AMD Undercuts Intel With Six-Core Phenom IIs

EconolineCrush writes "As Slashdot readers are no doubt aware, Intel's latest 'Gulftown' Core i7-980X is an absolute beast of a CPU. But its six cores don't come cheap; the 980X sells for over a grand, which is more than it would cost to build an entire system based on one of AMD's new six-core CPUs. The Phenom II X6 line starts at just $200 and includes a new Turbo capability that can opportunistically raise the clock speed of up to three cores when the others are idle. Although not as fast as the 980X, the new X6s are quick enough to offer compelling value versus even like-priced Intel CPUs. And the kicker: the X6s will work in a good number of older Socket AM2+ and AM3 motherboards with only a BIOS update."

Comments

Holy crap this is old.

[plague911]

But anyhow. I like AMD they are a good brand but to be honest their 6 core dose not undercut intel's 6 core. It maybe undercuts intel's 4core . But even than they only trade blows in quantitative analysis.

In short this posting is old and not very accurate. So doubly pointless

Re:Holy crap this is old.

[dimeglio]

Although not as fast as the 980X, the the new X6s are quick enough to offer compelling value versus even like-priced Intel CPUs. And the kicker: the X6s will work in a good number of older Socket AM2+ and AM3 motherboards with only a BIOS update.

So doubly pointless

Indeed as this is the "the" new X6s.

I still like the underdog and hope they do well. The latest and greatest is often overkill.

Re:Holy crap this is old.

[petermgreen]

Unfortunately for AMD there is little money in being the underdog of PC processors. Intels better process technology and income from high end chips like the 6-core i7* means they can set prices on the low-midrange stuff at a level that is comfortable enough for them while being extremely painful for AMD.

*Which unlike most extreme edition chips (which tend to cost a shitload of extra money for a marginal improvement over thier regular counterparts) doesn't seem that bad a deal to me. Afaict it will get you the performance of a lower end dual-quad (and more in processes where only some stages are multithreaded) at a similar CPU cost (2.4 GHz quad core 5500 series chips seem to be about $500 each, the 6-core i7 is about $1000) without the expense of a workstation board and chassis. The dual-quad soloution supports more ram though and has more total cache (though of course in the dual-quad that cache is split between two chips) which will be an advantage in some applications.

Re:Holy crap this is old.

[gstoddart]

But anyhow. I like AMD they are a good brand but to be honest their 6 core dose not undercut intel's 6 core.

If I R'd TFS correctly, by about $800 dollars. But you mean purely based on dick-swinging numbers I assume. :-P

See, for a lot of people (ie. non gamers and people not doing CPU intensive stuff) being CPU bound is rarely something they'll encounter. Multiple cores have the benefit of making the operating system more responsive since a busy app doesn't make the whole system crawl. My current Quad core has probably never had all four pegged at once, so I don't need faster. In fact, I'm not sure I've needed faster in a bunch of years.

Now, for me, the best way to get the most out of a machine is to put what sounds like an obscene amount of memory on it so it's future proof. You can survive software bloat if you piled on the memory up front. :-P

Re:Holy crap this is old.

[bloodninja]

There is another good reason to buy AMD: competition. Just think of what would happen to the computer industry if hardware were dominated by a single vendor (much as software is today). If you think that Microsoft is Teh Evil with it's stranglehold on software, a stranglehold on hardware would be much worse.

Always buy from the underdog.

Re:Holy crap this is old.

[sofar]

If everyone bought the underdog, wouldn't they be the market leader?

Re:Holy crap this is old.

[Anonymous Coward]

Unfortunately for AMD there is little money in being the underdog of PC processors.

I don't agree. If a shop is thinking of deploying 6-core CPUs then it's because it deals with highly parallel applications and if we are dealing with highly parallel applications then you get a decent performance bump by throwing cores at the problem. Knowing that, if we consider AMD's offering is priced a mere 20% as Intel's offering then it means that by the price you get a single 6-core Intel CPU you can get around 5 times the cores from AMD. And there's no way a 6-core CPU from intel can compete with 5 AMD processors on a highly parallel application.

Re:Holy crap this is old.

[Xiph1980]

No, because then the other company would be that underdog, and you'd buy from them.

Re:Holy crap this is old.

[petermgreen]

and if we are dealing with highly parallel applications then you get a decent performance bump by throwing cores at the problem
Only if those cores are actually fast enough. AMD hasn't even reached the top end of intels quad-core stuff let alone come close to touching the intel 6-core i7.

Given the smaller die area and the way intel is known for getting very high yealids I'd expect the intel eqivilients to these chips have a lower marginal cost. That means if intel desires they can push the prices to a level where AMD takes a loss on every chip without doing so themselves.

Re:Holy crap this is old.

[Phoghat]

I've always used AMD CPUs for their bang for the buck. I don't need bleeding edge performance.

Re:Holy crap this is old.

[Thundersnatch]

Funny, we've been buying all-Intel servers for the same reason recently. If you factor in total system costs, the CPU cost is dominated by the aggregate of others (chassis, storage, power, network switches, cooling). It therefore makes sense to have as few servers as possible, with as much throughput per server as possible.

We actually build a simple solver for several architectures when designing our most recent server cluster, and Intel-based system had a 5-year TCO nearly 30% lower than an AMD-based solution (using 1U servers and clustered iSCSI/NAS storage) of equivalent aggregate performance. In almost all designs, top-of-rack stackable switches and distributed bottom-of-rack UPSs were TCO winners as well.

Re:Holy crap this is old.

[Anonymous Coward]

and that gives you how many fps in crysis?

Re:Holy crap this is old.

[Anonymous Coward]

Your analysis is flawed. MIPS per WATT per Dollar is what matters. AMD wins.

Re:Holy crap this is old.

[CastrTroy]

You're right, if your goal is to pack a lot of processing power into a small volume. Because, let's face it, data center space is expensive. Way more than the CPUs. However, if you are building machines that don't need to be packed tightly, like for your home, or for employees in the office, then you don't really have to worry about density. In this usage scenario, you still need 1 PC per user, no matter how fast the machines are. So it makes sense to use cheaper machines, that may take up a bit more space, and may not be as powerful, if they still provide everything the user needs. In the server room, Intel makes the most sense, and probably will for a long time to come, but on the desktop, unless you go for the extreme high end, AMD is probably the better choice.

Re:Holy crap this is old.

[Ironhandx]

Um? Right. Get at least one or two facts straight before claiming to know something about the industry.

So #1 amd has reached the top end of the quad core, the new x6 matches them for most everything.

#2 AMD does and has had higher yields than almost every other single manufacturer in the business, not just Intel, everyone. Coupled with this the fact that their current manufacturing deal means they don't pay a single dime for wasted silicon and who the hell do you think is going to win the price war again? Intel was taking losses on certain chips not so long ago just to compete with *AMD* in the price war, not the other way around. AMD for almost 2 years with the Athlon 64 series was winning both performance and pricing. Intel was selling more expensive chips that didn't perform as well.

Oh, and heres a recent source for most of that. I thought I was going to have to go back a little ways, maybe into 09 to get something, but apparently AMDs manufacturers are outdoing themselves yet again so we don't need to worry about that.

http://www.brightsideofnews.com/news/2010/3/17/globalfoundries-produces-zero-defect-wafers2c-10025-yield.aspx

Re:Holy crap this is old.

[famanz]

Anandtech put up a good review of the new chip and included some benchmarks against some i5s and i7s. Here's a link to the review on a single page.

It was outdone by the i7 860 almost universally. The 860 has 4 cores and can be found for $200 if you live near a Microcenter. It also draws less power, generates less heat, and the superior intel turbo boost technology means it runs single-threaded applications faster.

I like AMD too but other than the mid/low range market they really just aren't competitive these days. I think the new chips are a step in the right direction and I'd love to see them get back to their Athlon glory days.

Re:Holy crap this is old.

[petermgreen]

See, for a lot of people (ie. non gamers and people not doing CPU intensive stuff) being CPU bound is rarely something they'll encounter.
True but I doubt those people were in the market for intels extreme edition chips in the first place.

Multiple cores have the benefit of making the operating system more responsive since a busy app doesn't make the whole system crawl.
This is true up to a point (it's very true going from single core to dual-core and somewhat true going from dual-core to quad-core) but you quickly get into diminishing returns. Typically an app will either hog just one or maybe two cores or it will be built to use as many cores as it can.

Assuming a comparable aggregate performance I'd definately take four faster cores over six slower ones.

Re:Holy crap this is old.

[jgagnon]

Considering Intel actually makes more money with each die shrink, I doubt they're going to change their strategy just because AMD goes out of business. They live an die by demand, and raising prices lowers demand, especially when we're living in a time where the average desktop can do a lot more than the average software on it is asking for.

Re:Holy crap this is old.

[Anonymous Coward]

Well, I figure a lot of people just get whatever the best bang for the buck is at the time - because they want to keep running the same chip for a good 3-6 years if they can. And the funny thing about that is that software grows a lot in a 3-6 year period.

A casual user probably isn't hitting the limits of a quad core CPU today, but casual programs are going to start taking advantage of extra available cores. Even a really simple example comes to mind: tabbed browsers with their own process per tab (and, if applicable, their own javascript engine and/or JVM and/or flash instances). High def video sends my triple core system into the 30-50% range, and it IS using multiple cores to do it. Future interpreted language implementations are potentially going to have both a virtual machine and compiler going at the same time, with a profiler running and deciding what parts are worth compiling on the side on another core; it'll make stuff run faster overall, IF you've got the extra cores lying around to handle that introspection on the side. Multithreading already happens in other common things too... like anything that does spell checking on a text field; these days, that's my word processor and my browser and my messenger program. Another complexity level on top of that will be translation of web pages.

We're already running a lot of background processes. Often any given one is just idling, but sometimes they'll just happen to all want to do something at the same time. With a chat client, messenger, firefox, and mp3 player going, my CPU utilization ranges from 3% to 17%. A much wider range than one might expect. And I have no virus scanner going, and my browser isn't running any java or flash right now.

Lastly, casual gaming will certainly grow to absorb whatever extra capacity is common. The game AI can certainly benefit from extra power, and richer UI design could use it too (even for just the simple and obvious stuff, like keeping lists and menus sorted for you in real time, and keeping the more relevant clickable things arranged efficiently as the game state changes). Certainly there's also the bloat vs development speed tradeoff, also; any time the CPU cycles are available to trade off vs publishing a game faster, there WILL be faster-published games making that deal, and those games are usually the ones aimed at casual play. Think about the occasional flash games that the average user plays every now and then, and just imagine them (or their future equivalent) taking more resources, and that's a pretty safe guess about the future.

Re:Holy crap this is old.

[Anonymous Coward]

You linked to BSON as if it was a serious source of information. You tried to claim AMD doesn't pay anything for wasted (defective) silicon. You tried to claim AMD has lower manufacturing costs than Intel. Taken together, all these things mean that you're pretty much a fanboy moron without a clue.

Intel was taking losses on certain chips not so long ago just to compete with *AMD* in the price war, not the other way around.

Got a source for this? (Doubt it.)

This should drive the i7 price down

[geekoid]

finally...

Re:This should drive the i7 price down

[MSFanBoi2]

Price/Performance of the i7's is actually quite decent.

Re:This should drive the i7 price down

[sznupi]

Yeah, a lot of people waiting for i7 price to drop instead of actually buying nice AMD product will surely result in drops of Intel CPU prices, right?

Re:This should drive the i7 price down

[Jah-Wren Ryel]

Yeah, a lot of people waiting for i7 price to drop instead of actually buying nice AMD product will surely result in drops of Intel CPU prices, right?

Of course it does. It doesn't matter why someone chooses to not buy a product, it only matters that they make that choice and thus the product doesn't sell. Companies have gone bankrupt because people chose to wait for a better deal.

Re:This should drive the i7 price down

[sznupi]

Intel will sell lots of CPUs regardless; it was just about this group waiting specifically for price decreases "forced" by potential attractiveness of a nice product from smaller competing vendor, a situation which GP basically subscribed to.

Re:This should drive the i7 price down

[Anonymous Coward]

Waitaminute, Intel makes real CPUs? I thought they were overpriced motherboard CPU socket fillers used in shipping.

Re:This should drive the i7 price down

[sootman]

Thus resulting in a lower price on the new i7 MacBook Pro!

waiting... waiting...

Re:This should drive the i7 price down

[alfredos]

Companies have gone bankrupt because people chose to wait for a better deal.[citation needed]

Re:This should drive the i7 price down

[Anonymous Coward]

LOL u phail quoting.

Re:This should drive the i7 price down

[Jah-Wren Ryel]

Figures that someone with such a high-uid would be ignorant of one of the most spectactular failures in the tech industry.
Probably happened before you were even born.

Re:This should drive the i7 price down

[alfredos]

Or maybe it happened in another country. However, perhaps you who have such a low-uid and are so old and wise can illuminate me?

Re:This should drive the i7 price down

[Beezlebub33]

This is one of the most famous mistakes in the technology industry. Learn about the Osborne effect.

Re:This should drive the i7 price down

[moosesocks]

No, you're right. The rising cost of Intel's high-end chips is killing Apple at the moment, primarily in their lower-end machines. The 13" MacBook Pro is no longer the alluring machine it once was, as it's got a seriously underpowered processor (by "New Computer" standards and comparison to the bigger machines). For the past several years, part of Apple's shtick has been that their tiny laptops are just as powerful as their big ones (I still love my 12" PowerBook).

The current generation Mac Mini is pathetic, even by low-end computer standards, and seems like a prime candidate for an AMD chip (even a 4-core model, although that risks competing directly with Apple's higher-end products). I've got a Core Duo Mini as my primary machine, and would love to replace it with a more modern equivalent, although the current models are barely an improvement over my 4-year-old machine.

Re:This should drive the i7 price down

[Jah-Wren Ryel]

It was Osborne Computer - they were as well known as Apple in their day.

Re:This should drive the i7 price down

[snadrus]

Then why did MS do a Get The Facts campaign?

I need a new computer

[Dayofswords]

all these cores and benchmarks...

i still run computer with one core and no modern graphics card

Re:I need a new computer

[Anonymous Coward]

all these cores and benchmarks...

i still run computer with one core and no modern graphics card

Me too, it makes a great router.

Re:I need a new computer

[Rogerborg]

Rest assured that most of the 1337 h4xx0rZ who will soon spew reams of artificial benchmark trivia are just demonstrating that what they really use their Maibatsu Monstrosity XP9000 system for is running a web browser.

Re:I need a new computer

[thoughtsatthemoment]

Please share the list of software that you use and enjoy.

Re:I need a new computer

[tepples]

On my Dell Mini 10 (2-thread Atom CPU, Intel GPU), I spend most of my time in Firefox, gedit, GIMP, ca65, and FCE Ultra. It was the same on my Eee PC 900 before it gave up the magic smoke.

Re:I need a new computer

[InlawBiker]

It gives us VM's - lots and lots of VM's. I can reproduce a production app environment entirely on one quiet little box, including the load-balancer, firewall and name servers. It used to take a half a rack of loud, expensive servers all with disks and other stuff that breaks and needs monitoring and replacing. I can't wait for the 8-core chips to become affordable.

Re:I need a new computer

[Anonymous Coward]

Sounds like the stuff I've used regularly on my K6-2 machine years back. Makes me wonder where all this GHz waste is really allocated on!

