Intel, AMD will not be able to bounce back from that mistake.

Congratulations!

[Anonymous Coward]

AMD is now only 1.5 years behind Intel!

Great job!

At this rate, by the 32nm node, they will be only 2 years behind Intel.

Re:Congratulations!

[postbigbang]

If you're hinging your comments on the wafer size, you're blinded by Intel propaganda. Take a look at AMD's SPECjbb numbers, their cost per socket/core, and their threading for virtualization. Then perhaps you'll stop being an Intel shill.

Re:Congratulations!

[jandrese]

Not that I'm an Intel fanboy, but SPECmarks are hardly the first thing that come to mind when I think about "reliable unbiased metrics". In fact the term that usually comes to mind when people start quoting SPECmark numbers is "lie".

Re:Congratulations!

[postbigbang]

You haven't examined SPECjbb, then, have you? It's a Java-based business transaction kit that seems to have quite a bit of both fairness and repeatability; it's not easily manipulated, like others I've seen. After running more than several thousand runs with it, I find it pretty reasonable in terms of systems performance comparison, rather than motherboard/subsystem/peripheral benchmarks-- which shed light only on one specific characteristic of a machine, or are operating systems-specific. Admittedly, it doesn't do things with GPUs, network I/O, and the like.

But to call it a lie is specious.

Re:Congratulations!

[Anonymous Coward]

Java based-business transaction quantifier, without network I/O? Not trying to be a troll, but how useful is that?

Re:Congratulations!

[RightSaidFred99]

I'm fairly certain he was referring to feature size (nm), where AMD is 1.5 years behind. He's undeniably correct there, your (correct) implication that AMD is still in the race on performance, price, and energy use is true but unrelated to his point.

Now, you may ask "so what, they've been ahead in manufacturing". And my answer would be to look at the financials. Smaller processes are cheaper, ergo AMD leaks money like a sieve - _that's_ "so what".

Oh please.

[Moryath]

The two companies take turns one-upping each other for the bleeding edge, but every time (10 years running) I've specced out a mid-range (home gamer, single CPU motherboard) to low-end (grandma's email/photo machine) machine, AMD's been the way to go. It's a lot like trying to decide which company's video boards to pick if you're trying to make a game machine without breaking the bank.

Some people are Intel partisans, some people AMD partisans. Benching them and looking at spec, I've consistently found that AMD's got faster chips (for the same $) up to the "sweet spot" in the curve where price starts shooting upwards during the times I've been buying, but I also know there were times I was not in the market when Intel had done a price cut and AMD hadn't caught up.

I'm not going to call someone an idiot for their CPU choice, as it's a long-term purchase decision that has to be balanced with other factors (motherboard choice, RAM, video board, power concerns, cooling solution, etc) anyways. In fact, I recommend consumers try to stay OFF the "bleeding edge" because they're basically throwing money away on it; even if you buy the latest, hottest chip right from the factory it's obsolete by the time you get it home. Your best bet is looking at the curve, because there's always a spot (usually between $150 and $250) where the price starts to jump up exponentially for only an incrementally "faster" product. Buy at the spot beyond which the relationship between price and performance fails to be linear and you'll turn out pretty happy.

Re:Oh please.

[Anonymous Coward]

Mod parent up.

Re:Oh please.

[Colonel Korn]

Since the C2D arrived, I've been going with Intel. I usually don't overclock, but the C2D handles it so well with such little effort that I based my purchase of a $200 ~2.2 GHz chip on that alone. With the addition of a $30 heatsink I had it at 3.4 GHz with temperatures under 60 C at load (below the temperature seen at stock speed with the stock cooler, implying good longevity), back when there were no 3.4 GHz Duos and the closest thing cost about $1000. I have several friends who had never OCed before who did the same thing, all ending up with 2.8-3.6 GHz chips that all are still working perfectly and speedily ~1.5 years later.

Re:Oh please.

[LWATCDR]

I tend to agree. The honest truth is that and AMD 780G motherboard and one of the low power X2s makes a great system for most users. If you want to play games throw on a 3870 or if you really need it a 4850.
I just built a system for my wife with an ASUS 780G motherboard, X2 and 4 Gigs of ram. Total cost was under $200 and it runs very well.
If you not into high end gaming then AMD seems like a great choice.
I can hardly wait for 45nm AMD desktop CPUs to start showing up. I really want one.

Re:Oh please.

[mfh]

The two companies take turns one-upping each other for the bleeding edge, but every time (10 years running) I've specced out a mid-range (home gamer, single CPU motherboard) to low-end (grandma's email/photo machine) machine, AMD's been the way to go. It's a lot like trying to decide which company's video boards to pick if you're trying to make a game machine without breaking the bank.

I have to agree. When the Quad cores shipped, I tested them and I compared the speed per dollar. AMD was half price for performance. If you understand that a decent graphics card, and having a nice power supply to run the show, then you are ahead of the game.

Value is what I look for when I buy things, not bleeding edge performance. Because money isn't a factor I could easily spend to get the best available but I have too much remorse wasting an extra thousand bucks on a slight increase. It's not worthwhile to me, considering the frame rates in games I get on my AMD system are good enough for 25man raiding in WoW, or world pvp.

Re:Oh please.

[Lonewolf666]

Starting with some of GP's requirements (game-capable PC but at a reasonable price) and wanting to use ECC RAM for reliability I ended up buying an AMD last year. It is an AMD Athlon64 X2 EE 4600, a dual core with 2x2.4 GHz, not overclocked. In practice, this machine is fast enough, especially considering that I don't run the very latest games.
The deciding factor in terms of Intel vs. AMD was that ECC capable mainboards for Intel are expensive. The cheapest C2D would have been not much more expensive than the Athlon (and a tad faster), but on the mainboard side the difference was 100 Euros or more.

Re:Oh please.

[daedae]

Frequently?

Do you have any actual backup for that? The last three desktops I've owned have all had AMD processors, and the only thing that's gone bad on any of them was the AGP slot went bad on one of the mobos after about 5 years.

Re:Oh please.

[Anonymous Coward]

Mid and high-class gaming rigs have been Core 2 Duo only for a year now.

Admittedly, I still recommend sticking on the Athlon 64 X2 5000+ 65w if you want decent non-gaming power on the cheap, though.

Re:Oh please.

[LWATCDR]

My Athlon 4200 X2 plays world in conflict and FSX just fine since I added a 3870. I would say that an X2 5000 will do just fine for a mid level gaming rig with an good video card.

Re:Oh please.

[sumdumass]

No he doesn't have any data to back that up. Well, not unless he is going back to the K6-2 and K6-3 days. I have used and supported many AMD system and only had 3 of them with problems with chipsets. Actually, this problems weren't even the chipsets, they were with cheap board manufacturers who used counterfeit parts and got a load of bad resisters. The ECS K7S5A had a short run on that.

Of course he might be referring to the older VIA chips that allowed the user to select optional components during the driver installs for windows 98 and early 2000 versions of the drivers but the problems there was with the USER not picking the correct options. I don't consider that as a bug in the chip either, that was back when people were expected to know what they were doing when mucking around with system drivers.

If you want to talk about buggy, I just put a Quad core pentuim system together. The store didn't have the mainboard I wanted so I settled for a paired down version of it. It said it supported my chip right on the box. After 3 days of blue screens, swapping memory out, and other problems, I found some bad reviews for the board online that said I needed a certain Bios Update before it would work with the quad core intels. Sure enough, there was a bios update at intel's site (the board shipped with a bios about 5 revisions out). I downloaded it and wouldn't you know it, the bios itself wouldn't fit onto a floppy (1.5 meg). SO I burnt a CD but the floppy boot disk couldn't find the SATA cdrom. So I spent the next day or so attempting to make a bootable CD with this bios update on it. No go with the freeware utilities and I ended up having to format the hardrive in Fat32 and copying the bios file there just to flash the bios. The techs and the shop I got the board from asked why I didn't just use the windows version of the bios updat and couldn't understand that if windows wouldn't stay running long enough to download it or even open the CDROM to run it, it probably wouldn't be wise to attempt to flash the bios from the OS in this state if I somehow did manage to get it on the machine. Anyways, 5 days to get something running that should have taken 45 minutes to 2 hours altogether. And yes, this was my fault because I didn't scope the board out before I decided to use it and I didn't know how to make a bootable CD that I could add files to. It wasn't the chipset's fault but I'm sure someone could somehow blame it if they wanted to. That's what some of the reviews were saying. Human error is often redirected to things or objects which is probably what the gp was doing.

Re:Oh please.

[Anonymous Coward]

Quick look at newegg shows comparable mb prices for either LGA775 OR AM2+/AM2 boards.

CPU prices are roughly comparable given the higher overall performance/clock of Core 2 based CPUs.

Core 2 FTW for me ATM, as I don't consider power use at all, and that's ALL that AMD has going for it ATM.

Although that said I'd wait until next year when Intel's new CPUs are out with integrated memory controller (forget what they call it), should pretty much put the final nail in the AMD actual performance coffin truly leaving them with only power consumption advantage if even that any longer.

As to mobile processors: AMD may as well just stay home. Their performance is horrible, and their power usage isn't enough better than the P & T Intel series, leaving only slightly cheaper cost a factor. Hell, even rpe-built notebooks with a decent GPU are within a few dollars of each other, but not worth it overall as back to the first point AMD's mobile CPUs are just not good enough(never were actually even when they had a clear desktop performance AND power advantage over Intel their mobile CPUs were pretty crappy).

Add into this equation that AMD bought ATI, and while their synthetic benchmarks look good, their game FPS look terrible, although I strongly suspect bad drivers from the performance disparity shown as nVidia cards don't do nearly as well in the synthetic benchmarks, but kill equivalent ATIs in price/game performance, although some of the 3rd party ATI purveyors seem to have some awesome stock coolers, e.g. IIRC the ASUS 4850 with their heatpipe heatsink/fan combo idles the 4850 c. 35C while running somewhere in the 50sC at load which is pretty damned good for a recent highish end card, too bad that they can't write drivers to really run their hw though... and we won't even talk about their mb chipsets...

Re:Oh please.

[Hatta]

I bought an E4500, expecting I could overclock it a bit. But I put the machine together, and honestly I don't know why I'd even bother. It's plenty fast enough for anything I do. At some point you reach diminishing returns, and honestly I don't care if a task finishes in 3 seconds instead of 2. Do you really see a difference between 2ghz and 3ghz for anything other than encoding video?

Re:Oh please.

[0xygen]

I have had a very similar experience - always being an AMD fan prior to this build (AthlonXP Barton was the last one) and now have a C2D E6750 which is getting a bit long in the tooth.

In my experience, I got fast DDR2 1066 RAM, lowered the FSB:CPU ratio, pushed the core voltage up to 1.4V and put the clock speed up to give about a 3.4 GHz clock speed, from the stock 2.6 GHz.
Everything still runs happily like this, even on the stock cooler.

I even can feel the difference in Far Cry 2 with all the physics set to Very High.

Re:Oh please.

[billcopc]

Err.... does ECC ram actually help with reliability, or does it merely ensure that errors get detected ?

Perhaps I'm a bit of a purist, but if ECC Ram is actually self-correcting, I would worry about how/why it got corrupted in the first place. I find it much cheaper and easier to buy good quality non-ECC Ram instead.

Re:Oh please.

[billcopc]

Do you really see a difference between 2ghz and 3ghz for anything other than encoding video?

Yes, it gives you more headroom for when your PC is bogged down with spyware and viruses!

Re:Oh please.