Re:I need a new computer

[Ethanol-fueled]

My dell laptop, a 6 year-old Latitude D600 with 1.6 gHz processor, 512 MB Ram, and integrated low-end ATI runs Ubuntu Studio 9.10, most compiz effects enabled(the only thing that really slows it down is a background skydome image, so I don't run that) with multiple desktops and multiple browsers(including instances of YouTube) or Jack/Ardour/Hydrogen running without slowing.

The only speed problems I have are the time it takes to fire up certain applications like OpenOffice and Gimp, but they run OK once open.

My next desktop will have a core2 duo or AMD equivalent and a more modern graphics card, making it WAY more than good enough, and the tower with monitor's still gonna cost less than an iPad!

Re:I need a new computer

[thoughtsatthemoment]

making it WAY more than good enough

I believe we programmers have the magic to make today's 32nm dual core to function like your old processor.

Re:I need a new computer

[skam240]

So why on earth are you even bothering to comment on this article? You clearly have no need for a top of the line system. Good for you! You're just like my mom! Does it make you feel superior to brag about your single core? Are you the computer ascetic of our generation?

Re:I need a new computer

[Sepiraph]

That's true, how much computing power does an average, non-gaming, non-power user need? Even if you throw in multi-media, including voip and video, I doubt your average user will be able to use all that computing power (also the later depends more on the GPU, which will become more and more important in the future). I suppose with the cost of cpu power getting lower and lower, it may not matter as much and we will see more hardware comes with some sort of computer (e.g. your entire house).

Recently, I built a new system based on the core7 i930, reasoning being there is simply no AMD cpu that can match its performance, and I do consider myself a power user when I am using it for emulating Cisco CCIE labs and probably be running several VM instances in the future.

Re:I need a new computer

[gullevek]

watching HD porn of course!

Re:I need a new computer

[Anonymous Coward]

lol u mad bro?

Re:I need a new computer

[Draek]

Perhaps because he's *really* thinking about upgrading? not all of us feel a physiological need to purchase every newest-and-shiniest the moment it's announced, but sometimes we do like to have shiny toys to play with and perhaps he just believes that this provides an affordable way to get a nice, sizeable upgrade from his current machine.

Hell, I'm hardly a ricer l337 g4m3rz yet if all I need is a couple hundred bucks and a BIOS update to run this chip, even *I* may be tempted in spite of my current dual-core still serving me well.

Not *all* posts here on Slashdot are sarcastic, though I can see why you'd get confused.

Re:I need a new computer

[MBGMorden]

That's true, how much computing power does an average, non-gaming, non-power user need?

Not much to be honest. I put multi-cores into the servers I build (no such thing as "fast enough" when I'm running complex SQL queries), but we long ago passed whats needed on the desktop. I'm even what most would consider to be a power user. I can notice faster processors when I encode video and the like, but 99% of what I do feels the same on any system past 1.6Ghz or so (and anything with 1.5GB+ of RAM) - regardless of the number of cores you stack on. I've continued to upgrade every now and then (currently on a Phenom 9850, so I'm not jumping the gun TOO much) because, well, that's what I do, but truthfully I don't need it.

Not to worry though. There are plenty of other things in computing to be excited about. Bigger/higher resolution monitors, bigger/faster hard drives (which as always been the slowest bottleneck), quieter/lower power systems at the same speeds, etc. We have plenty of room to improve with computers - it's just that improvement no longer means just faster and faster speeds.

Re:I need a new computer

[hairyfeet]

Yes but you a forgetting an important use for these new CPUs for "normal users": Futureproofing. I built my 67 year old dad a quad core, and most here would think I was nuts for giving an ordinary user a quad, but I thought that 3Ghz P4 would last him quite awhile, but then he started discovering Internet TV, and video IM, and soon when added to his normal programs that P4 started to slog. This way no matter what new things my dad decides to try in the future he is covered. The board can hold 16Gb of RAM, has an AMD 4200 and a PCIe X16 in case he becomes GPU bound, I just doubt dad will ever find a way to saturate a quad, which means less BS for me, which is always good.

Plus it isn't like you HAVE TO pound those cores all the time to see a benefit. When I built dad's I gave myself an AMD 925 quad, and thanks to having quad cores I have the machine with C&Q running most of the time at 800MHz, yet it never feels slow or sluggish. And because it is able to drop down so low without bogging down the fans can drop to virtually nothing, my apartment doesn't become a space heater, everything just runs nicer than the AMD 7550 CPU which it replaced. And considering I paid less than $650 with W7 HP X64 and 8Gb of DDR 800Mhz RAM, I'm sold on AMD. Sure having the largest ePeen is good for bragging rights or running a shitload of VMs, but honestly? This AMD quad is faster than I'll probably need for a loooong time.

Re:I need a new computer

[mwvdlee]

No, they use it for running artificial benchmarks.

It's like giving me all your money just to see how poor you can get; a pretty smart idea!

Re:I need a new computer

[mwvdlee]

no such thing as "fast enough" when I'm running complex SQL queries

If SQL itself is good enough, there definitely IS a "fast enough". If there really is no such thing as "fast enough" for your situation, your system would be leading the Top-500 list.

Re:I need a new computer

[mwvdlee]

Not *all* posts here on Slashdot are sarcastic, though I can see why you'd get confused.

Indeed. Some are ironic, right? right? ;)

Re:I need a new computer

[Vectormatic]

not to rain on your parade, but if your dad got a socket am2+/am3 machine, then why not build him a machine with a 'fast enough' cpu now (like the amd athlon II X2 240 in my dads machine), and leave the futureproofing to the socket/pricedrops? If your dad would feel his x2 is getting slow, dropping in an x4/x6 is trivial, and by that time, probably much cheaper/faster

i would say though, that if the budget had allowed for it, my dad would have an athlon II X4 as well, just for kicks..

*fights off urge to order a quad-core to replace the 7750 in his system*

Re:I need a new computer

[hairyfeet]

Because the quad was less than $125, which for a quad CPU was frankly beyond cheap. Hell if you built your dad an AMD you might want to look at an Athlon X4 while they are just $99, and use his X2 for a good netbox (or in my case the 7550 is gonna go into a BDay prezzy for my GF). Don't forget that CPUs quit being manufactured all the time, and if you miss out you might not get another chance. I tried it your way with dad's P4, but I held off on buying a Pentium D figuring I could get one later. Instead by the time his P4 was slogged down I found a Pentium D that would fit his motherboard would damned near cost more than a new CPU plus motherboard!

So don't be too skimpy, now is the time to score AMD quads. Believe me it won't be long before the 6 and 8 core CPUs are the norm, and it will be harder and harder to find the quads. If you used a good budget or business class board like I did for dad and me you'll find they don't get the BIOS updates for new CPUs nearly as much as the gamer boards. That is why I went ahead and got the biggest quads both mine and my dad's boards will take, because I know that mine will probably last me a good 7 years or more, hell maybe even a decade with the light gaming I do, and I honestly think that new AMD quad will last dad until the machine finally dies of old age.

Oh and OT, but does anybody know of a good remote desktop solution for someone old and completely clueless on W7 HP x64? Dad refused to listen to me on OSes and got W7 HP on both his office and home machine and would love to be able to remote in from home whenever the weather is bad, but everything I try is deemed "too complicated" for him to use. I can't leave him wired in 24/7 because he ends up screwing something up, so I need something simple enough he can just "clicky clicky" and at worst put in a password. So far I've tried both VNC and Comodo remote, but both never seemed to work well for him and I spent more time troubleshooting than he did using it. He has static IPs at both work and home, so all I need is a good and most importantly butt simple way for dad to get his work desktop at home. Anybody got any suggestions?

Re:I need a new computer

[Ndymium]

But... How can you surf the web with a car?

Re:I need a new computer

[shallot]

Recently, I built a new system based on the core7 i930, reasoning being there is simply no AMD cpu that can match its performance, and I do consider myself a power user when I am using it for emulating Cisco CCIE labs and probably be running several VM instances in the future.

I'm told that most Cisco machines have processors the speed of an Amiga 500 inside. How many thousands of those do you need to emulate that you actually need an i7? :)

Re:I need a new computer

[Lennie]

My Soekris box doesn't even have a graphics chip you insensitive clod !

Re:I need a new computer

[Lennie]

I would replace bigger/faster hard drivers with SSD, that's where the real improvements are now, I really hope the prices on those will go down soon.

Re:I need a new computer

[Vectormatic]

well you have a point there, high end chips for certain sockets tend to lose their value rather slowly

And as i said, had the budget allowed for it, then i probably would have dropped a quad in my dads machine, but he set me a budget X for a new web-machine, and i couldnt justify telling him to spend more on a quad.. he wouldnt see the difference for years anyway.

Re:I need a new computer

[amorsen]

I'm told that most Cisco machines have processors the speed of an Amiga 500 inside.

Then you were told wrong. Even the main CPU on current Cisco routers is magnitudes faster than that 7MHz MC68000, and many routers rely on the hardware to handle moving packets while the main CPU is just for running routing protocol daemons etc. When you emulate, you don't get hardware acceleration, so you better emulate a decently fast main CPU.

Re:I need a new computer

[jackharrer]

Logmein.com is the simplest I found. Web-based and free. Try it. There are other options but this one looks easiest for non-IT inclined person.

Re:I need a new computer

[Anonymous Coward]

lol u illiterate bro?

Re:I need a new computer

[pyite]

The emulation isn't the most efficient. Without a decent CPU, you cannot emulate complex routing topologies. The biggest problem is that once you're emulating too many devices, you start dropping packets between them causing routing protocols to fail and effectively making the lab useless.

Re:I need a new computer

[MBGMorden]

You mistake me. Saying "nothing is fast enough" doesn't indicate that money is no object. I still have to balance what we can afford to do with what I want to do. It's just that for example, I have a nightly batch job that runs on my server that recalculates fee information for real estate parcels. It's a fairly complex system which has to query buildings on the parcel for type, square footage, use, attributes that can exempt them from fees, etc. The process takes nearly 2 hours to complete a run each night. That's occuring during an already crowded schedule of lots of other things happening (many of which simply can't happen concurrently). Anything that can shave down that time is a good thing. Optimizing the code is on the schedule (and a lot of optimization has already been done), but throwing ever faster processors at the problem helps out too.

For things like that, I don't think there's ever a point where I'll say "You know what - I'm satisfied. I don't need any faster of a machine.".

Re:I need a new computer

[Beezlebub33]

I'm not sure what the problem is with VNC that you are having. If you need something pretty simple on the client end (i.e. at home, logging into the work computer), you can try Guacamole which uses HTML5 canvas to view it in a browser.

Re:I need a new computer

[pu'u_bear]

He is running W7 both places on a static IP? Why not just use remote desktop? I know, I dislike Windows A LOT, but for this problem, is there any reason he can't just use remote desktop? If you have access to the office machine, configure the firewall to only allow remote desktop to connect from the home IP, set the preferences on the home machine to connect to the work machine and start in full screen, and you are done. You can even link the rdp file to his desktop and he gets double-clicky access.

Re:I need a new computer

[Daengbo]

I just priced a full setup like the one you're describing (AMD X2/board/memory/box, HD, 19" LCD, Keyboard and mouse) for $335 with shipping. Just thought I'd let you know there's no need to wait.

Re:I need a new computer

[Anonymous Coward]

It gives us VM's - lots and lots of VM's. I can reproduce a production app environment entirely on one quiet little box, including the load-balancer, firewall and name servers. It used to take a half a rack of loud, expensive servers all with disks and other stuff that breaks and needs monitoring and replacing. I can't wait for the 8-core chips to become affordable.

8 core sock G34's start at 299 :)

Re:I need a new computer

[hairyfeet]

Because W7 HP can't be a server, only a client. It was one of those asshole moves MSFT did to try to "upsell" you to W7 Pro, which is why I pushed dad for pro but he got one of those family packs because he looooves HP so much.

Say what you want about MSFT they really got W7 right IMHO, this stupid RD problem not withstanding. It is easy for me to tweak HP X64 for my needs, while for those that are clueless like my dad it really is intuitive. My dad has found more features and learned to do more things on W7 than on any MSFT OS before it.

So while I would love to just RD it, from what I understand you can't remote from one home machine to another, just from home to pro. And dad loves HP too much for me to talk him into an upgrade. Nice idea though, damned shame it won't work. If you or anybody else knows how to get around the restriction I'd sure love it. I've set up RD for clueless XP Pro users and never had a lick of trouble, but dad really needs windows 7, as he would manage to "break" XP every time I turned around.

Re:I need a new computer

[bingoUV]

Have you tried this ?

Though in my experience, VNC is almost as trouble-free as rdesktop, I don't understand what your dad feels is complicated about VNC. Some network port blocked?

Both VNC and rdesktop have different quirks for clipboard, so I have to use both.

re AMD

[freddieb]

We all should hope AMD does well. I use AMD chips in about 90% of my systems. Value is the main reason. Intel makes excellent products however you invariably have to upgrade the motherboard to use a new chip. AMD has been kinder in this regard recently. I go with a middle of the pack system anyhow and I really appreciate the value AMD provides.

Re:re AMD

[RyuuzakiTetsuya]

Intel's kind of funny like that.

LGA775 was kicking around for like, 5 years before LGA1156, but, Socket 478 was around for about 3 years before *that*.

of course, there's also Socket 7, who can forget those days?

Re:re AMD

[bruno.fatia]

Yeah but try installing a new processor into an old motherboard with just a BIOS upgrade, see how that goes...

Re:re AMD

[angelwolf71885]

yah except with intel even if the new cpu used the same socket the new chipset was required for the new cpu forcing an upgrade 755 broke the mold but intel are back to there old tricks with 2 or 3 cpu sockets that ARNT compatible with each other amds guilty of it too 754 949 939.. but they eventually chose 939 for the definitive socket for the first gen 64 bit cpus then moved on to the AM sockets all are compatible with each other AMx doesn't work in AM3 BUT AM3 works in AMx

Re:re AMD

[RyuuzakiTetsuya]

I meant to say was that they seem to do what AMD did, which was float between like, 2 or 3 CPU socket types then settle on one(for the consumer level at least. If you're buying octo-core Xeons you're probably not building it yourself) for a good half decade.

Re:re AMD

[Anonymous Coward]

While it kind of sucks to have to upgrade the motherboard for a new chip, I went back to Intel from AMD on my last build and couldn't be happier. All three AMD systems I'd put together had their quirks -- from mild to major -- that I chalked up to shitty motherboard manufacturers (including one extremely well-known brand that shipped a malfunctioning board right back to me and called it tech support -- nevermind that their Northbridge chip was melting their passive cooling solution off the board. I never overclock, either.)

I finally decided to throw another couple of hundred dollars at my next system and went for the best price/performance chip at the time (Intel Pentium D 3.4) on an Intel motherboard, and found memory that had been tested with this combination (which seemed to be a real problem with AMD boards.) Everything has worked perfectly for five years -- and I've pushed the hardware with gaming, compiling, and virtual machine experimentation -- so I'm going to do the same for the next system.

When I go for a new chip, I've got to get a new board, but there are other advantages; my video card is a PCI Express x16 2.0 but my motherboard only has a 1.0 slot, which means it's running with half the bandwidth the card could use. To me, the difference would be enough to spring for a new motherboard even if I was able to drop a new CPU into this one. So I have to figure, if I can get five years life out of my computer systems, it's totally worth the premium given that other motherboard-related technology usually progresses a step or two over the same time period.

Re:re AMD

[TheRealFixer]

Personally, I miss Slot 1. So easy to upgrade/replace processors. Just a little bit too bulky, though...

Re:re AMD

[socceroos]

I certainly hope they do well. Many of my systems run AMD too. They do represent great value. Plus, I would hate to see Intel as the only desktop CPU manufacturer.

Re:re AMD

[RyuuzakiTetsuya]

Not only that, but all that stuff off die is quite frankly, slower.

Although I wish they'd go back to some sort of riser so we can not worry about destroying the motherboard when installing fans onto Core series processors.

Re:re AMD

[sznupi]

LGA775 was really two sockets though, with clear compatibility break around the middle. Also, wasn't there a recent story on /. that 1156 will be quite soon replaced?

Re:re AMD

[RyuuzakiTetsuya]

Yeah. LGA1366.

Cores vs performance

[Darkness404]

The problem is AMD is using an outdated architecture. More cores != more speed for general use. Yeah, if you are compiling your own software you can get things to work really fast with 6 cores but how many applications really take advantage of multiple cores? Very, very few. A single fast core can outperform a few slow cores in general usage and AMD seems only concerned with getting more and more cores on a single CPU die which really doesn't translate to great performance in the real world for general use.

Re:Cores vs performance

[h4rr4r]

Try transcoding some video one time kiddo.

Hell if I could get 24 atoms in one socket that would be fantastic for me.

Re:Cores vs performance

[by (1706743)]

$ ps aux | wc -l

If that prints anything less than 6, hats off to ya!

Re:Cores vs performance

[MrHanky]

It depends on what you consider "general use". Some cases demand more cores. GTA 4 is more or less unplayable on dual core systems, so an AMD is the cheapest option. As always, look at what you want to do, and then buy a computer, and don't be a fucking idiot.

Re:Cores vs performance

[Trepidity]

Almost all those processes spend almost all their time idle or blocking on something, though, not contending for a core.

Re:Cores vs performance

[blind biker]

This was common wisdom 5 years ago. Nowadays, there's a shitload of CPU-intensive applications making good use of additional cores. And the trend is towards more and more such applications. This, in turn, means that the architecture that allows for easier multi-core CPUs will win out.

Re:Cores vs performance

[dimeglio]

Well if your load average is always less than 0.10 your computer is likely overpowered.

Re:Cores vs performance

[TheRaven64]

$ top

If that prints more than a couple of processes in a running state, not sleeping waiting for I/O, you have quite an unusual workload.

Re:Cores vs performance

[Trepidity]

Indeed, which is why things like netbooks have gotten popular: modern computers are overpowered for most normal end-user applications.

Re:Cores vs performance

[Ryiah]

More cores != more speed for general use. Yeah, if you are compiling your own software you can get things to work really fast with 6 cores but how many applications really take advantage of multiple cores? Very, very few.

Even if it were true that very few general use applications take advantage of multiple cores, the average user is running more than one application. For example, I looked at a system a friend was running and she had Windows 7 and Norton AV on a Celeron 900. She typically does Office applications and runs a browser at the same time. Compared to her old system its pretty snappy, but it still isn't anywhere near as decent.

Re:Cores vs performance

[wisty]

Microsoft has a cunning plan to deal with that ...

Re:Cores vs performance

[h4rr4r]

Your GPU method only supports certain video types, this method supports far more.

Re:Cores vs performance

[palegray.net]

The more cores the better for me. I run a couple of home servers, which each run several VMs under KVM. Sure, this isn't general use, but I don't care about that. For all my "general use" tasks, the MacBook works just fine.

Re:Cores vs performance

[ld a,b]

Your real-world usage is what exactly? Playing badly designed games?

I want to play badly designed games *while* I am compiling, listening to some music and possibly leaving my browser on with some badly written JavaScript running. I also want my CPU not to melt.

You would need at least a 5GHz CPU to match a current dual-core CPU in this area. The ongoing trend is to have more and more things running and getting updated in real time. An it has been for a long time.

Files getting indexed, illegal files getting downloaded, stupid GUIs getting rendered, music getting played, Interpreted languages getting JIT-compiled ...

Gamers are still stuck in the microcomputer era. The real world isn't. And there isn't really a choice in the first place, the choice is more cores and a better experience or getting stuck at XGHz and having to pipe liquid Hydrogen into your home.

I think we will see more CPUs with more cores and likely more storage units to avoid resource starvation. More speed is just not possible.

Re:Cores vs performance

[hedwards]

I thought they got popular because they're cheap and easily portable.

Re:Cores vs performance

[ElectricTurtle]

I didn't know Jen-Hsun Huang posted on /.

Re:Cores vs performance

[NicknamesAreStupid]

More cores? Imagine how many 8051s you could put on a Phenom die, then imagine what you could do with it. You have an incredible imagination!

Re:Cores vs performance

[binaryspiral]

Virtualization, son! And when your single threaded game is using all the cycles, the chip increases its clock speed.

Re:Cores vs performance

[dbIII]

Do something with video or graphics and then all those cores suddenly do give you an extra 100% performance per core. For everything else at least two is always good simply to reduce the situations where a single application leaves you waiting and looking at an unmoving screen with a keyobard and mouse that won't listen to you.

Re:Cores vs performance

[Anonymous Coward]

Such as?

Re:Cores vs performance

[Anonymous Coward]

Because Nvidia is the ONLY manufacturer of GPUs...

Re:Cores vs performance

[sznupi]

BTW, for the part with leaving browser (+badly behaving JS) in the background, suspending its process works sensibly to unclog the machine / give you one more free core.

Re:Cores vs performance

[Anonymous Coward]

Back in the days before the Intel Core series, when AMD was doing well, Intel's outdated architecture was still better at encoding video. The only benchmark that Intel P4s would consistently beat AMD in was the video encoding benchmarks. But more people play games than encode video.

The iCore series completes the catchup that Intel had to do to get back into the performance gaming. Wider bus and on die memory controller easily won gaming performance benchmarks for AMD. The iCore series increased the bus, and put the memory controller on the die.

The Boost vs Turbo features sounds like Intel has the more innovative product.

These days its looking more like AMD is the one doing the catchup in innovation.

Re:Cores vs performance

[dangitman]

Indeed, which is why things like netbooks have gotten popular: modern computers are overpowered for most normal end-user applications.

That doesn't explain why netbooks are popular*, because they are underpowered for most normal end-use applications. Are you saying that people prefer underpowered computers?

*Note: there is no evidence that netbooks are popular.

Re:Cores vs performance

[Angst Badger]

Yeah, if you are compiling your own software you can get things to work really fast with 6 cores [...]

Welcome to Slashdot, user #1287218. As it happens, a lot of us are compiling our own software. And many of those that aren't are running multiple simultaneous processes that are doing a lot more than waiting for user input. I'm probably not alone here when I say that I can use as many cores as I can get -- in my case for rendering jobs and data analysis. My brother, who works in video production, probably wouldn't turn his nose up at a truckload of extra cores, either.

Anyway, point being that you're probably right about "general usage", but general users don't have the background to understand a discussion of parallel processing and will probably buy their next box from an equally clueless sales rep at Best Buy on the basis of the neato metallic red color of the case.

Re:Cores vs performance

[phantomfive]

I am amazed at the shifts that happen sometimes here on slashdot. A few years ago, multiple cores were the cat's meow and were going to change the world, and all the programmers who didn't learn Haskell or other non-parallel languages would be left behind. If you tried to post a comment saying that a parallel algorithm doesn't always make things better, you would get modded down or ignored, or laughed at.

Now it's gone to the other extreme, you've said multiple cores are essentially useless for the average person, and got modded up. Can there be no middle ground? Something like, multiple cores are better than not having them, but they aren't going to change the world?

Re:Cores vs performance

[Anonymous Coward]

Hey buddy! I compile shit all the time (several times per day). 6 cores beats 4, all day long. I also encode video. Not every day, but again, 6 beats 4. Scan line rendering? Again, 6 beats 4. If you are a gamer where your world is "duh clawk speedz" then yes, sparky, your world isn't getting any better. For me, more cores is more speed. I have applications on my desktop than can (today) support 64 threads of execution. With hyperthreading, that means 32 cores with hyperthreading, or 64 cores without, and I can use every bit of it, right now today (well, actually yesterday.....no, no, actually about 3 months ago). The compiling can take care of up to 512 threads. The freaky 8 way boxes that IBM builds and sells for about 60 grand (with the 8 core Xeon processors) offer 64 cores, or 128 threads. I can eat all of that, 4 times over, today.

Re:Cores vs performance

[MemoryDragon]

Simply use a decent graphics card and CUDA... beats the hell out of everything else.

Re:Cores vs performance

[Vectormatic]

Friends don't let friends run celerons? (or norton....)

Honestly dude, virtualize that mail server you are running on the P4 in the basement and give that thing to her, it's the right thing to do

Re:Cores vs performance

[RAMMS EIN]

``The problem is AMD is using an outdated architecture.''

Are they, really? Last I looked into it (admittedly a few years ago), I concluded AMD's architecture was _better_, with more scalable memory access (Intel's would choke when more than a few threads were running simultaneously), a better cache protocol (I forget the details there, but it also had something to do with multiple threads), and more instructions per clock tick. Compared to that, Intel had higher clock speeds, especially for memory, which _just_ gave it the edge in single-thread performance, at the cost of higher energy consumption.

So, while perhaps AMD's architecture is older, "outdated" implies that it also isn't as good as the current state of the art anymore. I would like to know if that is actually the case or not, and for what workloads one would prefer which model of CPU.

Re:Cores vs performance

[Anonymous Coward]

As times go on, having VM clients on even home machines makes sense. At best Web browser security is ending up at an impasse (mainly whack-a-mobile with add-ins now), so having the ability to browse in a VM that gets its snapshot rolled back as soon as you are done with that session adds a lot to personal security (as well as being anonymous due to shared objects being dropped). And with advances in "seamless" UIs that VMWare and XP Mode bring, as machines get faster, the CPU overhead of context switching into and out of a VM becomes less of an issue, so an app is just as usable inside a VM as it is running natively. What this will allow is not just isolating suspect stuff, but also keeping sensitive data in a protected VM where the data does not leave the VM unless it is connected (via the internal VM network) to print. Of course, if the host OS is compromised, all bets are off, but it goes a way to keep sensitive data isolated from anything else.

Re:Cores vs performance

[Narishma]

What games have you been playing? Most modern ones require a dual core and some like GTA 4 won't work well if you don't have at least a quad core.

Re:Cores vs performance

[drinkypoo]

Yeah, if you are compiling your own software you can get things to work really fast with 6 cores

Compiling doesn't provide a multithreading advantage. You have to be writing your own software, too.

but how many applications really take advantage of multiple cores?

Practically all complex games are multithreaded today. Essentially all multimedia applications are multithreaded. Or in other words, any application which needs to be specially coded to take advantage of multiple cores probably is already.

A single fast core can outperform a few slow cores in general usage

Only for legacy or other non-threaded applications. Both groups are dwindling. In addition most of the heaviest lifters have been multithreaded for a very long time, i.e. Photoshop.

and AMD seems only concerned with getting more and more cores on a single CPU die which really doesn't translate to great performance in the real world for general use.

AMD is concerned with the number of instructions they can retire in a given number of cycles, which is higher than intel's, and long has been... and probably long will be.

Re:Cores vs performance

[Rockoon]

The worst game port in the history of computing (in terms of efficiency) doesnt make for a good example.

Its almost like they are software emulating the Xenon chip.

Re:Cores vs performance

[Anonymous Coward]

such as the 'far more' video type, duh.

Re:Cores vs performance

[BLKMGK]

Video compression is my excuse - I'm tempted to buy the Intel 6core actually! I currently run a 4 core I7, with Hyperthreading I see 8 cores and I max that out 100% for hours at a time compressing HD video. Yes, the pseudo cores help. I went from 24 hour sessions on some movies to under 4 hours just by moving to this CPU. Additional instruction sets, 8 cores, and 4.2ghz vs my previous 2 core 3.2ghz are all responsible. this puppy is water cooled and still hits 82degress Celsius! Would I take more cores? Hell's yes! Would I want to move to an AMD 6core vs the Intel.... I dunno'. I will look at the benchmarks but my gut says no that it wouldn't be much of a help. I know that an Intel quad core running slower than my dual still sped things up but this I7 920 screams and with AMD's general lack of performance lately I'm not sure even another 2 cores would match it. I guess the benchmarks will tell.

Oh and not this isn't a profession or a job just something I do for my HTPC. I do it often enough that I want it faster too! Shows recorded on my DVR, BluRay, and other sources all provide grist for this particular CPU :-)

Re:Cores vs performance

[Beezlebub33]

and they do what we need 95+% of the time.

All this seems silly to me at this point. If I have a big computation job (and I do, often), I spin up a bunch of 8core machines on Amazon EC2. It's frigging $0.68 per computer hour. That's $0.085 per core hour. Why, oh why, would I want to actually buy my own computer when so much computational power is just sitting there for me to rent?

Re:Cores vs performance

[Cytotoxic]

No, this is Slashdot. Everyone here is smarter than everyone else in the world. Particularly anyone who runs a multi-billion dollar company. You should know that by now....

Re:Cores vs performance

[CAIMLAS]

Less than .10? Isn't that "idle, just sitting here chatting"?

Even on a fast system, doing something like watching Flash videos, playing flash games, etc. will take at least one core up to .9 or so. Sitting here at mostly idle, I've got a load of .3 or so, and this is a quad core system.

At .10, I'd imagine most computers would be "just turned on and waiting for a login".

(Never mind your fraudulent claim. "Overpowered" computers? Surely you're kidding. More power can almost always be used, if for nothing else then a more responsive experience.)

Re:Cores vs performance

[Daengbo]

There are far fewer transcoders that use the GPU than use the CPU, so the answer to your questions is ... any of them?

Re:Cores vs performance

[Alpha830RulZ]

Any sort of development running a local database is one example. I do Java/postgres work on a single machine, doing numerical analysis. When I'm doing a run, which is about once a day, two procs are slammed for 2 hours. The other two are still sufficient for browsing /. .

Re:Cores vs performance

[blind biker]

I do that, too, actually, but I'm using a 2-core AMD.

Re:Cores vs performance

[BLKMGK]

Yup and taking way longer to do it I'm sure. I've not got a clock ticking off deadlines for me or anything but when I did Watchmen for my HTPC and it took near 24hours to compress I nearly had a cow! Obviously I go for very high quality and max size compression and I also want to be able to use my computer for other things. 2cores just wasn't enough even overclocked sky high. 4cores made a huge difference, 6 would be even better. the Hyperthreading is icing on the cake! My current machine even beats CUDA implementations that only use the graphics card and I get way more flexibility on profiles and filters. If the Intel CPU weren't so expensive I'd own one now. If the AMD could prove itself better than my existing crazy setup I'd sure consider it if the price were right...

Re:Cores vs performance

[blind biker]

We're not the most discerning movie watchers. We're dead tired by the time we have the rare opportunity to at least sit down to watch a movie together, and even the crappiest quality will do.