[billcopc]

You might be surprised at the performance leap from a new CPU. I went from an X2 4800 with an 8800GTS, to an overclocked C2Q. I noticed a tremendous improvement in almost all games, despite running on the same GPU. For one, it eliminated any and all stuttering, even in older games. I'd say it pumped a good 30% more fps into my main games like WoW, LoTR, and the shooters of course.

Today's graphics cards are so ridiculously fast, they're very commonly limited by the CPU. It's like the 3Dfx days all over again!

Re:Oh please.

[Lord Apathy]

I think you pretty much nailed it. A few years ago I bought a amd 3000+ for 595 bucks because I wanted bleeding edge. A few days later a friend bought a amd 2800+ for 250 bucks. Less than half what I paid.

My extra almost 400 bucks got me nothing that mattered. Sure I was faster then he was but he could play all the same games I could just as well. Sure the web, read email, view porn. Just as good as I could.

The only place you could tell the difference was when encoding video or audio. I was just a few seconds faster. Well worth the 400 bucks....not.

Re:Oh please.

[Ed Avis]

What happens is that during normal operation of any RAM there is a small chance that a particular bit will get flipped. Cosmic rays are often blamed as the culprit; of course if you overclock and overvolt your memory you increase the chance of errors, but even good quality RAM running within spec will get an incorrect bit every so often. If you use non-ECC memory there is no chance to spot this error; it just returns the wrong data. The old parity memory added one extra check bit for every eight bits, so most of the time it could detect (but not correct) a one-bit error. ECC stands for error correcting code (look it up on Wikipedia) meaning that if one individual bit is corrupted it can recover the correct data. If you are really unlucky and two bits in the same code word get corrupted at the same time then you still have problems, but that is unlikely.

If you are using non-ECC RAM then may be getting corrupted from time to time, but you don't notice.

Re:Oh please.

[default luser]

Perhaps I'm a bit of a purist, but if ECC Ram is actually self-correcting, I would worry about how/why it got corrupted in the first place. I find it much cheaper and easier to buy good quality non-ECC Ram instead.

ECC isn't just really meant to counter hardware memory errors. It's also meant to prevent single bit flip error events.

Cosmic rays and background radiation can cause single bit flip events. With increasing data densities, the likelyhood of a high-energy particle colliding with a memory cell increases. The other with increasing data densities is: the higher the density, the smaller the cell, and therefore the amount of energy required to flip a bit from 0 to 1 is reduced.

It's a similar problem to the RAID5 array issue (posted recently here on Slashdot), where the data densities are so high that you may enccounter a disk read error before your entire array rebuilds. Higher data densities are making once unheard-of errors possible.

And sometimes, you will immediately see the effects of a bit flip. Other times, the bit flip may remain undiagnosed for years, which can be more annoying (you may not have a backup from that long ago).

So, it's no surprise that ECC is becomming more popular, especially for workstations. As for myself, my current desktop doesn't have ECC, but I'm seriously considering the extra cost of ECC in my next computer.

Re:Oh please.

[Lord Apathy]

I doubt he does. Of the half a dozen or so systems that I have had in the last few years all of them but one has been AMD. Every one of the AMD worked fine until I retired it, but one. I killed that one. I was flashing the bios and accidentally put my big foot on the power strip. Yeah, it was stone dead.

All the amd mb but that one still work fine. I have one I pulled out of storage after 2 years and powered it up. Its running perfectly. Hell, even the clock was still within acceptable time for being on ice for 2 years. The only intel board I had dropped dead after about a year and a half.

Re:Oh please.

[level_headed_midwest]

I have had a similar experience with machines I have had and built/administered:

1. K6-2/500 on a VIA MVP4 chipset: no problems at all
2. Celeron 900 on an i810/ICH1 chipset: no problems at all
3. P4-M 2.2 on an i845MP/ICH3M chipset: integrated Intel PRO/100 NIC died
4. Duron 1600 on an NForce 2 Ultra 400 chipset: no problems at all
5. X2 4200+ on an NForce 4 SLi chip: no problems at all
6. Dual 2.8 Xeon Irwindale on an E7320/6300 chipset: integrated Intel IDE controller was recognized intermittently
7. Pentium D 820 on an i945G/ICH7 chipset: southbridge PCIe controller went AWOL, knocking out the integrated NIC
8. C2D U7500 on an i945GM/ICH7M chipset: no problems at all
9. Mobile Sempron 3600+ (65 nm) on an AMD M690T/SB600 chipset: no problems at all

The AMD units have been good to me, while the Intel ones had the problems, particularly with the southbridges.

Re:Oh please.

[hairyfeet]

Well I don't know about him,but I can tell you that in the 13+ years I've been working PC repair I've thrown away MAYBE three burnt Intel chips. Compare that with more blown AMD CPUs and Mobos than I have even bothered to count. Hell I'm sitting here right now looking at an AMD board that blew the caps. I personally think it is because so many manufacturers OC them at the factory and then save a buck by putting shitty heatsinks on them.Let me give an example of Intel reliability:

We get a call at CPU solutions to come out and do some work on a PC for a trucking company. They had all their invoices,truck schedules,pretty much everything they needed to keep working on this machine and with no backups the boss started to panic when it started slowing down and called us. The first thing we notice is that I can't lift this tower onto the desk to look at it. It simply will not move,even after Doug and I tried together. It turned out this thing had been sitting for the past four years and the cleaning crew had waxed it to the spot. So necessity being the mother of invention I ask them if they have those dollies for rolling under a truck. Of course they did so they next thing I know I'm rolled under the desk trying not to get distracted by the cute secretaries long legs while I work on this thing. I pop the side and you lierally couldn't see the mobo. They smoke and the funk had made this solid mass like a brick sitting in the case. Really REALLY nasty. The fans looked liked they probably haven't turned in months. Yet when I cleaned all the crap out(putting some better fans and a smoke filter on the back) that Intel P4 purred like a kitten. I ran some tests and it came back dead solid perfect.

Now compare that to an Athlon 64 that Doug was putting together to sell. The fan must have been shitty because it quit spinning and was off less than 3 minutes when that AMD blew. Fried the chip and the mobo. It joined the pile of dead AMD chips we had in a bucket. There was ONE Intel chip in the bucket and it got it from a lightning strike. So while AMD may be cheaper I'll stick with the Celeron when I need to build a "Granny box" and the cheap Pentium Ds when building a budget gamer rig. Because I know those boxes will last. I have gone back to places that are still running old P2s and P3s that I built in some seriously harsh work conditions,and yet they keep right on doing their job,day after day,year after year. For me that extra bit of reliability is worth the extra cash. And I have YET to have someone bring in an Intel box I built because the CPU dies. HDD die,RAM or GPU needs upgrading,yes. But those Intel chips just take a serious beating and keep coming back for more.

Re:Oh please.

[Bengie]

First off I acknowledge
"I'm not going to call someone an idiot for their CPU choice" and " I recommend consumers try to stay OFF the "bleeding edge" "

As for AMD vs Intel. I find that after the P4 crap fest, Intel has been leading AMD is less power consumption, less heat, better overclocking. For a non-overclocking grandma who doesn't care if their comp generates a few extra watts of heat, this is all good.

The new Core i7 with 4 cores and 8 virtual threads has been tieing dual quad core Xeons. That puts it exactly 2xs faster than it's previous generation. Not only that, but people have gotten the first batches of chips to overlock to ~4.4ghz air cooled and 5.4ghz fridge cooled and run reliably. It also seems that the Core i7 used 1/2 the power to do the same work and when idle, used about 1/10th the power(15.5 watts total at idle). Now if you have a dual socket core i7 and use it for your home system, those dual CPUs will only use ~30watts idle and sub 60watts when playing video games that don't use all the cpus. That's a lot of raw CPU power on-demand with almost no electricity cost.(power usuage was cpu usuage, not total system. The buss links also close channels/etc to reduce power draw when less system I/O is going on)

Better hope AMD made this chip good.

Re:Oh please.

[lawaetf1]

Hah! that's priceless about the power strip.

Everyone knows you're supposed to sit perfectly still, holding your breath, squeezing your sphincter while any BIOS update goes through. Anything less than that shows disrespect resulting in consequences like yours.

Re:Oh please.

[Kamots]

They're only CPU bound at low-res.

At any decent res, you'll be GPU bound, even with the latest and greatest graphics card.

Also, *which* 8800GTS are you on? It's a horribly overloaded model number where there are even different cores used. You can tell which you have based on the ram, a 320 or 640meg flavor is the older card, a 512meg flavor is newer.

I'm thinking that you did something other than swap out your processor. WoW shouldn't be having any issues even on a single-core machine with a graphics card earlier than the one you had.

Re:Oh please.

[afidel]

Most modern ECC is multiword and can correct two bit errors so it's considerably better than the old single bit ECC. Going up just a bit there's IBM's Chipkill technology that throws a RAID controller on top of ECC ram to enable an entire chip to be lost. Chipkill is available from multiple vendors, not just IBM.

Re:Oh please.

[michrech]

Along with daedae, I'd like to see any sort of proof on that also. Nearly every machine I have had over the last 6+ years has been an AMD of one form or another. In fact, the only two non-AMD machines I have are an MSI Wind (I wanted to tinker with an Atom based machine in Linux, OSX, and Windows, to see how it performed) and an C2D 2.4ghz machine I slapped together (well before I acquired my Wind PC) in order to play with OSX.

I can't recall one occasion in which I've had bug-ridden and/or chipset incompatibility issues.

Re:Oh please.

[Anonymous Coward]

If you are only interested in interactive tasks, as are many desktop tasks,
then you won't notice much difference.

For everything else, there is a huge difference. Compiling comes to mind,
but anything involving tight loops of cpu-bound calculations will show
a big improvement from 2GhZ to 3GhZ.

Re:Oh please.

[Anonymous Coward]

The Core2Duo has looked great, and the last system I built I pretty much went in planning to buy an Intel for the first time in years.

But as my price kept going way up from my target I realized that half the cost of the system was the CPU and the Intel motherboard. Both parts were considerably cheaper with AMD for the same performance in the benchmarks I care about (gaming).

I ended up going with a $48 AMD chip with the same performance of a $170 Intel chip. The barebones kit (case+mobo) was also about $50 cheaper for AMD than for Intel.

But yes, I was telling this to a friend a month or two ago and he pointed me to newegg - sure enough, Intel had dropped prices dramatically and actually had chips for under $100, matching or even beating AMD at price/performance.

I just built a comp for my wife and AMD was cheaper again. Although I actually got her an Intel, just because that's what the dirt cheap Dell came with for $400. If I bought it myself I'd never be able to make the same system for the same price (and certainly not with that quiet+nice of a case). On the downside, I'm now waiting for SATA->Molex adapaters so that the Dell SATA-only PSU can power the graphics card. Lame.

Re:Oh please.

[LWATCDR]

maybe but I am waiting until the 45nm AMD desktop chips are out to upgrade. If the offer a good price/perf ratio I will stick with AMD.
The thing is right now I don't really want to jump to Intel since I will not spend the money for an i7 at this time. Seems like I would be buying into old tech. My current motherboard is very nice and has served me well for five years.

Re:Oh please.

[CAIMLAS]

I'm not going to agree with you, because I've noticed this too - since about 1997, it's held true.

The exception I'd like to make is that in the last year or so, I've noticed the number/quality/price of Intel boards has become more favorable, while the inverse is true for AMD boards. You'll pay 50% more for an AMD board of similar quality to a decent Intel board - and with the 'price sweet spot' for a processor being only about $120-$150, paying $50 more for a decent board doesn't seem to balance things out.

Personally, I think it's a wash right now.