It's what happens with a baby in the family.

So, the 2 core AMD is plenty fast for everything, really.

Re:Cores vs performance

[Anonymous Coward]

You can get 24 cores in a system today,

      http://www.newegg.ca/Product/Product.aspx?Item=N82E16819105264

That's 24 cores + motherboard for about $2500. Complete system is $3000+, depending on amount of RAM you want.

Re:Cores vs performance

[magus_melchior]

And there isn't really a choice in the first place, the choice is more cores and a better experience or getting stuck at XGHz and having to pipe liquid Hydrogen into your home.

Hydrogen?? Bah! Real overclockers use liquid oxygen!

I mean, who wouldn't want a rig to fail spectacularly when it dies? This way you get a boom and a fire.

Re:Cores vs performance

[Anonymous Coward]

Since nobody has been able to name a single one, I'm betting that you are talking about insignificant or non standard codecs that nobody has ever heard of.

All common and advanced video formats like h.264 and XviD/DivX have codecs with GPU support. The fact that you refer to a "transcoder" (which is not a real term, by the way) instead of a codec proves that you haven't a fucking clue about encoding video. Encoding software use the codecs that are installed, which means automatic GPU acceleration for a properly configured system.

Go learn about how all of this shit works before you open your mouth. Right now you look like a complete fucking idiot who might have used some simplistic crap software like MediaCoder a couple of times.

Re:Cores vs performance

[ElectricTurtle]

To hear them talk you'd think they were. And they're constantly harping about how GPGPUs are teh future and how processor companies need them so bad.

Re:Cores vs performance

[BikeHelmet]

The problem is AMD is using an outdated architecture. More cores != more speed for general use.

Ahh, but the problem with Intel is price. Everything else is great - just not price.

Very, very few. A single fast core can outperform a few slow cores in general usage and AMD seems only concerned with getting more and more cores on a single CPU die which really doesn't translate to great performance in the real world for general use.

True, but more cores for less money lets you spend money on more I/O - and your CPU strength really doesn't matter when disk I/O holds you up.

So regardless of whether 4 cores or 6 cores is better, cheaper IS better because it lets you spend more on HDDs or SSDs, which will have a far bigger impact on workstation use.

But if I could get 12 Phenom II cores in a system... that would be a dream come true for x264 encoding.

Re:Cores vs performance

[sjames]

Sorry, that's not possible. We as a society are deeply into the extreme. Even our deodorant and toothpaste must be exTREEEEEEEEEEEM! There can be no middle ground.

Value for money vs FanboiGasms

[w0mprat]

On a price performance basis AMDs Phenom IIs have consistenly been a better buy for some time now. To the point it's hard to suggest anyone buying intel at all, unless money is no object. (I don't know why I bought Intel anyway :S). Honest hardware review sites (that aren't far up the ass of vendors) are at the point of recommend AMD CPUs on a price/performance basis.

http://www.tomshardware.co.uk/best-gaming-cpu,review-31857.html

It seems Intel doesn't get even a "honorable mention" until page 3. At $120 price point, Core i3 gets a look in. Oh, they also don't recommend anything above about $160 to quote Tom's: "Best gaming CPU for $190: None".

To add further insult, money saved from AMD motherboards being cheaper (in particular SLI/xfire AMD boards are a good whack cheaper) will let you put money towards more storage, a SSD or a step up in CPU speed.

Re:Value for money vs FanboiGasms

[sznupi]

Plus the integrated GFX on most AMD boards is a bit more sensible than Intel one, meaning separate card can be more often ignored (or at the least the initial configuration not including it, and the machine will be still sensibly nice)

(yes, there's integrated Nvidia - not with latest Intel arch though; previously not so readily available...and for some reason motherboards for Intel with Nv GFX were consistently more expensive than for AMP CPUs with Nv GFX, at least where I am)

Re:Value for money vs FanboiGasms

[DDLKermit007]

Ummm...you do know that your BOARD dieing doesn't have jack to do with your CPU right? Unless of course you bought another board to find out somehow the CPU died which I never see happen outside of some dumbass not having good enough thermal control on the system. Buy a better board & PSU. I'll bet anything the board you used was a POS vs what you've used for your Intel rigs.

Re:Value for money vs FanboiGasms

[KingKaneOfNod]

The reason I buy Intel CPUs and chipsets is because I had a bad experience with an AMD X2 dual core CPU about a year ago. I'd had the CPU/motherboard for about 3 years and constantly battled compatibility issues with games (the main use for my computer); most games had timing issues (e.g. ran way too fast - kind of like playing a game made for an 16 Mhz 286 on a 40 Mhz 386), even after I'd applied all the XP patches and installed the AMD dual core optimizer. Then, to top it off the motherboard died (never had that happen before). So after that I bought a Core 2 DUO and have never looked back; no compatibility problems with any of the games the X2 had, and it has run very reliably.

The ironic thing is that my last two PCs before that had AMD chipsets that I had been happy with, but one bad experience like that has been enough to put me off AMD for good.

Re:Value for money vs FanboiGasms

[dimeglio]

Not my experience. All AMD based mobos and processors I purchased in the last 10 years have lasted many years past their useful life. I'd buy AMD again with confidence.

Re:Value for money vs FanboiGasms

[Nadaka]

I am running an M4A78T-E with ATI HD 3300 integrated graphics. It does surprisingly well. I have not doled out any money for new high end games for a while, but it easily handles games that brought my previous graphics card to its knees (it was top of the line in 04). I am eventually going to get a modern graphics card so I can play around with OpenCL, but I really have not felt a pressing need for it with my gaming habits.

Re:Value for money vs FanboiGasms

[angelwolf71885]

then DONT buy an SiS or VIA based mobo buy one with an AMD chip set or an NVIDIA chip set VIA if your desperate for cheep

Re:Value for money vs FanboiGasms

[LamboAlpha]

I will start this by saying that I have built four AMD desktop computers over the years and currently use three AMD systems (a single and quad core desktop and one dual core laptop). The quad core desktop was built as a server (video processing/storage/backup).

But, it is now time to upgrade my primary single core desktop computer (gaming and general use). I would prefer it if I could get an AMD system, however after doing the research, I am currently planning on building an Intel Core i7-930 based system. I completely agree that I could get an AMD system at a better price, but the performance would be worse than the Intel system. The point of this system is good performance at a decent price.

I had already planned to get the Intel based system before reading this review, but the Phenom II X6 review by Tom’s Hardware just reinforced my original decision. http://www.tomshardware.com/reviews/amd-phenom-ii-x6-1090t-890fx,2613-14.html

Re:Value for money vs FanboiGasms

[Jeremy Erwin]

It seems Intel doesn't get even a "honorable mention" until page 3. At $120 price point, Core i3 gets a look in. Oh, they also don't recommend anything above about $160 to quote Tom's: "Best gaming CPU for $190: None

and then... you stopped reading.

Best gaming CPU for $200:

Core i5-750

The new Core i5 brings top-of-the-line Nehalem-class performance at a $200 price point. We recently awarded it our Recommended Buy honor after seeing it stand up to more expensive CPUs in games and other demanding apps.

They don't recommend spending more than $200, though.

Re:Value for money vs FanboiGasms

[Anonymous Coward]

I've no particular preference regarding Intel vs. AMD, but I will point out that in practice the word "fanboi" always, without any exceptions ever, means "someone who likes something I don't".

Re:Value for money vs FanboiGasms

[Anonymous Coward]

every computer I have built over the last 15 years still works, both intel and amd.. maybe you just buy shit hardware.

Re:Value for money vs FanboiGasms

[AnonGCB]

What games are you playing that don't properly clock control? I've never had that hapen, and have been building with AMD since socket 754 was new, and playing games about that long as well. I call BS.

Re:Value for money vs FanboiGasms

[Galactic Dominator]

Your anecdotal stories are really only relevant to you. You'd be better off on /. presenting some sort of statistical evidence for your claim otherwise it's simply FUD and readers are correct to dismiss it as such. We're all here for conversation so if you have a real point bring it.

Re:Value for money vs FanboiGasms

[ElectricTurtle]

My counter anecdote is this: I have built nothing but AMD systems for my own use since my first K6-2 in 1998. I have never had an AMD chip or a motherboard for same fail on me. I would wager you just buy crappy boards.

Re:Value for money vs FanboiGasms

[sznupi]

You said yourself that if you go for AMD then you go for "cheap", meaning the cheap, crappy motherboards (and PSU most likely). You self select for higher failure rates of them.

Re:Value for money vs FanboiGasms

[sznupi]

Hell, I wish SiS were actually available - from K7 times they were a very solid, trouble-free chipsets (and with PCI implementation rivalling Intel). Of course with the problem of usually being put on the cheapest motherboards... (and this is the real reason for higher failure rates from GP poster)

Re:Value for money vs FanboiGasms

[sznupi]

It's hard not to get "good performance at a decent price" these days, at least if you're paying the tiniest bit of attention. Certainly works for AMD pretty much across their product spectrum.

Re:Value for money vs FanboiGasms

[elashish14]

There's another reason that I would consider Intel: in every benchmark/testing suite that I've seen, it almost always has lower power consumed. It probably amounts to little cost in the short run, but idle power draw actually is significant over long scales (roughly $1 for each watt over the course of a year of on-time). So after say a year of use, you can save about $15 choosing a i3 instead of an Athlon X4. It could be significant, especially if you plan on using your machine for a long time or with a lot of uptime.

Re:Value for money vs FanboiGasms

[angelwolf71885]

SiS did indeed have a damn good pci interconnect i had an SiS 330 chip set back in the day the turbo button such fond memory's and an AMD k6-2 the built in video card died in it and i replaced it but ive been an AMD fanboy since then replaced the SiS with an nforce2 and Athlon 1700+ Thoroughbred engineering sample and since then ive been an NVIDA fanboy but i defiantly enjoy ATI i would love to try one of the AMD chip sets after the merger with AMD VIA hav become quite good since they bought crisix im looking forward to the VIA nano 3000 64 bit

Re:Value for money vs FanboiGasms

[nabsltd]

Best gaming CPU for $200:

Core i5-750

The new Core i5 brings top-of-the-line Nehalem-class performance at a $200 price point. We recently awarded it our Recommended Buy honor after seeing it stand up to more expensive CPUs in games and other demanding apps.

They don't recommend spending more than $200, though.

I'm lucky in that I live near a MicroCenter store, and they are currently selling the i5-750 for $180 and the i7-860 for $200.

Since they both use the same socket, the extra $20 for a lot more performance (133-266MHz, depending on turbo-boost, plus twice the threads and better virtualization) is a no-brainer. Even with the $40 bundle discount on AMD CPU/motherboard from MicroCenter, Intel is still far and away the price/performance leader.

Re:Value for money vs FanboiGasms

[Anonymous Coward]

Hell, I wish SiS were actually available - from K7 times they were a very solid, trouble-free chipsets

HAHAHA!!! ....are you talking the same SIS I know?

Perhaps I'm the unlucky one who, for years, had experience with different SIS-equipped motherboards which not only had an (easily measurable) awful performance but also lots of bugs. Certain SIS chipsets required sacrificing a goat in order to be able to (badly) run Linux.

Re:Value for money vs FanboiGasms

[keeboo]

I'm not claiming it is a rule for AMD processors, but I've had a K6/233 (not K6-2 nor K6-3) which died after a couple of months.
The processor was properly cooled until the end, it was never overclocked and ran idle most of the time. Still, it just broke.

Re:Value for money vs FanboiGasms

[Anonymous Coward]

Hell I'm pretty much an intel guy and even *I* wouldn't insult amd's reliability that much, even dating back to the socket 5/7 era. With the exception of overheating parts (which AMD was initially notorious for, until intel decided to create the P4 just to say they could technically beat AMD in ALL areas of cpu design (:D), amd parts have rarely been any less reliable than intel so long as the peripheral components didn't fail (PS, video card, keyboard/mouse shorting, etc). Quite frankly out of the myriad of AMD based systems my friends have had, the only ones I've seen dead were due to fan failures, and almost to a T those were Athlon XP era parts from being the on board thermal diode. After that, there hasn't been a single failure I'ev seen that wasn't DOA from the factory.

Re:Value for money vs FanboiGasms

[sznupi]

Since K7 times. SiS 735, 745, 746, 748...negligibly slower in memory & cpu benchmarks than the fastest available chipsets, but very trouble-free and with fabulous PCI implementation, rivalling Intel - that was the reason why many videographers lowed them and bought en masse; and great value for the performance (chiefly because the only other working choice was...Intel chipset)

They worked like a champ, no real problems (especially MSI built on 745 was popular - it wasn't the bottom of the line motherboard, it was build to solid standards) - one would think we would notice if there were any, considering how much we pushed those machines.

There was also quite nice chipset in early K8 times, but by then hardly nobody noticing them took its toll. But every X360 ships with SiS southbridge...

Re:Value for money vs FanboiGasms

[twidarkling]

Your anecdotal stories are really only relevant to you.

I'm unsure if you didn't notice that I'm different from the GP of my other post or not, but I didn't offer any anecdotal stories about AMD sucking. In fact, I said the AMD CPU was more solid than the Intel, the opposite of the GP's own anecdotal claim. I simply pointed out that boards are connected to CPU. I also said "if" at the start, since it wasn't my claim. The other reply by sznupi has the good point that the quality is reflected in the pricing (again, not of AMD, of the motherboards, the choice of which is affected by choice of processor, though). AMD mobos tend to have lower price entries than Intel boards, which would be the ones most people unwittingly purchase and have failure rates on.

We're all here for conversation so if you have a real point bring it.

I'd suggest you read my post again in light of the realization it's not the same poster as the other. And as for how you reached +4 informative, I've no clue since you've no information in your post at all. Your post would have been better directed to Darkness404.

Re:Value for money vs FanboiGasms

[laughingcoyote]

I've got an Athlon XP 2100+ that serves as a general purpose test and data server, and a 900 MHz Athlon Thunderbird (yep, that old) that works as a file dump. Both have been in more or less continuous operation since I got them, and the only thing I've ever had to replace was the power supply on the 900 a few years ago.

That's quite aside from my 4 year old AMD laptop, which also still is kicking along quite well, or the AMD systems I've built over the years for others and are still doing fine. So I can tell you with assurance that the AMD systems in my experience have all lasted longer than "2-3 years", and in one case, coming up ten years now.

So when I went to build this system, did I go with a Phenom? I most certainly did. I've had very good experiences with AMD so far. Intel might be great too, but I've never had any type of speed problems with AMD chips, and they're half the cost. So I've really got no reason to try Intel anyway.

Re:Value for money vs FanboiGasms

[Kartu]

On a price performance basis AMDs Phenom IIs have consistenly been a better buy for some time now. To the point it's hard to suggest anyone buying intel at all, unless money is no object.

As a long time AMD fan, I beg to differ.
At roughly 190$ Core i5 750 (2.66Ghz) is an absolute "value" beast from Intel. It costs on par with Phenom II 965BE (3.4Ghz), consumes considerably less power idle/full load, is on par performance wise on stock clock, yet easily overclocks to 3.8-4Ghz with stock cooler/voltage and becomes simply unstoppable.

Re:Value for money vs FanboiGasms

[Vectormatic]

my 6,5 years old Athlon XP still runs smoothly

Granted, i have some trouble coming to terms with the fact that my Geforce 4 isnt the bees knees anymore (got a GTX260 in my main machine, but i likes me oldskool games), but the cpu/mobo are just fine and dandy.

Yes AMD had some quality issues in the past, but these days a decent AMD mobo and a matching cpu are a very trustworthy combo

Re:Value for money vs FanboiGasms

[Vectormatic]

i have to agree with you, i've had numerous AMD based gaming machines, socket 754, 939, am2+, and NEVER had any issues with clockscaling, nor has my little brother who is on his second AMD gaming rig

Re:Value for money vs FanboiGasms

[Vectormatic]

Certain SIS chipsets required sacrificing a goat in order to be able to (badly) run Linux.

hmm, i havent tried that yet on my mini-itx board (intel d201gly2, damn sis chipset doesnt have propper video drivers)

*searches e-bay for a cheap goat and a sacrifical knife*

Re:Value for money vs FanboiGasms

[Vectormatic]

if your main rig still is a single core machine, what kind of change in usage patterns do you expect that a Phenom II X4/X6 wont handle well enough? The quad server already handles your video processing..

good performance at a decent price is pretty much AMD's bread and butter, intel seems to go more for 'premium performance at a premium price'

Re:Value for money vs FanboiGasms

[alfredos]

The reason I buy AMD is because Intel sued my company because its name started with the same letters - without any bad faith (we didn't sell counterfeit processors, or try to impersonate as anything from them, etc). The company was alive and relatively well known in its small niche for many years before the Big Guys decided they wanted to piss us off. Heck, we were even a certified Intel reseller!

All of a sudden, all the things I had read about Intel's legal belligerence had a new and painful dimension.

The only Intel processors I buy now are those that come inside my favorite desktops - Macs. For the servers I use a company (AMD) that performs at a really nice price/performance point and that hasn't sued us to fill in the yearly legal battles report. Because I buy a lot of servers, I know that Intel has lost much more than they won. Do they care in the slightest? - I don't think so. It would be nice if they did. But the fact that they don't give a damn doesn't make me reconsider my decision.

Re:Value for money vs FanboiGasms

[GreatBunzinni]

That's odd. My 4yo AMD system still works like a champ. I guess my anecdotal evidence kind of cancels yours, doesn't it? Well, except that I'm at this moment typing in it while you have absolutely no way to provide a shred of evidence to back up your statement.

Re:Value for money vs FanboiGasms

[RavenChild]

That sir, is the suggested retail. I just bought an i7-930 for $199 (Microcenter). Just shop around and there will always be a good deal better than MSRP. The DDR3 and motherboard were what really killed me. But in all, I got a new top of the line i7 system with 6Gb of RAM for ~$750. I built a Phenom II X2(4) 555BE with the same specs and it came out to be only $100 cheaper in the end.

Re:Value for money vs FanboiGasms

[fredrik70]

Well, I have 2 AMD systems back home, one is from 2003 and the other is early 2006, both works fine, the older being my file server. About to upgrade and I cannot see any reason why not to go with AMD again.

Off course, this does not prove anything

Re:Value for money vs FanboiGasms

[Rockoon]

This.

A few months back I did some research into performance/price ratio and did up a little list:

Core2 Quad Q8200, score = 3255, price = $184, value = 17.69
Core2 Quad Q8300, score = 3570, price = $150, value = 23.80
Core2 Quad Q8400, score = 3668, price = $170, value = 21.58
Core2 Quad Q9400, score = 3756, price = $190, value = 19.77
Core2 Quad Q9505, score = 4016, price = $240, value = 16.73
Core2 Quad Q9550, score = 4291, price = $260, value = 16.50
Core2 Quad Q9650, score = 4559, price = $330, value = 13.82

Core i5-750, score = 4219, price = $195, value = 21.64
Core i7-860, score = 5570, price = $280, value = 19.89
Core i7-870, score = 5871, price = $540, value = 10.87
Core i7-920, score = 5590, price = $289, value = 19.34
Core i7-950, score = 6309, price = $570, value = 11.07
Core i7-960, score = 6727, price = $590, value = 11.40
Core i7-975, score = 7101, price = $970, value = 7.32

Phenom II x4 940 "Black", score = 3645, price = $156, value = 23.37
Phenom II x4 945 "Black", score = 3500, price = $150, value = 23.33
Phenom II x4 955 "Black", score = 3876, price = $160, value = 24.23
Phenom II x4 965 "Black", score = 4253, price = $180, value = 23.63

The 'score' is from cpubenchmark.net, the price was the lowest on newegg at the time, and the derived value is how many benchmark points you get per dollar.

The best of the bunch were the Q8300, the i5-750, and the Phenom II 955. Factoring in motherboard price and the Phenom II 955 is distinctly ahead, but both the i5 and the Q8300 are very nice looking as well. The i5-750 is of course the performance leader of the group, especially considering overclocking, but the Phenom is a "Black" so it is also highly overclockable.

But I think more interesting for many people would be a NOISE/PRICE ratio study. Whats the best performing CPU that can run with only a heatsink, with underclocking allowed. I am not equipped to do such a study, but I'm sure someone has the info necessary.

Re:Value for money vs FanboiGasms

[Rockoon]

My guess is that he had a game that was using RDTSC for timing on windows XP with Cool 'n' Quiet enabled. The XP scheduler loves to swap a process between cores (even when only one process/thread is using any CPU time), and the effect this has is that CnQ is constantly toying with the cpu frequencies of each processor, putting the RDTSC values out of sync with each other.

Prior to dual core CPU's, RDTSC was the accepted way to do timing (low latency, high resolution) Now its just foolish with all the stepping that CPU's do to save power and reduce heat. Intel's CPU's didnt have an issue here only because Intel cores couldn't be individually throttled. Now they can be with Intels turbo-whatever feature, so he would again experience the same issue with those same games if running XP.

Vista and 7 dont schedule the same way, instead trying to keep a process on the same core as much as possible to maximize cache efficiency.

Re:Value for money vs FanboiGasms

[Nalgas D. Lemur]

I'm even luckier and picked up a 750 back when they first came out and Micro Center was selling them for $150-160 for a few weeks. Ridiculously good deal at that price (supposedly they were $196 in lots of 1000 at the time, so they must've been losing a bunch of money selling them that cheap...), but it's slowly been going up since then, unfortunately, especially after the rest of the i5/i3 lines came out.

Re:Value for money vs FanboiGasms

[BikeHelmet]

To add further insult, money saved from AMD motherboards being cheaper (in particular SLI/xfire AMD boards are a good whack cheaper) will let you put money towards more storage, a SSD or a step up in CPU speed.

Or a better videocard.

With a highly limited budget, knocking $100 off the CPU might bump your GPU funds from $120 to $220 - which will give a HUGE framerate increase.

And all you lose is 10% performance for CPU-only tasks, and some overclockability.

Re:Value for money vs FanboiGasms

[cynyr]

Only reason I see to go Intel is the availability of mini-itx motherboards with PCI-E X16 slots. I have yet to find one for AM3. I found a AM2+ board, but it only has an 8X slot. boo.

ECC Support

[DAldredge]

And additional benefit of AMD processors is that they all support ECC RAM.

Re:ECC Support

[pslam]

This is a big reason I picked an AMD Phenom II over a Core i7 recently. To get ECC support from Intel, you need to buy a Xeon, at which point they charge you an extra $800-$1000 for the gates to be enabled. Screw that, I'll go with a chip 80% cheaper and 10% slower.

Re:ECC Support

[gullevek]

Yeah, but do you really need ECC RAM on a home machine?

Re:ECC Support

[drsmithy]

This is a big reason I picked an AMD Phenom II over a Core i7 recently. To get ECC support from Intel, you need to buy a Xeon, at which point they charge you an extra $800-$1000 for the gates to be enabled.

Huh ? You can buy a Xeon CPU for a few hundred bucks.

Re:ECC Support

[nabsltd]

To get ECC support from Intel, you need to buy a Xeon, at which point they charge you an extra $800-$1000 for the gates to be enabled.

Boy, when you make up numbers, you really reach deep into your ass, don't you?

Core i7-920 for $280 and the same-socket, indentical spec Xeon W3520 for $310.

The only issue might be that you need a motherboard that supports ECC, but $270 for this one isn't a lot more than the $200 or so you'd pay for a non-server board with equivalent build quality. Unless things have changed drastically since the last time I looked at AMD motherboards, not all of them support ECC, either.

Re:ECC Support

[DigiShaman]

You don't need ECC for home usage, but both personally and professionally, I highly recommend it.

With the stability of modern OSs such a OSX and Windows 7, people tend to leave their computer on 24/7. Eventually, a bit flip will take place. Question is, where? It might happen in an area which is about to be flushed out anyways. It could also happen where the kernel resides causing the OS to panic. Worse yet, having a bit flip could corrupt a file making data recovery that much more troublesome.

Re:ECC Support

[DigiShaman]

ECC support is disabled on all non-Xeon chips. There isn't a technical reason, it was purposeful market segmentation.

Per Intel FAQ

Does either the Intel® Core i7 processor or the Intel® Core i5 processor family support Error Correction Code (ECC) memory?
Neither family of desktop processors supports ECC memory. Typically ECC memory is used on servers and workstations rather than on desktop platforms. This is due to the price premium and extremely low likelihood of a data error occurring even on memory not utilizing ECC.

Knowing this before hand, that's why opted to build a new AMD based computer that has ECC enabled. The parts I used includes current prices below from Newegg.com and Crucial.com

Motherboard: Asus Crosshair III Formula = $199
CPU: AMD Phenom II X4 955 Black Edition = $159
RAM: Crucial 4GB kit (2GBx2) ECC DDR3-1333 (P/N: CT2KIT25672BA1339) = $149

Total (minus shipping) = $507

It's damn cheap for a fast performing ECC workstation.

Re:ECC Support

[chadruva]

Not only ECC, Virtualization Extensions are mostly guaranteed on all AMD cpus, I can't say the same for Intel ones.

Re:ECC Support

[Fjandr]

As memory increases in density and capacity, bit errors become vastly more common. The barriers to its use in home machines are manufacturers too cheap to put ECC support into chipsets and ignorance amongst even most tech-savvy consumers about the actual rate of correctable memory errors.

Re:ECC Support

[nabsltd]

ECC support is disabled on all non-Xeon chips. There isn't a technical reason, it was purposeful market segmentation.

Agreed, but the price difference isn't really that much for ECC with Intel.

I'm not a fanboi of either company, I just tend to buy the best price/performance at the time. I have a pair of dual-socket Opteron servers that were by far the best deal for the performance at the time, plus they use far less electricity than the Xeons of the same era.

But, I'm replacing them with dual-socket Xeon (Westmere) machines because it's by far the cheaper way to go for the same performance, and Intel has also leapfrogged AMD in lower electricity use.

It's damn cheap for a fast performing ECC workstation.

Right now, motherboard/CPU combos are just about perfectly priced so as to match their performance. So, your $360 parts are just about 75% the performance of $480 worth of Intel gear (Passmark 3931 vs. 5850 for a Core i7-930). Your ability to use ECC is a bonus, though, and I suspect that as standard memory grows to 8-16GB, Intel will have to re-think their ECC support stance.

The reason I'd still buy Intel myself for a desktop is that I can get the $280 CPU for $200 at a local store, which makes it only a 11% price increase for 48% more performance.

Re:ECC Support

[kueball]

I second this. ECC Support is primary reason when I use AMD desktop processors. For something like an Asterisk server, ECC is great peace of mind. I don't need a fast processor, just something will be reliably run for great lengths of time. Remember this DRAM error report based on Google's servers? Makes sense to use AMD desktop processors when you don't need a real server.

Re:ECC Support

[foxalopex]

Er, actually even the cheapest AMD motherboards support ECC. ie Asus M4A785TD-V EVO which on NewEgg lists for $114 and even has a built in video card if you're going to use it as a workstation system instead of a gaming system. It's one of the reasons why I pick AMD for workstations and server whitebox builds... They're hard to beat. :)

Re:ECC Support

[pslam]

Sure, if you pick a suitably low-end Xeon. Not sure why I'm bothering to reply to that.

All AM2/AM3 AMD motherboards I know of support ECC. In fact, all the Intel motherboards too - it's just that the ECC block is purposefully disabled in the Core i7 for no reason other than market segmentation.

Re:ECC Support

[BikeHelmet]

Wait - so your defense is Intel only costs $170 more for ECC support?

(High end AMD boards are going for $150 or less - not $200)

http://www.newegg.com/product/product.aspx?Item=N82E16813131631 (cheaper if you know where to look)

Maybe I'm a cheapskate, but I'd rather buy an AMD 6-core for $200, and save $280 ($110+$170) across all components.

Re:ECC Support

[Anonymous Coward]

Nice try. Official Intel specifications for Core i7-920 http://ark.intel.com/Product.aspx?id=37147

ECC Memory Supported: NO

Re:ECC Support

[sjames]

The GP did exaggerate the costs, but still, your more realistic prices show a $100 (~20%) increase in price to add ECC support on a system on the lower end of the speed range.

BIOS Update....

[Mark19960]

Does anyone think for a minute they will update the BIOS on a board when they can sell you a new one?

Re:BIOS Update....

[Anonymous Coward]

BIOS updates come out all the time.

Re:BIOS Update....

[DDLKermit007]

Yes actually...I've worked with so many boards that were made for AM2 that were made long before Phenom came out that work phenomenally with Phenom chips after a quick bios update. Now if your talking a prebuilt HP special POS, well that's your own fault.

Re:BIOS Update....

[Anonymous Coward]

Yes

Re:BIOS Update....

[D J Horn]

Yes. This is far from the first time a new CPU has been supported on older boards by updating BIOS.

Re:BIOS Update....

[Yaa 101]

ASUS/AMD certainly does, they have the most friendly system to upgrade your BIOS, just burn a firmware image on an empty CD, restart and let the computer startup from that cd and all goes automatically.
Don't forget to make a backup of your current bios and burn that on a separate CD.

Re:BIOS Update....

[Anonymous Coward]

My motherboard has a BIOS update for the hexacore phenoms... (Gigabyte GA-770TA-UD3)

Re:BIOS Update....

[luckymutt]

Does anyone think for a minute they will update the BIOS on a board when they can sell you a new one?

Yes, for example...

Serioulsy ...

[dnamaners]

There are more than a few things that AMD besides gaming and over clocking (Intel strong points) that make an AMD a good choice. I don't want to start holy war here but there is not much real gap here 10-5% in my tests at best. The price * power use thing shows AMD is a good choice in many places. Price alone makes me deploy more than a few AMD clusters. Don't just look at the max value on the "speedometer" to see how good a car is, we mostly drive at the speed limit. Take from it what you will.

Re:Serioulsy ...

[Anonymous Coward]

There is no speed limit in computing... and especially not in clusters. AMD competes in no arena intel has a presence. Its been very lop sided ever since the i7s came out, and it looks to stay that way for at least another year.

Re:Serioulsy ...

[Moheeheeko]

enjoy testing unstable INTEL chips then. The first gen of any intel product is how they stress test, all they do before they go on market is get one stable in the lab.

Re:Serioulsy ...

[Dishevel]

Come on though. Are not most of us guilty of testing Intel Chips?