Re:Oh please.

[GiMP]

Many systems have a replaceable BIOS chip. You can buy a new one, even flashed for you, for reasonable prices from several online sources.

Since we're talking about AMD here, when the original Barcelona came out, I was one of the first to get them. Since there weren't yet any motherboards that supported the chip without being flashed, I simply outsourced it and had a flashed BIOS chip mailed.

Re:Oh please.

[ergean]

Get your head out of your... If would've said Athlon XP I would say yeah... that shit happens.
I have a heap of them around. But not Athlon 64 all the motherboards I know from socket 754 upwards shut down when the temperature of the cpu hits 90*C. So stop banging the Intel drum. I just hate the shit cpus they made on socket 478. I would recommend anyone Pentium D or Celeron D for the love of god, they suck compared to Athlon64.

I usually take the Athlon64 way for anything budget. If there are money to spend I go for the C2D + P43/45 + 4850/9600gt. And don't buy crap motherboards. I think the P45 is the new BX chipset. :)

Re:Oh please.

[squallbsr]

Hmm, that completely sounds fishy to me.

Case in point, I have a C2D Q6600 on a DFI Mobo with Intel X38 chipset. I have had the hardest time getting Windows (Vista) to run in IDE hard drive mode reliably, I cannot even get it to boot in ACPI mode. Linux works great with ACPI, but refuses to boot with IDE mode.

I have built over 20 AMD based machines and 3 Intel machines. All of which have run Windows and Linux (and some Solaris), the Intel setups are the ones that I put together, fiddled with until they started working and have been afraid to touch for fear of them stopping to work again. My AMD machines (with the exception of 1 screwed up motherboard) have all worked flawlessly for years and weren't that hard to setup and install.

Re:Oh please.

[Kagura]

They're only CPU bound at low-res.

At any decent res, you'll be GPU bound, even with the latest and greatest graphics card.

No, this is wrong. I did the same thing as GP. I upgraded my video card from a Radeon X1900 to a new Nvidia GTX 280. This is running at 1680x1050 (maybe that counts as low res in your book?) Then, a couple weeks later I upgraded from an Athlon 4200+ to a new Intel Quad Core Q9550. The difference in stuttering and frame rate in games was around the 30% mentioned above. Oblivion's frame rate was among the greatest improved out of all games from the CPU upgrade.

In short, parent doesn't know what he's talking about. Grandparent was the one who was right.

Re:Oh please.

[glsunder]

I'm not sure where you're getting your info. At my last job, we used AMD processors in desktops since the K6 days without issues. In fact, there were a few K6 systems in use less than 2 years ago. That's hundreds of desktops without issues.

As for home, I currently have 6 computers in my house, 2 are intel based, 4 are amd based. I haven't had any issues with any of them except some flakiness on the c2d based laptop (due to intel's flash cache thing).

Now, all of those systems have decent quality parts. If you buy a $30 mobo, you can expect issues. I've stuck to 2 brands over the last 10+ years and haven't had any issues.

Re:Oh please.

[hairyfeet]

I honestly couldn't tell you if it was an XP or a 64,because when the bossman roasts his fingertips and is cussing a blue streak the smart money stays the hell out of his way. And as I said in my post it could simply be that a lot of mobo and OEMs crank up the OC from the factory and then use shit fans. I know that when the Sempron first came out folks were bring us burnt Sempron machines left and right. And slightly OT,but whoever came up with the Sempron should be smacked. That has to be the worst POS CPU I've ever come across. Even the lowest Celeron seems to stomp it into the ground.

The problem I have recommending AMD to my customers is threefold. One,by studiously avoiding "shit motherboards" as you so succinctly called them,you end up eating any savings you would have had from the cheaper AMD CPU. Secondly many of the machines I build and sell are for work environments,some of which are quite hard on machines. So the AMD advantage in some video games and multimedia benchmarks are simply irrelevant in that market but reliability is. Finally my customers who are buying home machines really need them to last,and I have had much better luck with Intel over the long haul than with AMD.

So while I am glad you have had great success with AMD,in this market I simply can't afford to take the risk. And unlike the old days where it was several hundred dollars between an AMD and Intel CPU, nowadays for the "middle of the road" dual cores that my customers are quite happy with there just isn't a big enough difference in price for me to switch. Finally just like the old saying "nobody ever got fired for buying Microsoft" even my most clueless customers know who Intel is,and I have never had a problem moving a machine with an "Intel Inside" sticker on it.

Re:Oh please.

[Agripa]

I have to agree with Lonewolf666. I have been agonizing over going with AMD over Intel but the ECC issue is a deal breaker. It is only supported on Intel's more expensive and older motherboards. DDR3 in combination with ECC is not supported at all ruling out anything recent and an FB-DIMM solution would be more expensive yet.

All of AMD's recent processors have really nice support (note 1) for non-registered ECC DDR2 with the caveat that not all systems have support in BIOS for it. Gigibyte motherboards supporting embedded graphics (780G and 790G) do not while similar Asus ones do.

My current system is an old P4 2.4C and 875P chipset with 2 GBytes of ECC DDR and my system before that is an HX chipset with 1 GByte of ECC SDRAM that now does home server duty.

As for how necessary ECC is . . . there are a couple of papers available online that discuss measured soft error rates versus density and memory size. Below 1 GByte it is not much of a concern but with large arrays it can become a limiting factor in data and system reliability. My fuzzy recollection is that errors could be daily for an 8 GByte memory array. For a majority of systems running games and media applications there is of course no issue.

Note 1: AMD Barcelonas support a variety of ECC modes including scrubbing and chipkill while the Intel systems I considered only support correction on read.

Re:Oh please.

[KDR_11k]

Hell my 64 3000+ with a GF6800 plays World in Conflict fine, that's not really an archievement.

Re:Oh please.

[yoshi_mon]

Exactly and in the IT realm there often is a 'keeping up with the Jones' idea that being on the bleeding edge makes things oh so much better. The reality is far from that in my experience when you look at actual end use. However I don't discount the fact that there are those who will benefit from having the very latest. Just that it's not as much as the media of course wants to make us believe.

Case in point at one point I had to replace a desktop a little over a year go that was just used for the very basics. Web, nominal apps, and at most a bit of lite gaming. I found a local brick and mortar that actually had very competitive prices vs mail ordering a system so I went and checked them out. They seemed cool enough so I bought there. As I was checking off the components I told them the CPU I wanted, a Celeron D 3.06, and the sales guy said kinda of in passing, "Oh so your really going low end then."

I was a bit taken back as I had already politely withstood his grilling on what the system was to be used for when I already knew exactly what parts I wanted. So he knew that, at least I hoped, that such a CPU was going to be far and away enough power. Not a huge deal in the grand scheme of things but it did serve to tarnish my overall impression of the place.

Re:Oh please.

[danbert8]

I don't think that's entirely accurate. You can't just switch out the processor. You have to switch out the entire motherboard too, so you have a different memory controller, a different northbridge, a different sound chip, a different sata controller, etc... What makes you think that the processor was the magical factor out of all those that eliminated the problem.

Additionally, you probably also did a complete reinstall of your operating system when you upgraded the parts, which de-bloats Windows and you also downloaded the latest drivers for your graphics card. Those two items alone could have given you huge performance increases without changing any hardware at all.

Re:Oh please.

[Anonymous Coward]

WoW is more CPU bound than you'd think. The entire user interface runs on interpreted LUA (a scripting language) and the UI has a lot of stuff to do.

Re:Oh please.

[Kamots]

I'd suggest that you view: http://hardocp.com/article.html?art=MTU4MCwsLGhlbnRodXNpYXN0
Notice which resolutions changing out the CPU doesn't matter at...

And yes, 1680x1050 is borderline low-res. 1600x1200 which is what I game at is as well. high-res really starts at 1920x1200.

Also, as another poster has said, you swapped out way more than your proc. If I'm remembering right... didn't the 4200 run with DDR and not DDR2? Yes, if you're running on something that old, I'd imagine that yes, you could be CPU limited (or more likely memory bandwidth limited...)

However, with any modern system, for a decent resolution and AA setting, you're GPU limited.

Re:Oh please.

[Anonymous Coward]

I tried a few Asus boards and their support for ECC is now very limited (compared to their older motherboards). There's not even an option for reporting and logging errors, only one general ECC option. It is of course possible to set everything through the PCI interface of the processor, but I think this is kind of a PITA. That's a reason I now buy Gigabyte GA-MA770-DS3 boards for desktops. It doesn't have integrated graphics, but with Vista and Beryl, a real video card is not really a waste of money.

Re:Oh please.

[xrobertcmx]

I don't know, I run an Athlon X2 6000+ w/3870 and have no stuttering, no issues in any games I play. I even get to enjoy Forged Alliances at high settings with almost 0 slowdowns while streaming di.fm through VLC in the background.

Re:Oh please.

[puetzk]

Ordinary ECC will correct any single-bit error in a 64 bit word, and detect all 2 bit and most other errors. Chipkill ECC (same RAM, fancier memory controller) spread the codes across multiple words and is stronger still.

Average soft error rates for DRAM are about 1 bit flip/gigabyte/month (sea level up to about 10x that in Denver), mostly due to cosmic rays.

Re:Oh please.

[Anonymous Coward]

What's the difference in power consumption between 2.2 and 3.4 GHz at idle and under load?

Re:Oh please.

[Anonymous Coward]

Galactic cosmic rays are not responsible for most soft errors in a terrestrial environment. They can be viewed as an indirect cause when they collide with particles in the atmosphere generating thermal neutrons that can cause upsets in ICs on the Earth's surface. But these thermal neutrons can be generated by many particles other than GCR. Also, alpha particles are emitted when impurities in metallization and packaging decay. These alpha particles are ionizing and are a significant source of terrestrial single-event upsets.

Re:Oh please.

[QuasiEvil]

Couldn't agree more. I just finished building a quad core machine for home (my desktop was nearing three years old, and figured I might as well change it out in conjunction with upgrading to Ubuntu 8.10), and found AMD to be the more cost-effective choice. I have no ambitions of having the fastest proc out there (or the most overclockable, since I don't overclock) - I want the most cost effective processor under about $200.

The other thing to consider with AMD is it's not just the processor price, it's proc+mb price. Sure, a Q6600 Kentsfield produces slightly better benchmarks and is only slightly more expensive. That said, good Intel motherboards are pricey ($225+), whereas good, reliable AMD boards can be had for $120.

Bear in mind I'm not a gamer, I'm a programmer and a photographer. My main concerns are stability, reliability, performance/$, noise, and power consumption, in that order. I need a machine that can run a ton of processes without bogging, but the speed of any single thread isn't all that critical. So, better parallelism and greater I/O bandwidth is the name of the game for me - your requirements may differ.

Re:Oh please.

[billcopc]

Actually I run my games at 1920x1200, and it's the 320mb.

I'm still hitting the GPU's limits, but the averages are higher / smoother. On the X2, I used to get a fair bit of stuttering whenever something "happened" in-game, be it a big explosion or other game event.

I specifically remember Lost Planet, which chugged on my X2 anytime one of those huge bugs was on-screen, while the same scenes ran perfectly on my Quad without hesitation. I'm sure part of that improvement comes from the increased memory bandwidth, going from DDR1-400 to DDR2-1000, making texture swaps less noticeable.

I'm not disagreeing with you, my GPU didn't magically get faster, but you can't describe a gaming system just by identifying the edge cases. Most of the time you're somewhere in the middle, and that's where improvements are most noticeable.