Re:Serioulsy ...

[Gr8Apes]

So where's that 8P Intel PC? Granted, it's less of an issue now that there are 6 core CPUs out there, but let's not forget that Intel still can't effectively put one together.

Let's look at that Top 500 list. The top 2 are AMD systems, and that the 3 Intel systems are in the bottom 5.

If I want to browse the web and not heat my house at the same time, AMD really does offer the better chip, and cheaper by far too.

If I want to play that FPS and have an extra frame or two, then the Intel chip is a winner. If I want to participate in something like Folding@Home, then Intel is a winner if power and heat aren't considered. If I'm rendering video, Intel might be a winner, depends on whether heat matters.

Very few things even most on /. would utilize a computer for will only see an Intel advantage maybe 1% of the time.

After all - does using an AMD or Intel chip make any difference rendering /.?

Re:Serioulsy ...

[mwvdlee]

Warning! Car analogy comming up.

Very few things even most on /. would utilize a computer for will only see an Intel advantage maybe 1% of the time.

After all - does using an AMD or Intel chip make any difference rendering /.?

My cars spends some 23 hours a day going 0mph. Yet I'm sure glad it can go over a 100mph when I need it.

Re:Serioulsy ...

[shallot]

Warning! Car analogy comming up.

Very few things even most on /. would utilize a computer for will only see an Intel advantage maybe 1% of the time.

After all - does using an AMD or Intel chip make any difference rendering /.?

My cars spends some 23 hours a day going 0mph. Yet I'm sure glad it can go over a 100mph when I need it.

Your car analogy is lousy, because my 2000 Toyota Yaris can and does sometimes go 160km/h, and in relative terms it cost me much less than my CPU. If I was buying a car comparable to my CPU, I would be driving an SUV twice the size. Hm, on the other hand, thanks for that analogy, it put things into perspective. Maybe next time I shop for CPUs it will save me some money :)

Re:Serioulsy ...

[Lennie]

And AMD does just fine at 100mph, just is it doesn't do 200mph like the Intel one.

Re:Serioulsy ...

[WaroDaBeast]

Mod parent up. The truth is, a lot of smaller cars are able to go 100 mph -- even a Twingo I can.

Re:Serioulsy ...

[asukasoryu]

I think the key is whether or not going 100mph cause the car to explode. Some cars are designed to perform well at 100mph while others barely have the capability.

Re:Serioulsy ...

[dfghjk]

What a load of crap. I'd be embarrassed to have my name associated with that post.

Re:Serioulsy ...

[kkwst2]

The key is the probability of a Toyota Yaris exploding while traveling at 100 mph? Wait, what were we talking about?

Re:Serioulsy ...

[DrgnDancer]

Your car analogy is faulty. Let's say you have two cars; an Intel and an AMD. The Intel car cost 50% more and has 25% poorer gas milage. Also it's air conditioner doesn't work as well. It can, however, go 10% faster. Let's say the max speed of the AMD car is 150mph, but the Intel car can hit 165mph. Either speed is well above the speed limit, but it is undoubtedly true that the Intel car is faster.

Which car is better? Unless you have a really good reason to need to go 165mph, I'd rather have the AMD car, honestly. Not for my race car obviously; but for most day to day travel needs the AMD car is cheaper to buy, cheaper to operate, more comfortable, and fast enough by a long margin.

Re:Serioulsy ...

[Endo13]

Cars have a lot more speed these days than they need, and a lot more perhaps than you personally assume. My '98 Honda Accord with a V6 hit 115-120MPH with no trouble at all, and had rock solid handling at that speed. I don't know how fast it could safely have gone, because I didn't trust myself to push it to the limit. (Oh, and it came with tires from the factory rated for 149MPH.) If a stock Accord from 12 years ago was safe at that speed, I think it's fair to assume that most cars on the road today can probably hit 100MPH easily and handle safely at that speed, given appropriate road conditions.

So yeah, it's a pretty good analogy. Most dual-core and better processors these days have all the power you need as a home user. For the first time ever, we're in a place where even most machines from 5 years ago are still perfectly viable for most tasks.

Re:Serioulsy ...

[Gr8Apes]

Honestly, the analogy should be more along the lines of I have an AMD car that can go 130mph, and an Intel car that can go 155mph. While the AMD car at 130mph is rock solid, imagine my surprise at 155mph when the Intel car's back end tends to lift up and the engine's running in the red.

If I wanted to go 230mph, I'd buy a Power car.

Re:Serioulsy ...

[thetoadwarrior]

The i7 is too expensive for me. I can play games well enough with AMD and a decent graphics card. Maybe it'll be outdated a bit sooner but it's much much cheaper. I'm still only playing games like TF2 and doing development. I don't really want to pay an boat load. PC gaming doesn't have enough big titles to justify the big costs these days.

Re:Serioulsy ...

[mwvdlee]

What I meant was that even if you rarely ever need the power, when you DO need it, it really is important.

You could buy a moped instead of a car for those 23 hours. It's a lot cheaper if it's just sitting there. But when you DO need to travel hundreds of miles, the speed DOES become imporant.

You could buy a CPU going the equivalent of 10MPH as it will be enough for 99% of the time, but for that other 1% you really do need more performance.

Re:Serioulsy ...

[Anonymous Coward]

So where's that 8P Intel PC? Granted, it's less of an issue now that there are 6 core CPUs out there, but let's not forget that Intel still can't effectively put one together.

Um... what planet are you from? On mine, Intel has been shipping Nehalem-EX for a while now. 8 cores on one die, with glueless support for up to 8 CPU sockets on a board.

Let's look at that Top 500 list. The top 2 are AMD systems, and that the 3 Intel systems are in the bottom 5.

ROFL. You conveniently didn't link to http://www.top500.org/charts/list/34/procfam , which shows that Intel thoroughly dominates the Top 500 supercomputer list. There's a lot more than just 3 Intel based computers in the list!

Yes, the top 2 happen to be AMD; it's more or less meaningless. Massively parallel supercomputers are more about the interconnect tying all the processors together than the processors themselves. Did you notice that a ton of the machines in the Top500 are Blue Gene? IBM Blue Gene supercomputers are one of the few left that use a custom designed processor instead of a commodity CPU. That custom CPU isn't anything special, though: it's a PowerPC core derived not from IBM's firebreathing POWERn line, but instead from their old PPC440 low power embedded control CPU. You would not want to use one of the CPU nodes from a Blue Gene to run even a web browser. But Blue Gene gangs metric fucktons of them together and the result is a competitive supercomputer.

The top 2 machines in the list are at the top because the organizations which commissioned them (United States DOE and Oak Ridge National Laboratory) have very large budgets to buy larger supercomputers than anyone else, so they can buy more nodes. Look at the processor counts on each.

(Also note that #2 isn't a pure Opteron machine, it's an Opteron/PowerCell hybrid. I'd guess the Opterons are for communications and the Cell processors are what provide the FP grunt.)

If I want to browse the web and not heat my house at the same time, AMD really does offer the better chip, and cheaper by far too.

Once again, what planet are you on? On mine, for the last several years, virtually every review comparing an AMD notebook to an Intel notebook has found that roughly equivalent Intel notebooks outperform AMD's and also have much better battery life. You need to pull your head out of the sand, it isn't the P4 vs. Athlon days any more.

If I want to play that FPS and have an extra frame or two, then the Intel chip is a winner. If I want to participate in something like Folding@Home, then Intel is a winner if power and heat aren't considered. If I'm rendering video, Intel might be a winner, depends on whether heat matters.

Available evidence suggests that Intel is far, far ahead in terms of power efficiency at full load. In other words, if you do Folding @ Home, and divide performance by the watts used, Intel has a much better perf/W ratio than AMD. Since AMD and Intel ship similar wattage rating CPUs, that translates into an absolute performance advantage for Intel too.

Once again, Intel has stuff just a bit better than the P4 now. The memes you blindly parrot are a bit out of date.

Missed the Mark.

[Anonymous Coward]

Too bad the flagship product doesn't out perform even an Intel Core i5 750....

As an AMD fanboi, I can really say I am disappointed in this missed opportunity.

Depends what you do.

[FatSean]

I have four cores. I run an IDE and an AppServer at all times, which uses up at least two cores. Then there is my bit-torrent app and...

Seems like you can easily use all those cores.

Re:Depends what you do.

[forkazoo]

I have four cores. I run an IDE and an AppServer at all times, which uses up at least two cores. Then there is my bit-torrent app and...

Seems like you can easily use all those cores.

Stop using Azureus for your bit torrent client, and downloading a file will no longer require a fill core with of CPU time.

Or, if you are using a sane torrent client, what the hell kind of internet connection are you using that you are still CPU bound on a file transfer?!?!

Re:Depends what you do.

[Anonymous Coward]

what the hell kind of internet connection are you using that you are still CPU bound on a file transfer?!?!

  Analog 10G fiber with software-based modem firmware?

Re:Depends what you do.

[nabsltd]

Stop using Azureus for your bit torrent client, and downloading a file will no longer require a fill core with of CPU time.

Over the past month, my torrent VM (running Azureus) has averaged 603MHz CPU usage.

I suspect it's that high because of the fairly high average transfer rate (1.5MB/s disk, 984KB/s network), plus the usual suspects of other background programs (anti-virus, etc.).

Re:Depends what you do.

[Anonymous Coward]

I can't imagine the bittorrent app using very much CPU power at all. I use rtorrent on an ancient AMD Thunderbird 1100Mhz computer running Debian.

At any given time with 20-25 torrents running CPU usage hovers around 3-4 percent and most of that is probably the GNU Screen session. Even hashing files at the limit of the hard drive only uses 35% CPU.

Re:Depends what you do.

[GreatBunzinni]

And if you rely on the make utility to build your project then you can take advantage of how many cores you wish by passing the --jobs flag, which is quite able to use all available cores quite nicely.

Re:Depends what you do.

[Anonymous Coward]

Uh, I can download at 2MB/s using uTorrent and barely top 1-2% CPU on my aging 2.1GHz Athlon XP. Maybe you should profile the machine to see if it's the VM HDD driver causing the slowness.

Re:Depends what you do.

[Pinchiukas]

If a bittorrent client uses up an entire core, you're doing something wrong.

Re:Depends what you do.

[Daengbo]

Apparently, four cores isn't enough to render Slashdot.

Re:Depends what you do.

[Daengbo]

RTorrent is amazingly efficient. I think you could probably run an rtorrent server on a programmable calculator if you wanted to.

Re:Depends what you do.

[BikeHelmet]

Seems like you can easily use all those cores.

Modern games can use 4 cores fairly well. Sometimes well enough to impact UI responsiveness.

Therefore: Six!

AMD

[Antisyzygy]

AMD basically has a processor that has a high performance/price ratio for any budget. I will be loyal to AMD for quite some time. Im seriously considering tattooing AMD on myself.

Re:AMD

[Anonymous Coward]

Bah, you're nothing new. Want to see my tattoo?

love to see one of those....

[SethJohnson]

older Socket AM2+ and AM3 motherboards with only a BIOS update.

Isn't that an oxymoron?!? A BIOS update on and older AM2+ mobo?

Seth

Re:love to see one of those....

[Yaa 101]

Dunno but you are considered a normal moron if you don't.

Re:love to see one of those....

[Vectormatic]

how so? my AM2+ board had the needed bios update available before the new chips even hit store shelves

But I'm going to undercut AMD

[mjwx]

But I'm going to undercut AMD with seven cores.

Re:But I'm going to undercut AMD

[gandhi_2]

Or 7 cores and an aloe strip.

I have been waiting for this day!

[bigredradio]

includes a new Turbo capability that can opportunistically raise the clock speed

Does this mean I can get my turbo button back on my computer?

Re:I have been waiting for this day!

[rrohbeck]

And the 7 segment clock speed display!

Re:I have been waiting for this day!

[Anonymous Coward]

hear hear... i want that damned button back too. and the lcd speed display to show the difference beteween 66 and 33... lol

But wait, there's more!!!

[PPalmgren]

If you act now, we'll throw in this brand new nose hair trimmer for FREE!!!! Get rid of those pesky nose hairs with our patented root-ripper design that leaves your nose feeling clean for months. Also, if you order within 24 hours we'll include a 29 foot garden hose!

Certain restrictions apply. $19.95 shipping & handling, delivery within continental US and Canada only. ...I love AMD, but COME ON man. Make it a LITTLE less obvious.

Transcoding is not common

[tepples]

Try transcoding some video one time kiddo.

That's not necessarily the best example. Most people do not produce original high-definition video, and the United States (home of Slashdot editors) has made it illegal to break DRM in order to transcode major-label video. Besides, GPU-accelerated transcoding can use more cores than a CPU-only transcoder ever can because it is embarrassingly SIMD.

Re:Transcoding is not common

[Anonymous Coward]

Most people are just doing things like email and web browsing. Give them a 5 year old single core processor and they'll be fine - hence the popularity of netbooks and Atom processors.

If you're really taxing the processor, odds are that you're doing something that could use more cores.

Re:Transcoding is not common

[Lennie]

On the other hand Youtube is filled with people doing HD-video, so I guess it's not such a small fraction of the users anymore, it's very close to mainstream actually.

Re:Transcoding is not common

[tepples]

On the other hand Youtube is filled with people doing HD-video, so I guess it's not such a small fraction of the users anymore

People who upload to YouTube are a small fraction of people who visit YouTube, and as I understand it, people who upload homemade videos produced in a format bigger than 480p are a small fraction of that.

Re:ill pit my i7-920 against any AMD 6 core

[DDLKermit007]

And guaranteed the AMD systems you have crap out were probably using junk boards & PSUs. Unless the chip fizzled, the CPU choice had nothing to do with it. Plus really? Your going to talk about prototypes vs whats on the market? Both have stuff that's way ahead of whats on the market. That's why the call it's a (spell it with me now) P-R-O-T-O-T-Y-P-E. AMD has it's strengths. It's why Intel's x64 architecture is the same as AMD's. They got their shit together first, and made it work.

Re:ill pit my i7-920 against any AMD 6 core

[Com2Kid]

and I've had AMD systems last as long as your Intel systems.

Quite frankly the quality of power supply, memory, and motherboard, are more important than the CPU. Both Intel and AMD make reliable CPUs, and you can find cut rate MoBos sold for either, and many low cost PCs using AMD CPUs also come with a low cost supply.

Apps that sleep

[tepples]

I have four cores. I run an IDE and an AppServer at all times, which uses up at least two cores.

The app server uses a core only when someone is using the app. And what does the IDE do for you between keystrokes?

Then there is my bit-torrent app

Network bound, probably sleeping much of the time. Or what am I missing?

Re:Apps that sleep

[crazycheetah]

I've noticed some IDEs are annoyingly more CPU hogging than a lot of other applications I run. Code::Blocks, for example, seems to start eating up CPU usage after being kept open with not even a very large amount of files open. It gets worse with more files open, but I've seen it happen with relatively small projects. I've seen the same with other IDEs, also--I've just been using Code::Blocks with one project I've got going lately, so it's most fresh in my mind. I'm not quite sure what it's doing, but some of them seem to like eating up CPU even when the window is off on another virtual desktop. That seems to affect me more on Linux, though. On Windows, I usually close what I'm not using, because it drives me nuts having so many windows open with only one desktop area to work with. As such, I'm not sure how the comparison goes...

Re:Apps that sleep

[Anonymous Coward]

The app server uses a core only when someone is using the app. And what does the IDE do for you between keystrokes?

Apparently, you have never used Eclipse.

Re:Apps that sleep

[sznupi]

It's a sad world we're living in if IDEs, of all things, are poorly written...

Re:Apps that sleep

[grimsweep]

Modern IDEs do quite a bit these days beyond just organizing files and giving you a color coded editor. Between parsing your code as you write it, context-sensitive auto-complete, and dynamic recompilation of the files you're changing, there's plenty of things for it to do to try and make your life easier as a developer.

Re:Apps that sleep

[sznupi]

Though that doesn't explain eating the CPU while the app just sits there, as the parent poster described.

Re:Apps that sleep

[Anonymous Coward]

My IDE is precompiling code in the background to check the entire assembly for errors.

Re:Apps that sleep

[Anonymous Coward]

Maybe if Eclipse did more on Windows than just launch and exit without any errors or messages saying why it does so, it actually would be of serious consideration.

Re:Apps that sleep

[7-Vodka]

I take it you've never heard of emacs.

It does everything. EVERYTHING.

Re:Apps that sleep

[ifrag]

Network bound, probably sleeping much of the time. Or what am I missing?

You are missing the virus scanner double scanning every block, once as it comes through the pipe, and again as it writes to disk.

And for those unfortunate users running McAfee, that's probably a core worth of work.

Re:Apps that sleep

[tomthegeek]

Does it run Linux?

Re:Apps that sleep

[FatSean]

I have my IDE set up to auto-publish every time it re-builds....and it builds every time I touch a file...which happens quite often.

I've found that my older multi-CPU systems age more gracefully...the UI doesn't get as sluggish despite the OS getting heavier and heavier.

Also has nice overclocking prospects

[alvinrod]

Anandtech managed to get a stable 4.0 GHz overclock with air cooling. It makes an already great deal all that much better in my opinion.

Re:Also has nice overclocking prospects

[nabsltd]

Anandtech managed to get a stable 4.0 GHz overclock with air cooling. It makes an already great deal all that much better in my opinion.

How is a $299 6 core/6 thread chip at 4GHz a better deal than a $199 4 core/8 thread chip that can also be overclocked on air to the same speed, and benchmarks far faster at that point?

Re:Also has nice overclocking prospects

[Anonymous Coward]

You are getting ripped off on that AMD CPU, NewEgg has it for $204
http://www.newegg.com/Product/Product.aspx?Item=N82E16819103851

Then again their i7 930 is $288
http://www.newegg.com/Product/Product.aspx?Item=N82E16819115225

Re:Also has nice overclocking prospects

[Anonymous Coward]

OK I'll bite. More MIPS per watt on the phenom, twice the L2 cache per core - also I ain't gonna overclock shit. That's all I need is some unstable crap in my rig. I'd rather get the chip designed for the higher clock rate.

Re:Also has nice overclocking prospects

[MarkVVV]

Because HyperThreading != 2*(cpu.getNumberOfCores())

Re:Also has nice overclocking prospects

[nabsltd]

You are getting ripped off on that AMD CPU, NewEgg has it for $204 http://www.newegg.com/Product/Product.aspx?Item=N82E16819103851

That isn't the unlocked 1090T you are referencing, this is.

Re:Also has nice overclocking prospects

[nabsltd]

True, but all the benchmarks show that the i7-920 (and the i7-930 and i7-860) end up doing better on multi-threaded tasks than the 6-core Phenoms. So, it works out that 8 threads really is better than 6 cores in most cases.

As a matter of fact, the 4-core i5-750 (2.66GHz) can beat the 6-core 1090T (3.2GHz) in some benchmarks, and isn't very far behind in the rest. If you overclock the i5-750 to 3.2GHz (which is easy to do even with the stock cooler), it far surpasses the best that AMD offers.

On the other hand, the 890FX chipset is better than anything Intel has out...42 lanes of PCI Express 2.0 is pretty awesome.

Re:Also has nice overclocking prospects

[nabsltd]

More MIPS per watt on the phenom

Nope.

Both the i5-750 and i7-920 use considerably less power at load and both benchmark as good as (or better than) the 1090T.

Re:Also has nice overclocking prospects

[BikeHelmet]

The $195 6 core/6 thread chip overclocks to 4ghz too, according to xtremesystems.org.

That puts value quite firmly on AMD's side. Expect price drops for Intel, soon.

Re:Also has nice overclocking prospects

[nabsltd]

The $195 6 core/6 thread chip overclocks to 4ghz too, according to xtremesystems.org.

That puts value quite firmly on AMD's side. Expect price drops for Intel, soon.

Not when the $200 i5-750 can beat either chip in many benchmarks at slower base clock speeds, and can also easily be overclocked to fast enough (3.4-3.8GHz) to beat out the Phenom X6 chips at 4GHz.

If the X6 had solidly bested the i7's and i5's in all (or even most) benchmarks, then I'd say Intel would be dropping prices, but since it actually loses in most, I don't see Intel as being very worried. They might release a non-"Extreme" 6-core i7 at a $400 or so price point to really squash AMD, but I don't see price drops happening on existing Intel chips because of this AMD release.

Re:Also has nice overclocking prospects

[BikeHelmet]

Who needs benchmarks? This is about marketing, and right now AMD has more cores. They must be better!

P.S. Some of those same benchmarks used to tell you that a fast dual-core beats a quad-core. Try running multiple benchmarks at once, and suddenly it's a different story. Many people went with quads despite them being worse for gaming, because of them being "future-proof" and "fast-enough". Given more cores for the same price(and a negligible performance difference), many people will go for more cores.

And yes, it is negligible. Nobody cares if the i5-750 is 20% faster at X task, when it's only 10 vs 12 seconds.

Re:Also has nice overclocking prospects

[nabsltd]

And yes, it is negligible. Nobody cares if the i5-750 is 20% faster at X task, when it's only 10 vs 12 seconds.

Gamers seriously care about 20% faster frame rates, especially at the edge cases. The difference between 28fps and 34fps is sometimes the difference between playable and unplayable.

One other advantage to the Intel chips is TurboBoost, which works well compared to the new AMD Turbo Core. You end up with a "have your cake and eat it to" config where you have 4 cores to handle multiple tasks, but a single task gets as much as 666MHz more speed.

I sound like an Intel fanboi because right now I am, since their chips are by far the best price/performance. Previously AMD had that crown, but they let it slip away. Even while Intel was taking over the desktop, AMD was still the best server chip, but even that is now in jeopardy. If Intel ever starts selling slower-clocked, less expensive 6-core Xeons, AMD is done in the server market, too.

Re:Also has nice overclocking prospects

[BikeHelmet]

Gamers seriously care about 20% faster frame rates, especially at the edge cases. The difference between 28fps and 34fps is sometimes the difference between playable and unplayable.

True. But quite often these benchmarks are done at low quality. Few gamers care about 200fps vs 240fps.

And as soon as you raise the quality, so the GPU is the bottleneck, suddenly the difference drops below 20%. Benchmarks like this suggest 8-16% is more common than 20+%

There are a few games that truly benefit from fewer, faster cores. But it's not like we're comparing 28->34fps. For most, any modern CPU and GPU will manage 60+ fps. Take a look at L4D2 - even Dragon Age is way above 100fps.

Given identical prices, I think a lot of people will go for the extra cores. Whether that's a smart choice is determined by the buyer, because only the buyer knows how he will use it.

Last time around many people picked up Q6600's rather than E6600's, because even though it wasn't as fast in games, more cores meant they'd be set for when games used more cores. The odd time they ran multiple multi-core programs at once, they'd also be set.

I'm guessing many will go for more cores this time around, too.

Transcoding is common

[Anonymous Coward]

Way to belittle an excellent example of the usefulness of multiple cores.

Original HD production? Illegal to break DRM? So do you think it's only the occasional cinematographer or pirate who uses software based on FFmpeg and X.264 to transcode vids for HTPC or iPod? Because neither currently targets OpenCL or CUDA.

Re:Transcoding is common

[tepples]

So do you think it's only the occasional cinematographer or pirate who uses software based on FFmpeg and X.264

AMD and Slashdot are located in the United States. In the United States, the encoding process that x264 uses is subject to royalty-bearing patents, and royalty-bearing patent licenses are incompatible with the copyright licenses for FFmpeg and x264. So anyone using FFmpeg and x264 is a pirate, except possibly in one or two corner cases that someone is likely to chime in to clarify.

to transcode vids for HTPC or iPod?

Before this discussion succumbs to Layne's Law, let us clarify something: Are we referring to major-label video or homemade video?

Cores is the new MHz

[pankajmay]

I have been wondering for quite some time - do regular joe consumers really need all those cores? OR is everyone buying into the marketing hype of processor manufacturers without thinking whether we would actually need that many cores??

First of all, any computer organization text will inform you that as the number of cores increase - scheduling amongst those cores becomes an exponentially costly issue in itself. This scheduling/load balancing of course has to be ultra low latency to maintain a reasonable throughput.
Not to mention the fact, that on software side managing threading and choosing instructions to parallelize is a big headache. Many decent programmers cannot get it right so that in itself defeats the presence of different cores.

Secondly - unless you are continuously doing protein folding, calculating eigen values of huge matrices, or are acting as a node for traffic in your part of the world -- most people's processor cores will spend a majority of their time idling or spin-lock. Is it any surprise then that both Intel and AMD are advertising technologies to power down three cores, boosting the power for the other three?? Simply because most end-users will rarely utilize all six of their cores simultaneously. Yes, that is even true no matter if you are doing heavy video transcoding or running multiple servers, and playing games simultaneously - you will still leave your cores without any task simply because unless the bandwidth of the memory bus catches up, your cores will be waiting for data to process.
This is why Intel's i-series architecture is superior to AMDs and likely the fact their processors cost more, because they have addressed the memory bus issue.

You have to realize your computer acts like a chain and it is only as fast as its weakest link.

I have been advising people that any new dual or quad processor will suffice - they should instead spend that extra money on buying a better motherboard, speedier RAM, and of course high-speed HDD.

Trust me when I say that just that approach above will yield systems that are actually much faster than coupling an i7/Mega-core behemoth with an old hard-disk and crappy RAM.

It is an altogether different matter that computers are already so speedy that most users cannot for the love of God discern between the speeds of any recent dual-core and a top-of-the-line processor - and it is not their fault -- the advantages now we are talking about are incremental. The power is present but cannot be harnessed. So any gloating is moot.

Re:Cores is the new MHz

[Anonymous Coward]

Don't know about average joes, but for work, a quad core drops compile jobs from 17 mins to 3.
For home use, I usually pull 40-80% load on 4 cores, so I would say I get good use out it.

Re:Cores is the new MHz

[thoughtsatthemoment]

For homer users dual core is sufficient, but for professionals more cores can make a significant difference like: 1. A bunch of VMs; 2. Visual Studio with the /MP option. 3. Graphics rendering. With more cores available, software developers are also encouraged to use more parallel algorithms. As to the MHZs, I'd give it to the nanometers. I believe they could have gone straight to 11nm or something, but they just have to make each step feel like a breakthrough. Maybe that's true, but I just got this cynic feeling that they are milking us.

Re:Cores is the new MHz

[pankajmay]

But see that is the issue here -- every one is throwing around numbers convinced that it is solely the number of cores that is causing the improvement.

What part of that 14 minutes improvement -- is actually the result of a speedy RAM, improvements to the processor-memory bus interface? If we strip it down to its bare, I have a high confidence that the cores added only a fraction of improvement.
Also factor in

Data flowing through your processor is not a water stream -- you provide 6 pumps and viola! You pump out more water.

Data has to be managed and marshalled through the cores - that takes what? You guessed it - more processing power. One of your cores will be exclusively dealing in managing data flow through other cores -- so even though you are under the illusion that you are putting a healthy load -- the fact of the matter is that in absolute terms, you did not gain any great leap in your computing power.

Re:Cores is the new MHz

[Anonymous Coward]

"The power is present but cannot be harnessed"

g++ would like to have a word with you

Re:Cores is the new MHz

[dbIII]

"Regular Joe" users play with home videos and photo editing - two embarrassingly parallel situations right there where you can always find something to do with as many CPU cores as you can get.
Four is a relative large number though.

Re:Cores is the new MHz

[pankajmay]

Again, I suspect the improvements noticed are not just due to the cores, but also improvements in RAM, memory bus, processor-memory interface, and speedy hard drive.

Look I am not against multiple cores - dual or even quad cores are actually beneficial since both the hardware and software can effectively utilize them without a lot of complications. But beyond quad core, I am not convinced.

The only one way I can argue for the need of more than quad cores is to allow for a future unknown resource-intensive application. But for most present day applications, even for Graphics rendering (which by the way - has its own dedicated processor!), spending money on more cores is purely money down the drain.

Parallelization of algorithms is actually not a trivial task at all.
Many programmers discover this fact after innocuously assuming that they can just do it by reading a book or utilizing a library in their otherwise procedural programming paradigm. It is only after they have banged their heads, that many realize that parallelization actually entails a completely different thinking about programming. And these are the lucky programmers.

The worst are the ones who naiively implement utilizing a library and happily go about blissfully ignorant about the issues involved.

Compound this with the fact that most common languages used today (C, C++ etc.) really didn't think of parallelization as an integrated aspect. It is mostly tacked on as an after thought. There is work being done in developing effective strategies for parallel algorithms -- but this work is still very pedantic in nature and you will need a whole new generation of software developers to catch up to that.

What bothers me is not that Intel or AMD are coming out with ever more cores... that is simply an artifact of technological progress. You cannot expect that to stop just because the rest of the world has to catch up.
But the fact that bugs me is that nagging suspicion that we are not really using all the computing power already at our disposal and we are being deluged with more. Simplistic metrics like my program runs faster, ergo more cores = faster is fallacious. It will only be valid as a metric if you change nothing on the computer except the number of cores.

But since processors are packaged with memory controllers, conducting such an experiment is only difficult for regular people. You never know whether it was the improvement in underlying memory interface that got you that extra speed or the tacking of actual cores.

Re:Cores is the new MHz

[Chirs]

"What part of that 14 minutes improvement -- is actually the result of a speedy RAM, improvements to the processor-memory bus interface? If we strip it down to its bare, I have a high confidence that the cores added only a fraction of improvement."

Totally depends on the task. Compiling is probably one of the best tasks to just throw more cores at since the required memory bandwidth isn't too bad. Besides, any of the chips with on-board memory controllers have pretty decent bandwidth to start with.

Re:Cores is the new MHz

[QuantumRiff]

I used to have a lab setup of a half dozen machines. Last week, I built an Active directory domain, SQL server, SCCM server, and a few clients, all running on a dual xeon workstation (dual quad cores) and 8 GB of ram. That one workstation costs a little bit more than the very nice thinkpad notebook I have to do normal work tasks. I am using a fraction of the power, teeny bit of the cost, and i have my desk area back. Sure, a few more disks would be nice, but its darn nice to just spin up another VM (or restore to an older snapshot) and not have to mess with driver issues.