Clock

[Anonymous Coward]

"dramatically increase" ... 2.7GHz still a bit far from the 5GHz of the Power6 from IBM

Re:Clock

[Anonymous Coward]

Comparing two CPUs based on clock speed alone is like comparing the speed of two cars by measuring only the RPMs of the tires. It won't get you anywhere ... you need to know the size of the tires as well!

Thus concludes my first /. car analogy. Thank you.

Re:Clock

[Anarke_Incarnate]

As another poster mentioned, frequency is not indicative of performance. You can run 9Ghz over a copper wire. That doesn't make it fast, and 2.4Ghz phones are not faster than 900mhz phones. You should also study I/O speed, instructions per clock, memory latency, cache coherency, failed branch prediction penalties, pipeline depth and penalty for misses, etc.

A CPU can run at much lower clock and still rip apart another CPU depending on application specifics.

...and so?

[tchernobog]

AMD says will result in 'dramatic performance and performance-per-watt gains.'

Okay, that's marketing talk. I think that at virtually *ANY* presentation of a new CPU in the last twenty years someone had said that.

Me, I just have a 6-yrs-old P4 laptop which, compared to nowadays new models w/ Core Duo, isn't much different.

This because there are other bottlenecks: hd speed, RAM, etc.

So, why upgrade, for a desktop user? Even for middle business servers, we live with two 8-yrs-old Sun machines which are more than adequate for keeping up all the services we need internally. We never have CPU spikes.

Sometimes I just wonder if all this isn't just a grab at customer pockets.

Re:...and so?

[robogun]

Me, I just have a 6-yrs-old P4 laptop which, compared to nowadays new models w/ Core Duo, isn't much different.

I once had a P4 laptop, replaced with a Centrino. I definitely do not miss the P4's howling fan & poor battery life, the Centrino (half of a Core Duo) runs much quieter and cooler.

Re:...and so?

[Pulzar]

Me, I just have a 6-yrs-old P4 laptop which, compared to nowadays new models w/ Core Duo, isn't much different.
This because there are other bottlenecks: hd speed, RAM, etc.

A 6-year old P4 laptop is very different from a new one, in just about every way except for maybe hard-drives which only got marginally faster. The CPUs use less power, video cards (and even chipsets) have hardware video decoding for many hours of DVD watching on one battery, the RAM is tremendously faster than 6 years ago, screens are brighter... and, well, CPUs themselves are a lot more powerful allowing you to do many workstation-type tasks on a laptop.

So, sure, if you only use your laptop to check email, surf the web, and write documents, and you don't care for long battery life, a 6 year old laptop is fine. It's not a grab at consumer's pockets if you don't need it, as there's plenty of us who do need more out of our computers. Mind you, this article is about a new *server* chip, so...:

Even for middle business servers, we live with two 8-yrs-old Sun machines which are more than adequate for keeping up all the services we need internally. We never have CPU spikes.

Not every server is a "middle business" server. Large-scale database servers, high-load web servers, simulation machines, rendering, research, video editing... all of these need (1) as much computational power as possible, and (2) for as little electric power as possible.

Re:...and so?

[thedonger]

Sometimes I just wonder if all this isn't just a grab at customer pockets.

Never underestimate a person's desire for the shiniest, newest thing.

Like you, I am happy - for now - with my Athlon XP 1500+ CPU paired with an Nvidia 128MB eGForce Ti video card. Old school, yet it handled nearly flawlessly all but the last level in Portal. Most people don't need the power of the last two years worth of hardware improvements. But remove price as a factor (i.e., you have more than enough money) or add status, and why the hell not get something bad-ass?

I wonder what would happen to the overall CPU picture is suddenly everyone stopped biting at the bleeding edge tech and waited for the inevitable price drop? Perhaps Intel and AMD begging the govt for a bailout? That would be funny.

Re:...and so?

[pinkocommie]

Don't know about you but try playing back a 1080p H.264 video file and watch it choke to death and then some.

Re:...and so?

[Archangel Michael]

Ha

But there's your error .... I need six cores just to keep up the the crap I'm running

One Core for Crapware
One Core for Anti Crapware
One Core for Virii
One Core for Antivirus
One Core for Applications

and

One Core to rule them all and bind them!

Re:...and so?

[JAlexoi]

It's not aimed at you, who is not planning new purchase. But at people that ARE planning to purchase new machines.
When I bought my last system, my swing vote was overall energy efficiency - AMD beats Intel, with a big baseball bat in that area.

Re:...and so?

[marcosdumay]

"Okay, that's marketing talk. I think that at virtually *ANY* presentation of a new CPU in the last twenty years someone had said that."

Decreasing the feature size has always leaded to big improvements on top performance, performance/watt and price on the past. I'm not saying that this trend will continue on the future, but it still hods for 45nm.

"Even for middle business servers, we live with two 8-yrs-old Sun machines which are more than adequate for keeping up all the services we need internally."

I see... You use neither the MS plataform nor the newest enterprizy technology at your company. I wasn't expecting you to understand such things on those circunstances.

Re:...and so?

[SeeManRun]

AMD says will result in 'dramatic performance and performance-per-watt gains.'

Okay, that's marketing talk. I think that at virtually *ANY* presentation of a new CPU in the last twenty years someone had said that.

Me, I just have a 6-yrs-old P4 laptop which, compared to nowadays new models w/ Core Duo, isn't much different.

This because there are other bottlenecks: hd speed, RAM, etc.

So, why upgrade, for a desktop user? Even for middle business servers, we live with two 8-yrs-old Sun machines which are more than adequate for keeping up all the services we need internally. We never have CPU spikes.

Sometimes I just wonder if all this isn't just a grab at customer pockets.

You are wrong. Everything else being equal, a CPU upgrade won't do much for you. However, that is rarely what happens. If you have a 3 year old computer and compare it to a new computer, the speed improvements will be vast in every single area of usage, except maybe Internet download speeds. If it doesn't feel faster, then put the same software on it and it sure will be.

Re:...and so?

[initdeep]

get a real video card that does the decoding on the GPU and the clock speed of the cpu matters very little.

Re:...and so?

[Anonymous Coward]

How did you get here, you little ungrateful clueless brat?

Get the fuck off my lawn!

(And oh, the reasons why: "128mb GeForce Ti card" - I know 5 different GeForce Ti cards off the top of my head, and video memory has nothing to do with performance. Also, "it can run every level in Portal" - when we're talking about server processors. Get a clue, grow up, get some real hardware, and then return.)

Re:...and so?

[savanik]

So, why upgrade, for a desktop user? Even for middle business servers, we live with two 8-yrs-old Sun machines which are more than adequate for keeping up all the services we need internally. We never have CPU spikes.

I actually ran into a very good reason to upgrade recently - hardware cycles. Computers last about five years, not due to any particular deficiency in the hardware, but just because eventually the things break. If information is important to your company, you need to be sure that if your computers break, the parts to repair them exist.

Case in point, I have a friend who's a networking consultant, and at one point he had to pull a business's data off of an old XT machine with an MFM hard drive. Yes. Those. No ethernet port. 5 1/4" floppy, which nothing reads these days. No MFM controllers for modern machines. No expansion ports to put in something useful. He ended up finally using the serial port and Xmodem (it didn't have enough memory for Y or Zmodem) to pull the data off.

So for business, the question for upgrading isn't so much about having the latest and greatest as it is about business continuity. For the home user, it's more about running the latest and greatest game. If you don't fall into either of those categories, you may be perfectly happy with an old Pentium 5.

Me, I <3 Fallout 3.

Re:...and so?

[thedonger]

I believe I was responding to a commenter who was talking about a P4 laptop. Not server-related. You are an idiot.

Re:...and so?

[BLKMGK]

So you're advocating Windows? I'd love to do H.264 decoding on the GPU in Linux, which driver and which video card do that for me?

I use XBMC, we're stuck with using the CPU for now but at least we can use both cores. So far the AMD CPUs haven't fared well with that software for full 1080P H.264 decoding either.

Re:...and so?

[des09]

Get off MY lawn, coward.

Re:...and so?

[ShadowRangerRIT]

While AMD uses the least energy, a number of articles I've read indicate (e.g. this Tom's Hardware article shows a win for Intel about a year and a half ago, even before Intel's move to 45 nm) that if you factor in performance to get perf-per-watt, the Intel and AMD lines are much more similar. Granted, when neither chip is doing anything, AMD wins, but when placed under load Intel's slightly greater power usage seems to be offset by doing more actual *work*. I run Folding@Home to keep my machine busy doing something useful, alongside some video encoding, so for me, and anyone else with fairly high CPU loads, Intel is probably a better deal.

Re:...and so?

[Anonymous Coward]

meh,

we are talking about server processors here no?

Re:...and so?

[Belial6]

Sometimes it's cheaper to buy the new stuff than not to. I bought a new Athlon X2 5400 a little over a year ago to replace my Athlon 1.2. I plugged the whole system into a Kill-A-Watt to see the power draw. I calculated that it would take 10 months for the energy savings to completely pay for the upgrade. So, my faster computer is now not only free, but actually saving me money. Because of this, I have also upgraded my wife's computer, my kid's computer, and my server. My wife's computer and my server should be breaking even over the next month or so, and my kid's computer will have payed itself off in June.

I also downgraded my speakers. My old Klipsch 5.1 surrounds sound speakers sounded great, but they drew something like 45 watts. I replaced them with a generic set of 2.1 speakers that don't sound as good, but are more than adequate for the purpose, and I am now only drawing 2 watts.

Re:...and so?

[Chandon Seldon]

I wonder what would happen to the overall CPU picture is suddenly everyone stopped biting at the bleeding edge tech and waited for the inevitable price drop?

There would simply be no price drop. The price drop comes from the company having made their money on that generation and released the new generation. If both Intel and AMD were to drastically slow down selling high end processors, they'd stop building new fabs - and run their current-generation fabs until they had made their money on them.

This would be really bad for everyone, because we'd miss out on the benifits of progress in computer hardware technology. Sure, you'll claim to only run applications that would basically run fine on a Commodore 64, but that's probably not even true - since nearly every website you visit has reasonably high end hardware backing it. And that's completely ignoring things like bioinformatics, where fast computers are quietly but drastically transforming human life for the better.

Old school, yet it handled nearly flawlessly all but the last level in Portal.

I suggest upgrading. I know you're putting a lot of mental effort into convincing yourself that game graphics quality it's irrelevant so you can avoid getting a new computer, but there's like 5 years of good games that you're missing out on - or effectively missing out on - with that rig. It's like watching an action movie on a 12" TV - it's possible, but it ruins the movie compared to watching it with some immersion.

Oh, and if you upgrade, humanity will thank you for supporting technological progress. You never know - your money could make a fab break even just enough sooner to build a new fab a month earlier, which could give a biotech lab one generation better processors, which could enable the breakthrough that creates a medical treatment that saves your life one day...

Re:...and so?

[johnny0099]

You're in there:

http://www.nvidia.com/page/purevideo.html

Re:...and so?

[hairyfeet]

While that is true one mustn't forget What Intel and AMD giveth,MSFT taketh away which is why I have had so many customers saying "Get this Vista shit off my PC!" because they brought home their shiny new toy only to find it getting its ass handed to it by their P4 running XP. So if you got the money to go bleeding edge you might notice the difference,but at least for my customers the Vista resource black hole ate all the gains and then some. That is why I'm hoping Win7 doesn't suck,because frankly I'm tired of playing "find a driver" for laptops that are Vista only.

Re:...and so?

[Anonymous Coward]

OTOH, if hardware development slowed, software developers would be forced make version 2.0 faster by writing good code, rather than by upping the system requirements on even-more-bloated-than-ever-crap. I'd take that over your life-saving medical hardware.

Re:...and so?

[BLKMGK]

From their download link ->

"NVIDIA® PureVideoâ Decoder software, a plug-in for Microsoft® Windows® Media Player and Media Center Edition, delivers smooth MPEG2 video and unmatched color fidelity when watching DVDs, videos, and TV programs on your PC. It features Dolby® Digital surround sound audio and supports the MPEG2 video features on your Microsoft DirectX® compatible graphics processor."

I'm sorry, I do not see Linux listed there anywhere, do you? Yes, my HARDWARE is H.264 capable but unfortunately no manufacturer to include ATI, Intel, or NVIDIA, has released a DRIVER for Linux that enables it. What I really really want is feature parity between the Windows drivers and the drivers for Linux. They own the code right? They built the hardware right? So why do I only get MPEG2 acceleration? Yes, I've read that they cannot open source that code because of licenses (so much for owning) and I've heard them claim other things too. Where's the beef? License it for Linux already. Do it as a binary blob closed source like the existing NVIDIA drivers and document it such that it's easily handled for the programmers. That's what I want, I'm not holding my breath. Yes there are efforts to write this support using bits already exposed but WOW would it be better if the manufacturers would simply do it in the damned drivers! Can I have audio over HDMI while we're at it? Apples and oranges maybe? I still want it...

And no apparently CoreAVC isn't an option for the XBMC guys, they do not want to code for a pay for play library.

Re:...and so?

[Anonymous Coward]

Already here with amdcccle (AMD Catalyst Control Center Linux Edition) release 8.12 and later with Radeon HD3xxx and HD4xxx GPUs (UVD and UVD2). Catalyst 8.13, to be released in December, 2008, will do the H.264 encode using the GPU using a Radeon HD46xx or HD48xx card. I bought the ASUS HD4670.

Re:...and so?

[Anonymous Coward]

1080p can be decoded fine in software on any cpu > 1.8ghz dual core, as long as ffmpeg was compiled with smp support.

Re:...and so?

[compro01]

and the GGGGP is talking about his 6-year old P4 laptop, which would definitely not be dual core and likely clock somewhere in the 2ghz range, which I know for a fact will not do 1080p anything in a nice manner.

Re:...and so?

[Chandon Seldon]

OTOH, if hardware development slowed, software developers would be forced make version 2.0 faster by writing good code, rather than by upping the system requirements on even-more-bloated-than-ever-crap.

Not only do you want to die of some potentially treatable genetic defect, you want your software to be crappier and take longer to write because the programmers had to spend all their time obfuscating previously elegant and readable code so the machine liked it better or, alternatively, simply stripping out features?

Screw that. I'll take maintainable code running on fast computers.

Re:...and so?

[toddestan]

Really? I've tried the "it will pay for itself in energy costs" calculations before, and I've found it just doesn't work out. According to the kill-a-watt, my computer (the box only) currently draws about 150-180W of power depending on the load (it's an old-school Athlon with dual video cards and 3 harddrives). That means it costs me roughly 1.5 cents per hour to run. That sounds bad, but that's about 36 cents a day, or about $132 a year to run, assuming I run it 24/7 (which I don't). So even if I was to upgrade, it would probably be about a year to pay off if I went with the absolute cheapest stuff I could find, to well over a decade for something more like what I would want without being excessive. So electricity must be really expensive where you are for something like that to pay off that quickly. Generally I've found that it's almost always cheaper just to keep what you have so long as it's still useful and keep using it instead of upgrading to less power-hungry components with the goal to save money.

Re:...and so?

[Anonymous Coward]

Ah, the dual joy and pain of living in the country where it's far too cold far too often.
Basically, every Watt I save in one place has to be replaced with one Watt of (electric) heating - the end result is that from September to May I can leave my computers on 24/7 and it costs me no more or less than turning them off.

The downside is that this makes upgrading less tempting.

Re:...and so?

[afidel]

Yeah but if you throw in ram and the northbridge the Intel solution loses bigtime, especially in servers where the number of DIMM's goes up significantly.

Re:...and so?

[johnny0099]

Yeah, sorry, just mpeg 2 for linux. Bummer. Now I'm mildy pissed.

Re:...and so?

[AncientPC]

I'd have to second toddestan. Assuming you went from an idle ~300W hog to a ~40W machine running 24/7/365 at $0.10/KWh you'd only save $222.76/yr.

And to be perfectly honest, I've had a hard time hitting ~300W. That only happens when I play games or run Folding@Home using my GPU (ATI x1950). Even my home file server with 8 hdds and an Athlon XP idles around 120W.

Re:...and so?

[juhaz]

With a little luck, all of them will, pretty soon.

NVidia just released their HW accelerated video decoding API
http://www.phoronix.com/scan.php?page=article&item=nvidia_180_vdpau&num=1
http://www.nvnews.net/vbulletin/showthread.php?t=123091

AMD pretty close to doing the same with their alternative
http://www.phoronix.com/scan.php?page=article&item=amd_xvmc_xvba&num=1

And Intel is backing VAAPI as well as apparently considering extending XvMC to support more recent video codecs as a stop-gap measure.

It's rather unfortunate that they're all separate and mutually incompatible (and presumably proprietary, for the first two), which will no doubt hinder adoption by application developers, but it's a lot more than what we had until now.

Re:...and so?

[Belial6]

Here in California, we start at 12 cents a kwh and on my bills have gotten up 32 cents a kwh on the tail end of my bills. PG&E ramps up the cost of electricity the more you use. This means that energy savings always come from the most expensive part of the bill. The upgrades I made of new motherboard/processor/memory cost me ~$160 and dramatically sped up my machines. So, to pay for the machines I needed to reduce my energy usage by $160/$.32 = 500 kwh. My systems went from ~140watts idle to ~60 watts idle and ~180 watts under load to ~100 watts under load. So, we can say that they save ~80 watts given that few operations really push these new processors. My server runs 24/7 so that would be only 260 days to pay for itself. My wife's old machine would not properly suspend, so it ran basically 24/7 without load. That means that for 4 hours a day it reduces the watts by 80 and for 20 hours a day it reduces it by ~100. That means it pays for itself in a little less than 260 days. My system is under much heavier load and finishes these tasks much faster and can then move back down to a non-load state. While it runs on average ~16 hours a day, it has higher savings because we are reducing 180 watts down to 60. The least savings comes from my son's computer. That one might take a couple of years to pay itself off, but it is likely that he will not get another upgrade within 3 years so it will also pay for itself while giving him a better user experience.

It is great that there are places in the US that get their electricity at less than a third of what we here in California pay, but for us, it often makes sense to upgrade from a strictly cost point of view.

Re:...and so?

[johnny0099]

And then out of no where...

http://www.phoronix.com/scan.php?page=article&item=nvidia_180_vdpau&num=1

crazy

Re:...and so?

[BLKMGK]

Very nice and good info - thanks! It's a shame it's not a single easy way to access acceleration but I guess we'll see if the XBMC or other developers use it. The XBMC guys seem to like using curtting edge ffmpeg code so my fingers are crossed for NVIDIA at least! :-)

Re:...and so?

[BLKMGK]

Progress - maybe. If the XBMC developers pick up on the ffmpeg patches and use them I will be SO very very happy! They grabbed the CABAC stuff back when it was being REJECTED by the ffmpeg guys so fingers crossed they like this too :-)

Please tell me AMD is not betting it all on SIMD.

[Ostracus]

"If AMD is betting the company on an improved production process..."

They're not. There's a reason they bought ATI and it's not just because they want to get into the graphics business.

What about motherboard integration?

[VorlonFog]

So, what southbridge chipset are these going to use, now that nVidia has completely lost any credibility for supporting AMD processors? (note the HP and Dell laptops failing all year and the numerous Register stories about faulty self-destructing GPUs.

Re:What about motherboard integration?

[LWATCDR]

How about AMD/ATI?
They have bee producing some very good chipsets for the desktop.

great hallmark

[corsendonk]

yea to be honest, I don't think most people associate 'Shanghai' with something nifty (altough it is one of the most thriving city in China).

Re:great hallmark

[rickb928]

Thriving on *what* is the question that comes to mind for me... Thriving on manufacturing nifty cell phones, cool. Thriving on hosting pr0n and v1@35a sites, not so much. Thriving on midnight hotel-room kidney transplants, eewww! Make them stop!

Shanghai could the the melamine capital of China for all we know.

Re:great hallmark

[Malc]

Thriving on life. I've just lived there for four months and wish I could go back.

Re:great hallmark

[rickb928]

Probably beats liking in Detroit...

Re:great hallmark

[wild_berry]

Barcelona, Budapest (not released), Shanghai (HT3.0, 45nm), Istanbul (6-core response to Dunnington), Montreal (missing in action), Sao Paolo (6-core, DDR3, due 2010), Magny-Cours (12-core, DDR3, due 2010). All Formula One race circuits.

Hooray!

[Hassman]

Yay! AMD leading the world into the ... past?

Making Me Feel Old

[withoutfeathers]

The first computer I ever worked on (as a data entry operator in the mid '70s) was an IBM S/360 mainframe with 64KB of "main" (physical) memory.

The first computer that I was a primary operator on, a S/360-135 plug-compatible 2Pi, had 768KB when it was delivered and was eventually bumped to 1.25MB shortly before I moved on to programming.

The computer upon which I wrote my first professional (COBOL) program was an IBM 3033 with a (for then) eye-popping 4MB of physical memory.

The first computer I ever owned was an RCA COSMAC with 4KB of memory.

The first DIY computer I ever assembled completely from parts (about 15 years ago) had 4MB of interleaved DRAM and a 256KB SRAM cache and was considered somewhat amazing by everyone who saw how fast it ran OS/2. I eventually boosted it up to 16MB

Now you get 8MB of on die cache with your four cores... And I still can't get a decent flying car.

Re:Making Me Feel Old

[girlintraining]

Well, speaking to both the feeling old part and the flying car part... Gravity is a bitch, ain't it? :)

Re:Making Me Feel Old

[valnar]

You're absolutely right. The flying cars today do suck.

Re:Making Me Feel Old

[MadKeithV]

No it's gravity that sucks.

Re:Making Me Feel Old

[Anonymous Coward]

"You're absolutely right. The flying cars today do suck."

have you ever wondered why hard drives are mirror polished? is it because a mirror finish makes for less possibility of a head crash, or are there tiny people with magnetic boots running around putting big 1 or 0 signs as a giant eye looks for the 0 and 1 signs? does the little guy down there have a pen and paper to write down the last 30 signs they held up? so that data can be recovered even if it's been formatted?

Disclaimer: I am paranoid schizophrenic.

Re:Making Me Feel Old

[slapout]

The way most people drive, I don't think flying cars are such a hot idea anymore :-)

Re:Making Me Feel Old

[Anonymous Coward]

Actually, it pulls.

Re:Making Me Feel Old

[pablodiazgutierrez]

If only we had a sort of Moore's law for gravity... Or at least batteries, for crying out loud!

Re:Making Me Feel Old

[Anonymous Coward]

I'm General Lee, you insensitive clod!

Re:Making Me Feel Old

[magus_melchior]

Agreed. These buckets of bolts only fly for a few seconds, tops, and only if you take a speed bump at 80 mph.

Re:Please tell me AMD is not betting it all on shr

[marcosdumay]

If by "more of the same" you mean better, cheaper and less power hungry CPUs, well, I guess that could work for AMD.

About Time

[ShakaUVM]

It's about time... I mean, seriously. The CPUs coming out of AMD have stagnated in the last few years. The Phenoms are decent enough, I guess, if you have apps that can take advantage of the three or four cores, but they clock at slower than comparable X2s, and two cores is still the optimal point on the diminishing returns curve (on adding more cores).

I remember the 90s and early 00s when you were basically required to upgrade your processor every year or two or be hopelessly behind when the latest game came out. Now, I'm running the same machine I was back in '04, except with a new video card and an upgrade from a 3800+ (2.4Ghz) to a 4800+X2 (2.6Ghz) a year and a half ago.

I got curious how far I was behind these days, and found that as far as everything goes, a 4800X2 is still about as good a chip as anything AMD produces, only about 30% below the top chips AMD makes right now.

By contrast, Intel has the E8500 which is not only significantly faster, but is heavily, heavily OCable as well. I think Moore's Law has finally broken down for AMD.

Re:About Time

[powerlord]

I remember the 90s and early 00s when you were basically required to upgrade your processor every year or two or be hopelessly behind when the latest game came out. Now, I'm running the same machine I was back in '04, except with a new video card and an upgrade from a 3800+ (2.4Ghz) to a 4800+X2 (2.6Ghz) a year and a half ago.

Yeah ... around 5 years ago I got tired to the treadmill and switched to console gaming, with the occasional light puzzle game on the PC.

Suddenly my PC is now lasting 3 years without needing any parts upgraded (although the power supply needed to be replaced, but that was just poor QA by the manufacturer). I replaced my console 2 years ago (got a PS3 in Dec '06), and it looks set to last for another 4-5 at least.

On the whole I think console gaming seems to be a much cheaper way to go.

I haven't had to worry about:
- compatibility specs
- upgrading hardware
- one game's install breaking another

and best of all I don't have to sit down at a desk in front of a computer (something I do all day at work)

Re:About Time

[ShakaUVM]

Yeah, I know what you mean. I was seized by a sudden crisis when I picked up Fallout 3 at the Gamestop. PC version or PS3? On one hand, the PS3 version runs better, and plays on a 46" LCD television. The PC version crashes occasionally (as do all Bethesda games -- they're allergic to making quality software), but... the PC has a mouse. While the VATS system kind of cheats for you on the whole aiming issue, playing something that requires aim (aka COD4) is still much more enjoyable on the PC than a console.

Re:About Time

[Ecuador]

Think of it this way, the current AMD's are not much behind in similar clock speeds, but are a bit behind at top clock speeds (well, ok, more than a bit if you OC). Even if I buy C2D when I want fast workstations, I would definitely say AMD is withing the same performance "generation" and they are very competitive at the mid-low segment.
So, your perception that Intel is improving very fast is based on the fact that they were running on a laughable (PR dept driven) architecture for years. From P3 -> P4 they went BACK in performance (not just IPC, the first P4's were actually slower than existing P3's).
What I also find utterly ridiculous is how most (fortunately not all) of the publications etc of the era found the P4 as a great architecture and some even compared it favorably to the Athlon. I was pulling my hair as I was seeing our daily tasks (e.g. building a symbian app on a VM running over linux) running over 5 TIMES FASTER on a 64bit AMD machine compared to a (more expensive) P4 that was its "competitor" (and "faster" in several 32bit windows "canned" benchmarks). Intel eventually adopted AMD64, went back to the P3 arch to evolve it into Core and now everyone and their mom admit P4 is crap, while AMD with a not much changed architecture since the first Athlon 64 so many years ago, has indeed fallen behind, but not by that much.

Re:About Time

[powerlord]

VATS was actually what sold me. I'm more interested in the exploration and story and am more of a turn-based game player vs. a "twitch" gamer, and have a hard-time aiming for headshots all the time.

I've found VATS to be a nice compromise. (of course I also like that its use is optional :) )

On the other hand, besides the PC game being buggy, it is also more "open" in that it can be modded.

This was a factor but since my desktop (can't call it a gaming PC anymore its too behind the curve), can't support Fallout 3, it was no choice. :shrug:

I think not sitting in front of the computer has actually increased my gaming time since it feel less like "work" :)

Re:About Time

[Anonymous Coward]

"...except with a new video card..."

Back then, CPU was important for gaming. Now, it's the GPU. Newer games like Crysis need powerful GPUs to run, while powerful CPUs aren't as important for most games (MOST - there are exceptions like Supreme Commander).

Looking at games like Crysis, I'd still say there are performance hogs that need upgrading to run well. It's just that the focus is now more on GPUs than CPUs.

Re:About Time

[ShakaUVM]

Eh, I'm basing this on actual benchmarks. (http://www.cpubenchmark.net/common_cpus.html)

As far as dual-cores go, the 3.16GHz E8500 is king, and isn't that expensive (I think ~$180), whereas a 3.1Ghz 6000+X2 is $90, but the 8500 is twice as fast and is exceptionally overclockable. So price/performance-wise, if I had to buy a CPU today, it'd be an Intel, and this is coming from a person who has never bought an Intel (except for the heavily OCable original Celeron line back in the day). Quite simply, Intel has won this round of the CPU wars.

I did buy a Phenom 9500 for the last machine I put together, since I did a lot of MATLAB coding on it, and MATLAB autoparallelizes its matrix math, but the gaming performance on it was pretty meh.

Re:About Time

[Jesus_666]

On the other hand, besides the PC game being buggy, it is also more "open" in that it can be modded.

In theory. I think they still don't want to release the toolkit to the general public.

Re:About Time

[ShakaUVM]

>>On the other hand, besides the PC game being buggy, it is also more "open" in that it can be modded.

Yeah, in theory a mod or two might make it to the PS3 as for-pay DLC, but modding was the only salvation for the steaming-pile-of-crap game mechanics that was Oblivion, but mod issues also caused a huge number of crashes in Oblivion. So again, I kind of dithered on the purchase on that front. Do I want to be able to fix any bread-dead decisions Bethesda made? Or do I want a game that doesn't crash?

In the end, I was very pleasantly surprised that they actually did the game mechanics near-perfectly in Fallout 3, so the lack of mod tools hasn't been an issue yet.

Re:About Time

[Anonymous Coward]

Also, what a lot of people don't understand (or even look at) is how many instructions per clock cycle the chips do.

It used to be that this is why AMD's ran hot, they executed more instructions per cycle, whereas Intel didn't. So, the AMD's could run at a slower clock cycle than an Intel and still get the same amount done.

I still try to use AMD whenever possible. AMD & ATI!

Re:About Time

[Ecuador]

Eh, you missed most of the point of my post.
And no, I do not base anything on canned benchmarks. Neither should anyone. I take the exact code that will run daily on the intended machine and I time it. The perl & C code that we run at work (on Suse 64bit) is about 10% faster at the same clock speed on AMD than on current intel. Thus, the latest workstation I bought was an intel E8600 running at 3.8GHz, since I would not be able to get an AMD to 3.4GHz (on stock air cooling) that would provide similar performance (for our workload). A typical run of our pipeline takes 10h+ so I can probably save a couple of hours over the fastest AMD. However, the point of my post was that AMD processors are still relevant, they are not a generation behind, they can't compete only in the upper segment (that brings back some pre-2000 memories). You can say Intel has "won the CPU wars", but AMD is not trailing by much and it is a good choise for cheaper PC's.
Now, I gave the examle of compiling a project with Visual Studio on a VM over linux. The particular project takes less than 10 seconds on an Athlon 64 @ 2.2 GHz, almost the same time on a C2D @ 2.66GHz and about A MINUTE on a P4 @ 3GHz (bought about the same time as the Athlon 64). Most of the difference I guess comes from 64bit vs 32bit host, however the fact is that Intel was a whole generation behind which made the Core seem an even bigger leap forward.

Re:About Time

[ShakaUVM]

Sure, but what relevance does the P4 discussion have to the chips today?

Besides, you can't see much difference in performance when you have a 10s compile time.

When looking at benchmarks in games (which is the primary thing that interests me besides MATLAB performance) the Intels again blow the AMDs out of the water. As you say, AMD can't compete in the upper segment. However, whereas "upper segment" processors used to mean you were paying $600 to $1000 for a chip, the E8500 is $180 or so, which is consumer-priced, in my book.

Re:About Time

[Anonymous Coward]

Perhaps with English not being my native language, our communication breaks. I was referring to your comment that "AMD have stagnated in the last few years" etc, implying that Intel has not stagnated. My point is that while AMD is losing the mid-high segment (mid being the $150-$300 now I guess), they have slowly been advancing all these years with the botched launch of the phenom (bug, lower clocks than expected) being an exception. Your perception of Intel advancing fast in performance is exactly because they not only "stagnated" themselves but even went backwards in design by adopting the P4 (and then going back to P3/Pentium M). After they launched C2D they went years ahead of their P4 (which was even sold concurrently), and even leapfrogged AMD. However for per-core performance don't expect Intel to manage other P4->Core jumps, unless they go again down a stupid path and have to correct once more after some years.

Jackie Chan

[glock22ownr]

Is Jackie Chan or Owen Wilson somehow involved?

Alice must be pleased...

[Suzuran]

Will there be a Hourai as well? ^_^

Re:Alice must be pleased...

[Muffinmasher]

(Score: -1, Forced Reference)

Re:Please tell me AMD is not betting it all on shr

[chrb]

Of course it worked for Intel. Higher resolution lithography processes mean you can fit multiple cores in the same space as a single core from a decade ago. It means that the latency for critical paths is reduced, which means you can run the chips at a higher clockspeed. It means current consumed by transistor switching is reduced, so that chips can run at a lower power whilst maintaining or increasing throughput (thought interestingly leakage current increases as feature size shrinks).

Manufacturing process improvements are the number one driver of processor advances. It is obviously true that processor architecture changes, but mostly this is a response to new developments allowing more circuitry to fit in the same space. The latest Core processors have basically the same pipeline design as the original Pentium Pro. If you could go back and re-design the PP using our "new" architectural advances but older technology process, you would end up with a pretty similar design, since the process itself imposes such huge constraints on the architecture.

This bodes well for the company

[default luser]

Just an off-the-cuff calculation on my part shows power consumption dropped over %50 over Barcelona, clock-for-clock.

This is good news, because when AMD moved from 90nm to 65nm, their leakage was so bad that the power consumption only dropped around %10 clock-for-clock. Combine this with better cache architecture (larger, and faster), and AMD may have a winner in the server space.

I'm not sure if they're going to take back the desktop anytime soon. Intel doesn't have the FBDIMM downside on desktop systems, and I'm fairly sure that Shanghai didn't add major microarchitecure changes, so a quad-core Core2, let alone an i7, should continue to dominate the desktop.

However, it is nice to know that the market once again will have a choice in processors. AMD's 65nm offerings were spanked in terms of performance and power consumption by Intel's lineup, but Shanghai will at least compete on the power front, if not the performance front. We shall see what happens when AMD releases their desktop version.

Re:This bodes well for the company

[Anonymous Coward]

there is no competition. AMD ruleeeeeeeezzzz!!!!

Re:This bodes well for the company

[Penguinoflight]

Power consumption is actually one of the areas where intel has been soundly beat, year after year.
Even 65nm processors from AMD use less power than Intel's 45nm procs, and Intel doesn't have an on-chip memory controller.

Add in the extra power consumption of an Intel northbridge, and intel's offerings are usually about double the power consumption of a similarly clocked AMD system.

AMD's real problems are in acheiving high clock speeds, and solving their fabrication process. If AMD's 45nm process is as improved as they say it is, and with their fabrication/design company split they should be able to get that side of their business under control.

Re:This bodes well for the company

[squallbsr]

AMD has also been doing much better in Database performance with regard to MySQL and a very specific workload. Intel Core 2 Quad was choking on the data where AMD would just hum along without issue. This is primarily an email data storage and email access back-end using MySQL. The MySQL engineers were even involved in this research.

Re:This bodes well for the company

[default luser]

I would disagree that Intel's power consumption has been "higher." In terms of performance per watt on the desktop, Intel has been impressively competitive.

For example, check out this review of Intel's 45nm Core2 procesors. The true consumption of the processors puts AMD's Athlon 64 X2 far in last. Even if you take into account the power consumed by a memory controller (5w or less, typical), the 45nm Core2 processors have very low power consumption, and the 65nm cores are competitive.

Now, let's read into those results. First of all, the Athlon 64 X2 6000+ listed (a typo, it's a 6400+ in the rest of the article) is a 90nm core, so it's not a fair comparison. You can add to the fray the Brisbane 5050e, which has 45w peak power consumption - this is more competitive with the Core 2 power consumption, although it's not amazing.

At the end of the day, the sad fact is this: every AMD dual core on the market will lose in a desktop performance battle with either the E8200 (27w load) or the E6550 (40w load). You need only look at the article I linked above, where an X2 clocked at 3.2 GHz cannot beat either of the above processors. Even ignoring the 45nm E8200, if you just compare 65nm to 65nm, the Intel chip offers similar power consumption (40w peak) with an incredible performance boost to boot.

And this trend isn't limited to dual-core processors. Intel's 65nm quad beats Barcelona in terms of performance per watt. On the desktop, Intel has the performance/watt leadership.

Re:This bodes well for the company

[Arterion]

I've noticed that for normal desktop use, the overall performance of the system isn't nearly as reliant on the processor as it has been in the past. My system at home has a Core2 Duo at 1.86 Ghz, and mine at work has a Core2 Duo at 3.16 Ghz (overclock to 3.8 Ghz).

Honestly, I can't really tell much of a different between opening Word and surfing the web.

At this point, I think "joe user" performance improvements are going to come from places beyond the CPU: the chipset, RAM speed (since quantity is a non-issue at today's prices), and storage speed.

I recently replaced a failed original hard drive in a Dell 4600c. It has a 2 Ghz Northwood P4 in it. I was really surprised just how much replacing the original 40GB hard drive with a new Seagate 7200.10 improved the performance of the machine.

I think it's the little things like that which will make the computers seem faster for most users. Unless you're crunching big numbers (gaming, video, scientific, audio, etc.) I think any run-of-the-mill dual core processor is going to be as good at a top-of-the-line model.

Re:This bodes well for the company

[BOFHelsinki]

AMD's real problems are in acheiving high clock speeds, and solving their fabrication process. If AMD's 45nm process is as improved as they say it is, and with their fabrication/design company split they should be able to get that side of their business under control.

The jury is still out whether splitting the fabbing into The Foundry Company was a proactive and beneficial improvement or a desperate and potentially harmful necessity. The worst case scenario is that AMD doesn't get preferential treatment for every chip run they want; the best case scenario is obviously completely different. Will be interesting to see. One would expect that the splitting causes at least some disturbance to the flow from design to fabbing -- at the minimum they have hard work ahead to make sure it stays seamless. It's hard to see anything in the split that as such helps with their process challenges (against Intel's evident leadership in reaching new process nodes). In other words, the split doesn't automagically solve any problems. (That said, I'm personally rooting for them; have been building AMD rigs since the Athlon 750 and I dislike the vitriolic bashing they are getting nowadays, even if they really dropped the ball with the way they marketed the problem ridden Phenom.)

Re:Hey, Libertarians!

[Anonymous Coward]

Somalia doesn't have strong property rights or equal treatment before the law. But if you've ever dealt with a government beaurocrat, you'd realize how stupid your argument is.

Chipset?...

[valnar]

Despite any advances AMD makes in CPU's, they still have such a sub par selection of chipset vendors. I'm very happy that Intel makes both the best CPU's and chipsets at the moment. It makes the decision easy. Because if AMD came out with a killer CPU but I had to resort to an NVidia/AMD/VIA chipset to run it, that would be quite a conundrum.

Re:Chipset?...

[SargentDU]

"Despite any advances AMD makes in CPU's, they still have such a sub par selection of chipset vendors. I'm very happy that Intel makes both the best CPU's and chipsets at the moment. It makes the decision easy. Because if AMD came out with a killer CPU but I had to resort to an NVidia/AMD/VIA chipset to run it, that would be quite a conundrum."

What are you complaining about? AMD also makes chipsets you said so yourself. AMD chipset. So they also make the CPU and chipset. Go slobber somewhere other.

Re:Chipset?...

[valnar]

My comment is on the inferiority of AMD chipsets. Can't you read?

Re:Chipset?...

[Chris Burke]

Despite any advances AMD makes in CPU's, they still have such a sub par selection of chipset vendors.

What's wrong with their PCI-E bridge chip? It converts PCI-E to HT and back pretty well afaik. Or maybe you meant the southbridge? Yeah, that USB and SATA logic is really cramping my gaming rig.

The performance-interesting parts of the northbridge are on the CPU in AMD architectures (and now intel ones too), and they're great. I'm not sure what you're complaining about.

Re:Chipset?...

[TheThiefMaster]

What are you saying, nVidia chipsets rock!

We've had someone stick up for amd, and now for nvidia. Anyone want to stick up for via chipsets? anyone? oh...

Re:Chipset?...

[valnar]

The problems with AMD (ATI), NVidia and especially VIA (though not common much more) chipsets are well documented. I do not need to go into them here. Suffice to say if you aren't encountering any problems with your particular rig, good for you.

Re:Chipset?...

[Chris Burke]

Oh you're talking about ancient-in-computer-terms history. Fun, but not really relevant.

Re:Chipset?...

[m50d]

Or maybe you meant the southbridge? Yeah, that USB and SATA logic is really cramping my gaming rig.

If it's not causing you any trouble, good. But I've never been able to get one with consistently working USB. Which yes, does rather cramp my gaming when my mouse stops working halfway through.

Future proofing?

[leandrod]

Perhaps the rationale behind buying nearer the bleeding edge than your sweet spot is not having to replace as often?

Re:Future proofing?

[wisty]

Or you can put the $500 you saved on the stock market, and by the time you need to upgrade you can use the money you saved, along with any capital gains and dividends to buy, um, a packet of waffles.

Re:Future proofing?

[Anonymous Coward]

Perhaps the rationale behind buying nearer the bleeding edge than your sweet spot is not having to replace as often?

Yes, but this needs to be balanced against the money you saved from buying closer to the sweet spot meaning you have more money to replace sooner (if you actually need to).

I've been building systems for ~15 years and I've always found that buying 1-3 notches below the "top of the line" usually gives a very reasonable bang for your buck.

The sweet-spot is definitely there, its just a matter of exploiting it.

Re:Future proofing?

[Moryath]

If you're buying "bleeding edge" you're not going to be satisfied with your purchase 2-3 years from now (about my replacement cycle for my personal box, and even then a lot of my components like hard drive / sound card / DVD drive tend to last through 3-4 iterations), unless your tastes suddenly radically change and you're no longer interested in the "bleeding edge" games you were trying to run.

Plus, consider the following two options:

#1 - "Bleeding edge" rig. Blow $900 on processor, $1200 on dual video boards, $400 on RAM, $800 or so on miscellaneous other components. Total system cost around $3000.

#2 - "Decent Gaming" rig, single $300 video board, $200 processor, etc. Total cost: $900 if you really push your luck.

I'll take my $2000, buy more games, take girlfriend to dinner, stick some in a rainy-day fund, etc. One of these years you need to run the numbers and then you'll figure out that the "savings" you claim are there from buying at bleeding edge aren't really there at all. Even if I spend $900 every 2 years upgrading my PC, it takes me 5-6 years to equal the cost of your rig, and I guarantee you're going to turn around and want to rebuild to get back to the bleeding edge because you'll be "disappointed" that your 2-3 year old "bleeding edge" machine is only getting 15 fps in the timedemo mode of CallOfUnrealCrysisDoomQuakeTournament 3: Yet Another Non-Scaling Tech Demo Masquerading As A Game.

Re:Future proofing?

[sumdumass]

It depends on where the sweet spot is and what else is going on.

I have found that if the sweet spot in capable of providing solutions at the time of the build, then it is still capable 5 years down the road. Vista being the exception of course. Even if your gaming, you will find that it is still fast enough although you want something faster. I just retired some Athlon XP 2100 machines last month which had nothing wrong with them, they were just 5 or 6 years old and when parts started failing, we decided to replace the entire bunch. I expect this latest batch which is in the sweet spot too, to be usable for another 4-5 years before component failure gets us replacing them again. Most companies I know of try to work 3-4 years replacement cycles on the same principle. But I know a few still sporting Novel netware 3x servers on 486 machines and a few sub 1 gig workstations running specialty programs on win95 and win98.

Re:Future proofing?

[zippthorne]

And what's so bad about replacing often? If, at the end of ten years, you end up having had more total cycles available AND having spent less money, AND have the additional reliability factor of not holding on to equipment for very long after it's no longer under warranty, what's so bad about that?

Re:Future proofing?

[leandrod]

If you're buying "bleeding edge" you're not going to be satisfied with your purchase 2-3 years from now

Speak for yourself. I tend to stay with stuff until it breaks.

Besides, I never said I am buying bleeding edge. I merely suggested buying nearer it than the original poster's sweet spot.

unless your tastes suddenly radically change and you're no longer interested in the "bleeding edge" games you were trying to run.

I run no games.

Re:Future proofing?

[0xygen]

Agreed - even if you buy "sweet spot" rigs twice as often as the bleeding edge guy, on average over the time involved you will have a better gaming experience.

"Bleeding edge" guy is essentially paying for the R&D which makes our 2nd generation cards, motherboards and processors cheaper and less power hungry.

Re:Future proofing?

[Windows_NT]

you have a girlfriend?
Im gonna get me one of those computers!
... Anyone wanna lend me $2k so i can get a hottie?

Re:Future proofing?

[Anonymous Coward]

I run no games.

...attempt no landings there...

Re:Future proofing?

[Sj0]

You subscribe to the bathtub curve of reliability. For many components, this isn't an accurate model.

For many components, after you get past infant mortality, the devices remain consistently reliable. I've seen 386s and 486s that are still running, day in day out, today. PDP11s simply don't die, and there are some that are just sitting in a corner quietly doing mission critical tasks in industry.

All you have to do is identify common failure modes and do maintenance to mitigate them. For example, the dominant failure mode of PCs tends to be fan failure due to dust build-up, followed by hard disk failure. The first can be avoided by cleaning the fans regularly on a schedule chosen based on the rate of dust build-up. The second isn't really a dominant failure mode -- I don't know anyone who has ever had to buy a replacement hard drive. Regardless, to protect against failure, simply keep back-ups of important data.

Re:Future proofing?

[Chris Burke]

Besides, I never said I am buying bleeding edge. I merely suggested buying nearer it than the original poster's sweet spot.

Well the whole point of the 'sweet spot' is that it is the highest performing part before the price for incremental performance drastically increases. It's the bend of the hockey-stick curve of price/performance. So no, it never makes sense to buy anything past the sweet spot if your goal is to optimize your computer purchases for amount spent vs longevity, because the extra amount spent will not buy you a proportionate amount of longevity.

There's basically two general cases where it makes any sense to buy past the 'sweet spot':
1) You have lots of disposable income so price efficiency isn't a big concern, and you can afford to upgrade as often as you want. In this case, you'd always be buying the bleeding edge. Vendors love these folks.

2) You have a one-time chunk of disposable cash to spend on bragging rights, but you don't care that in a year it won't be worth bragging about any more. In this case the only reason not to buy the bleeding edge is if your chunk of change isn't big enough to do that.

I've been in the second boat before, where I bought as much bleeding edge stuff as I could just for the fun of it, just because I could. Sure getting all that extra horsepower meant that the computer felt fast for longer than it would have otherwise, but not at anywhere close to the ratio of the extra cost I paid versus a more modest system. When I finally did upgrade to a new "sweet spot" system, it still utterly blew away the old one and I wondered how I was able to survive so long.

Re:Future proofing?

[Airline_Sickness_Bag]

I don't know anyone who has ever had to buy a replacement hard drive.

I've gone through 20-30 drives between home & work. Of course, quite a few were the infamous IBM deathstar drives. But I've had IBM, seagate, conner, maxtor, and HP drives fail. Some have failed after only a year (the deathstars less than that). I try to get cases with good cooling, and add fans for the hard drives, but I still have some fail.

Re:Future proofing?

[kraemate]

Huh?
This is classic consumerism at work.

What about all the waste you generate? You think getting a new computer every 6 months and dumping it in a landfill is worth the 100$ you save per decade?

And the effort going into buying something at the sweet-spot every 6 months?

Re:Future proofing?

[Sj0]

My two questions would be,

1. How many hard drives do you have between home and work that 20-30 could fail?

2. What the hell are you doing with them that they're failing at such a rate(assuming you don't have 20,000-30,000 hard drives? Studying EARTHQUAKES?

barcelona does 2.5Ghz now

[easycheesy]

Small correction here.

AMD Barcelona processors clocked at 2.5Ghz have been available for quite some time. They are however a 105 watt part.

reference list here

Tags are seriously broken

[Anonymous Coward]

Seriously, you click on the triangle by the tag, start entering a tag, and the menu pops *behind* the story below it. What good is that, does nobody notice this? Can we just get rid of tags already?

Re:Please tell me AMD is not betting it all on SIM

[initdeep]

and that reason would be?

oh, i know, to bleed money ferociously and take your company bottom line so low that it nearly killed it!!!!!

oh and now you can spin off parts of it (and hope to sell them to someone) so you can at least try to remain viable?

AMD has been so poorly run over the last 4 years it's just plain sad.

HOPEFULLY, they can get their shit together and remain viable, but this new chip isnt going to do it for the most part.
there just isnt going to be enough call for it wit the way spending on IT is slowing and the fact that except for the very top end, most servers out there are under utilized to begin with.

Re:Hey, Libertarians!

[Anonymous Coward]

YHBT. YHL. HAND.

Soo much power

[ripdajacker]

I'm not that old, but I remember when I had my first x86 pc with a whopping 133 mhz and 16mb of ram (which was virtually the top of the line back then).

The problem is with these kids get all that processing power and do not know what to do with it.

My 450mhz machine, which was essentially what replaced my commodore 64 was total overkill for my needs back then.

Hell I could do most of the programming in emacs, vi or even nano, but since I am blessed with a 2.3GHz dualcore machine at work and a ~3GHz machine at home, I tend to use an actual IDE, but I could be at least as productive without.

Today I've seen kids at age of 10 having a laptop at 1.5GHz or something, which doesn't teach them real computing, back then where freeing one extra MB of ram was essential for winamp not lagging.

If I had that computing power back then, I would be the happiest kid in the world.

Faster than the Ahtlon64 6400+

[Svenne]

So, thish seems as good a place as any to ask; what's the fastest AMD CPU available? I have a Socket AM2 6400+ and I'm looking for an upgrade without changing the motherboard. I'm talking single core operation, that is, I don't care if a good threaded app runs faster on a quad core Phenom than on my dual core 6400+, I just need it to run one application that doesn't thread, on one core, really fast.

Re:Faster than the Ahtlon64 6400+

[wild_berry]

The Phenoms are AM2+ and, while they sit in the same socket, they will at-least require a BIOS update if they are to work; it's likely that they won't. There may be tiny IPC and process improvements in the last set of X2's, but I doubt any such would be worth it.

Re:Faster than the Ahtlon64 6400+

[ergean]

The improvement is small, don't waste your money on that. Depending on what you do, you can improve the performance with some dual channel memory with better timing.

Refractive fluid

[jarek]

The exotic "refractive fluid" in question is deionized pure water, for those who wonder. However, make no mistake, this is really a small wonder. In the early 2000, few people believed immersion lithography at 45nm would ever become mainstream. However, 157nm became toooooo expensive and none of the problems anticipated with immersion lithography appeared. It is almost, as with the bike, to be considered a discovery. Hence, a full node thanks to the refractive index of water (1.44 at 193nm)

immersion liquid

[megamike23]

What liquid is AMD using for immersion? Water? Last I heard moving to second generation liquids with a higher index of refraction was impossible because no one could make a prism that would allow the light to transmit into the fluid - kind of like when you are swimming in a pool and you look up and see a mirror like reflection instead of seeing outside perfectly

Both?

[Anonymous Coward]

Why they will offer "performance and performance-per-watt" gains is beyond me. (Notice the AND).

Re: ECC RAM

[Lonewolf666]

Accorcing to Wikipedia (http://en.wikipedia.org/wiki/Hamming_code#Hamming_codes_with_additional_parity_.28SECDED.29), a scheme that can correct 1 bit error and detect 2 is typically used. So it can correct single errors. The most common reason is some form of radiation (see http://en.wikipedia.org/wiki/Soft_error#Causes_of_soft_errors).

Against at least one of those (cosmic rays), even quality RAM is not immune. This said, only the vendors of quality RAM seem to be in the business of making the ECC version anyway. Which makes it more expensive, because you're buying more of the good stuff (the parity bits for the SECDEC algorithm require an extra chip on the module). But considering the overall price of RAM, that should be not a problem.

a day late and a dollar short

[DragonTHC]

This is hardly exciting. AMD should have released these a year ago. Now they are irrelevant.

You are missing out..

[Junta]

I might not have said much if you said you had a P4 desktop, but...

The power envelope of a non-netburst processor makes laptops much much better. Heat and battery life on my P4 laptop were quite unbearable.

Convenience

[leandrod]

And what's so bad about replacing often?

Convenience. I tend to use e highest quality equipment I can afford, and to run free software, so I don't need to reinstall or worry much about machines and OS. Having to reinstall because e machine is obsolete is someðing I want to postpone as much as poßible.

Re:Convenience

[zippthorne]

Wait.. why do you pay the apple premium if you're not going to use the apple OS?

Depends...

[leandrod]

it never makes sense to buy anything past the sweet spot if your goal is to optimize your computer purchases for amount spent vs longevity, because the extra amount spent will not buy you a proportionate amount of longevity.

It all depends on how much one hates shopping and reinstalling.

Re:Depends...

[Chris Burke]

Well then you're not optimizing money spent vs longevity, or anything really. And you're still not really buying yourself much, if any, time. You'd really want to be buying the bleeding-edge in that case, at the maximum price premium, to have any significant impact.

Re:Depends...

[marcansoft]

The latter only applies to Windows folken. I've stuck with the same Linux install across three major system upgrades (CPU/mobo/RAM/VGA/etc), which amounts to about 5 years. The only reason I reinstalled it in the first place was to switch to a 64-bit system, and I've kept that same 64-bit install across two AMD CPUs and now an Intel quad core. I'm about to buy a laptop and will probably transfer this same install to it, while keeping a clone on the desktop.

How will this compare to Nehalem

[shadowofwind]

...for I/O bound applications? Anyone have info on that yet?

Why limit oneself?

[leandrod]

if the sweet spot in capable of providing solutions at the time of the build, then it is still capable 5 years down the road.

Why limit oneself to five years? If I run free software and high-quality components, I can avoid the haßle of a new setup for longer.

You have a point

[leandrod]

I've stuck with the same Linux install across three major system upgrades (CPU/mobo/RAM/VGA/etc), which amounts to about 5 years.

You have a strong point there.

Fix a messed up bios chip- IOpener

[link-error]

  When hacking my Iopeners, I learned you can pull the bios chip from a running computer, put in the bad one, then just re-flash it.
    Just have to be careful putting it in not to short the pins. Worked great.

  -Mike

amd MB prices are way lower then core i7 mb prices

[Joe The Dragon]

amd MB prices are way lower then core i7 mb prices.

This is all fine and dandy... But...

[mcnazar]

.. does it run Linux?

780g boards have way better video some even have s

[Joe The Dragon]

780g boards have way better video some even have side port ram so you don't need a video card and don't use system ram for video.

Re:Please tell me AMD is not betting it all on SIM

[Muffinmasher]

It was to diversify so they don't die at the first sign of increased pressure from intel. ATI's Graphics cards are now extremely competitive with nVidia's, especially for the price point.

Re:Please tell me AMD is not betting it all on shr

[waferbuster]

The higher Numerical Aperature lithography tools are definitely helping for making narrow lines (hence faster transisters) for both Intel and AMD alike. However, the biggest advantage Intel has in the chip-making business is the use of hafnium for forming the gate of the transister. As Gordon Moore put it, "It's the biggest change in transister technology in the past 40 years."

The rest of the industry is feverishly trying to match/duplicate the hafnium process improvements which Intel discovered. Unless there's some equivalent breakthrough at AMD (which is highly unlikely), Intel will retain the crown for performance.

Disclaimer: I work for Intel, the above is my opinion and I am not a spokesman for Intel. Heck, I'm just a lowly peon. I'm not even authorized to tell you the time of day!

Faster CPUs are less important

[bradbury]

Faster CPUs are lees important than "intelligent programming". I am sitting here with an old Prescott P4 CPU which is essentialy doing nothing and yet 13+% of the CPU time is being consumed by galeon and 30+% of the CPU time is being consumed by X.

Dedicate some of the design and engineering expertise to the software, rather than the hardware.

Silence

[leandrod]

why do you pay the apple premium if you're not going to use the apple OS?

I know of no other so silent machine, which endures so much.

Re:Silence

[zippthorne]

I think you missed the question.

Why would you dump the perfectly good operating system which is already installed and specifically designed/configured around the hardware?

For the price of even a low-end apple, you could get one mondo silent pc (some assembly required, maybe)

There is still a need for more CPU power!

[Ed Avis]

I recently discovered gnofract4d and it reminds you that, whatever you think about CPU usage for web browsing or programming, computers still aren't fast enough :-p. It's like the old days when the first thing you'd do with that shiny new Pentium-66 workstation was to see how fast it could run Fractint. It seems we have to wait at least a decade for high-resolution fractals zoomed in real time. (There is XaoS but it has a limited choice of fractal types and needs to interpolate pixels. The answer may be to use GPU hardware, as FFFF does, but preferably using a compiler like Brahma that translates a high-level language to GPU calls.)

Re:There is still a need for more CPU power!

[Ed Avis]

Actually, I tried the latest XaoS (it has improved a lot since ten years ago) and it does support a wide range of fractals, though not as many as gnofract4d.

Not in the US

[leandrod]

Why would you dump the perfectly good operating system which is already installed and specifically designed/configured around the hardware?

Because it is not perfectly good for me. I dislike having to pay for upgrades and additional software, being restricted to the original CDs to reinstall -- they have a way of being lost --, I miß Debian's package management and so on. I really value free software.

you could get one mondo silent pc

I am not in the US. In fact, sometimes I can't even get an Apple down here. And I really can't be bothered to aßemble my own.