Re:Cores is the new MHz

[Korin43]

This is the only reason I'm interested in multi-core CPUs. "make" can be run in parallel quite well. Also, it's nice that when a program starts taking up 100% of a core, all of my other programs keep running like nothing happened.

Re:Cores is the new MHz

[Anonymous Coward]

You can rather easily profile the impact of multi-core on compile times given the common compiler option to select the number of cores used. Substantial portions of my compile times are CPU bound and scale almost linearly with core count. Even hyperthreading helps significantly. It's the difference between a 20 minute build and a 90 minute one to go from 1 to 8 (where 4 are "virtual" cores from hyperthreading). Moving from 1 moderate HDD and 2 fast ones in RAID 0 with a separate OS drive, and 4GB RAM with good timings to 12GB RAM with great timings, gave me a fraction of that (but it did virtually eliminate thrash-related delays for building multiple versions in parallel rather than serially, which I need to do more often than you'd think, and which the cores did not help with substantially).

Re:Cores is the new MHz

[pclminion]

What part of that 14 minutes improvement -- is actually the result of a speedy RAM, improvements to the processor-memory bus interface? If we strip it down to its bare, I have a high confidence that the cores added only a fraction of improvement.

Totally off base. First thing I tried when I got a new quad core dev box at work was to try the standard build with 1, 2, 3, and 4 cores. Imagine my total lack of surprise when the 2 core build ran in about half the time, the 3 core build in about a third the time, and the 4 core build in about a quarter of the time. How shocking.

No, it's not perfect, as in T(N) = T(1)/N, but it's certainly much stronger a correlation than "only a fraction."

Hell, I'd love to see it turned on for all the official build machines, but it's just a little bit scary, since I have seen it fail (as in the build driver crashing) about 1 in a 100 times. But for daily dev work, it's awesome.

Re:Cores is the new MHz

[thoughtsatthemoment]

The only one way I can argue for the need of more than quad cores is to allow for a future unknown resource-intensive application. But for most present day applications, even for Graphics rendering (which by the way - has its own dedicated processor!), spending money on more cores is purely money down the drain. Parallelization of algorithms is actually not a trivial task at all.

Let's take Visual Studio as an example. It can compile each project in a separate thread. In a large solution, each additional core will cut the total compiling time in half. If it takes an hour to compile on a single core, a dual 8-core xeon will take about 4 minutes (or 2 if hyper threading is really that good). I'd call that money well spent.

And the parallelization here is actually quite simple as it is done on the project level. No complicated algorithms are needed.

Re:Cores is the new MHz

[thoughtsatthemoment]

And the parallelization here is actually quite simple as it is done on the project level.

It should be at file level.

Re:Cores is the new MHz

[pankajmay]

No, it's not perfect, as in T(N) = T(1)/N, but it's certainly much stronger a correlation than "only a fraction."
May be true. But do not forget "cum hoc ergo propter hoc" fallacy http://en.wikipedia.org/wiki/Correlation_does_not_imply_causation .

<quote><p>... but it's just a little bit scary, since I have seen it fail (as in the build driver crashing) about 1 in a 100 times. But for daily dev work, it's awesome.</p></quote>

And why does it fail about 1 in 100 times? It is mysteries like these that are the real cause of worry. For mission critical applications, not knowing the cause of failure is a deal breaker.
See my comment below about understanding parallelization.

Re:Cores is the new MHz

[MemoryDragon]

I think the need of those cores will come from code level not from the users. This will be a gradual but ongoing shift.
I think the biggest need will come from parallelizing the algorithmic functionalities in the data structures like sorting tree operations etc... Those are the easiest to tackle and I wonder why the big libraries have not yet started to do this.
(JVM, .Net etc...). I personally think due to the fear of the rising memory consumption.

The second step will be to parallelize more parts of a program, the hardest to tackle, and no functional languages are not the solution only part of it because they mostly lack things needed for big programs like namespaces etc...
A mixed approach will be what will happen, but unfortunately with that come immutable data structures and a rise in memory consumption.

The third step will be highly parallel operations which are already done which currently are offloaded to the gpu will be partially shifted back, when it makes sense, so there also will be an increase and the demand for the user will come.

But there is no free ride anymore of plastering a new processor in and your performance problems are gone (heck for years I tried to program an 5-6 year old machines to not even get into the habit of pushing something out of the door which does only well on current gen machines)

Re:Cores is the new MHz

[kcbnac]

I have been wondering for quite some time - do regular joe consumers really need all those cores? OR is everyone buying into the marketing hype of processor manufacturers without thinking whether we would actually need that many cores??

First of all, any computer organization text will inform you that as the number of cores increase - scheduling amongst those cores becomes an exponentially costly issue in itself. This scheduling/load balancing of course has to be ultra low latency to maintain a reasonable throughput.
Not to mention the fact, that on software side managing threading and choosing instructions to parallelize is a big headache. Many decent programmers cannot get it right so that in itself defeats the presence of different cores.

Is it any surprise then that both Intel and AMD are advertising technologies to power down three cores, boosting the power for the other three?

Nope, not at all.

This gives us the best of both worlds. When doing something massively parallel, we get 6 cores. (or 8, or whatever)

When we're doing something that is more linear (say the latest Grand Theft Auto or whatever game we enjoy) we get half the cores running faster. I don't know of any games that'll take advantage of more than 3 cores anyway - so I want it to run fast.

With this feature I don't *HAVE* to pick Fast Cores or Many Cores. I can have both. And it will optimize itself for me, nothing I have to toggle.

I love it.

Simply because most end-users will rarely utilize all six of their cores simultaneously. Yes, that is even true no matter if you are doing heavy video transcoding or running multiple servers, and playing games simultaneously - you will still leave your cores without any task simply because unless the bandwidth of the memory bus catches up, your cores will be waiting for data to process.
This is why Intel's i-series architecture is superior to AMDs and likely the fact their processors cost more, because they have addressed the memory bus issue.

You have to realize your computer acts like a chain and it is only as fast as its weakest link.

Agreed. These days the weakest link technologically are the programs we use (not always optimized to be speedy when they can just eat more CPU cycles), the storage mediums we use, and the connection to the internet for some folks.

I have been advising people that any new dual or quad processor will suffice - they should instead spend that extra money on buying a better motherboard, speedier RAM, and of course high-speed HDD.

Trust me when I say that just that approach above will yield systems that are actually much faster than coupling an i7/Mega-core behemoth with an old hard-disk and crappy RAM.

It is an altogether different matter that computers are already so speedy that most users cannot for the love of God discern between the speeds of any recent dual-core and a top-of-the-line processor - and it is not their fault -- the advantages now we are talking about are incremental. The power is present but cannot be harnessed. So any gloating is moot.

Agreed, spending more on CPU at some point isn't worth it. Make sure you've got enough for your apps (some want 2.5GHz on 1 or 2 cores, but can't thread out past that) and enough cores for your apps and a background thread or two, and that should be plenty for quite some time.

When I buy a car, I don't buy it to be "fast enough" for commuting. I buy it for the intention of commuting, but I want some extra margin, for safety. I want to have "Oh shit" power, not "Oh shit + 5" power. Also I want it to last. At this point the only things worth potentially upgrading in a new quad-core box would be storage devices if the price/GB comes down on SSDs. Otherwise I'll probably stay with my 1TB fast drive and 1.5TB storage drive in my quad-core (2.8) box with 8GB of RAM.

Re:Cores is the new MHz

[Anonymous Coward]

Oh, please!
Memory, bus or hard disk speeds have improved fractionally in the last decade, as top CPU speed. Graphics performance has, but guess what? It's due to more parallelism in the graphics pipeline.

So more cores = faster may not be true for many common tasks, but it is, in a number of important tasks.

If for your workload that is not true, you may keep using your current computer, because a new one will probably not get much faster at it anyway.

Re:Cores is the new MHz

[the_B0fh]

time make -j1

make clean

time make -j2

make clean

time make -j4

etc

Re:Cores is the new MHz

[jvkjvk]

No, it's not perfect, as in T(N) = T(1)/N, but it's certainly much stronger a correlation than "only a fraction."

May be true. But do not forget "cum hoc ergo propter hoc" fallacy http://en.wikipedia.org/wiki/Correlation_does_not_imply_causation .

Umm. Yeah. Just what else do you suppose is causing the speedup in compile times, based on the description of the experiment he gave?

Correlation may not imply causation, but it is a good starting point to look for it.

I get that you don't particularly care for multiple cores, but at least keep your complaints in the realm of the reasonable.

In this particular case quoting latin makes you appear stupid (even though you are probably not) since it is well proven that more cores cause compile times to be lower for many types of compliation (i.e., the "embarrasingly parallel" ones).

Regards.

Re:Cores is the new MHz

[zeropointburn]

This is a bit of a rant, and I don't really mean to direct it at you, so I apologize if it sounds that way.

  Are you suggesting that chipmakers should apply their shiniest new memory controller technology to their 3+ year old processing technology and try to sell that instead of the latest and greatest? To be blunt, the technical community would cry foul. We know better, and would not be satisfied with anything other than their absolute best (at the top of the line). Neither side would dare make that decision, since the competition would simply release all their best tech in one package and annihilate them in every benchmark known, possibly aside from pure memory throughput.

  Each use case is a little different. Even generalizations for types of data processing (graphics rendering, compiling, transcoding, etc.) are not as homogenous as they appear. The best solution is to decide what you can afford to spend, then get the best performance and reliability that you can get for that money. Sometimes that means going for 6 cores instead of 2 or 4. Sometimes that means Intel instead of AMD. Sometimes that means an old, cheap processor and a really nice motherboard and power supply.
  To slashdot readers in general: I don't know where you get your average joe examples, but the average joe users I know have XP or Vista (eating 1 and 2 cores respectively [yes, I'm in the vista-hater fanboi camp]), heavy-duty antivirus (eating a core and half the hd bandwidth all by itself), and they have purchased dvd backup software and cd ripping software to go with their (or their kids') ipod. They like to watch movies on hulu or netflix, they like to transcode whole dvd's and they get pissed if it takes more than half an hour. They also use big, bloated MS Office to take simple notes. Their kids play flash games, rip cd's, chat, and pretend to do their homework all at the same time. My personal examples are all of people that would benefit more from a 6-core processor than they would from a x% improvement in memory bandwidth. Certainly they would notice a difference in either case, so the best choice would be a chip that has both and that's what AMD is offering. If they had more money than they could possibly spend, sure I'd set them up with a kilowatt PSU and a dual-quad Intel box and a SSD raid and four top of the line graphics cards and 4 big widescreen flatscreen monitors. Of course at that point they could have bought nice AMD systems for every kid on their block and a very nice machine for themselves at the same price.

Re:Cores is the new MHz

[BikeHelmet]

First of all, any computer organization text will inform you that as the number of cores increase - scheduling amongst those cores becomes an exponentially costly issue in itself. This scheduling/load balancing of course has to be ultra low latency to maintain a reasonable throughput.
Not to mention the fact, that on software side managing threading and choosing instructions to parallelize is a big headache. Many decent programmers cannot get it right so that in itself defeats the presence of different cores.

That's why we need to rethink our current architectures. I once heard that when it comes to games, most games would be faster running off a single core rather than attempting to split things to two cores. With early dual-core systems, this seemed quite true.

But if we simplified the cores a bit, we'd lose 40% performance each... and be able to pack in 20x as many. Just think - 120 cores available. Anything that you can run in parallel has a core available. That's going to be a lot easier to code for, since new assumptions can be made.

Intel has been working on it for years. Look up their 80-core prototypes.

I guarantee that with so many cores available, we can figure out ways to split algorithms to multiple cores and gain speed. It won't happen overnight, but it will happen. Apple is already preparing for it, with their Grand Central Dispatch.

Re:Cores is the new MHz

[pankajmay]

I agree.
Intel has also been doing its part. The Thread building blocks library is being increasingly used. http://www.threadingbuildingblocks.org/

Re:Cores is the new MHz

[sjames]

Hell, I'd love to see it turned on for all the official build machines, but it's just a little bit scary, since I have seen it fail (as in the build driver crashing) about 1 in a 100 times. But for daily dev work, it's awesome.

That can be fixed by making sure make knows about all of the dependencies. Parallel build failures happen when the serial build just happened to work due to the order make processed the rules (but could break at any time).

This chip from AMD was launched WEEKS ago!

[Anonymous Coward]

This story is so out dated now it's crazy. Slashdot needs to wake up and follow tech again.

Fuck everything

[oldhack]

Fuck everything. We are going to five cores.

gcc

[Anonymous Coward]

All the big cpu hog apps like gcc support multiple cores quite well. If your not running big cpu hog apps, just get yourself a nice high end laptop, and plug it into a big monitor, then you'll have portability and all the performance you'll ever need.

The BIG News that the SOCKET stays the same AM3

[Proudrooster]

This is what I love about AMD, other than the price and ability to overclock. I can upgrade the BIOS and pop this CPU in my system without throwing out the MOBO and having to reload the O/S. If this were Intel, I would have to buy a new motherboard to support the slightly new CPU.

Thank you AMD for not playing socket-a-paloozo like Intel!

Oh, BTW, my 3.2Ghz AMD Phenom II X4 955 Black Edition runs stable at 3.8Ghz for $160. :) If I ever win the lotto I will buy an an i7, until then, GO AMD!

Re:Cores vs performance - VMware

[seifried]

For all us virtualization types more cheaper cores = more better. The future is in virtualization and I think AMD gets this.

There is no such thing as PIRACY...

[ub3r n3u7r4l1st]

In Zhongguo.

FCE Ultra is FOSS btw.

Nobody expects you to ....

[ub3r n3u7r4l1st]

abuse an AMD processor like overclocking.

I have built 8 machines since the socket 7 era and not a single problem come through. My K6-2 333 can still run as of today (w/ Win98 of course).

I don't really care for AMD at all

[Sycraft-fu]

And yet I still hope they do well. Competition is good for everyone. The better AMD's offerings, the better Intel's response has to be if they wish to compete.

You can really see that back in the Athlon era. Suddenly AMD launched to the top performance wise, they had a chip that was powerful and relatively cheap compared to the P3. What happened? Intel cut prices, but also released a huge speed bump. Whereas previously it was in the realm of 600MHz the P3s topped out at, they started shipping 933MHZ P3s in rather short order. Clearly Intel was producing chips that could work faster, they simply didn't bin them higher because there was no need. Game them a way to bring out speed improvement for not cost later. However with AMD's competition, they had to do it sooner.

Then of course there's the P4. It wasn't a bad architecture over all, but it didn't work as well as it should. Main reason was speeds didn't scale like Intel thought they would. Initial tests showed they should be able to get 10GHz out of them in time, but real world it didn't work out. Ok but they were still plenty good chips, they performed well enough for what most people used. Intel could have simply refined the design... But that wasn't an option because AMD's offerings were so strong. So instead Intel had to do a redesign, and from that we got the Core 2, which is an extremely solid architecture.

The fight with AMD is what keeps the costs low and quality high on both sides. So, even if you don't care for AMD, as I don't, you still have to like the company and that they are around. I want to see AMD going for Intel's throat on a continual basis. I want to see both companies have to push their technologies to the limit to make the very best chips possible, and then sell them as cheap as possible. The competition is a wonderful thing.

Re:I don't really care for AMD at all

[mirix]

Whereas previously it was in the realm of 600MHz the P3s topped out at, they started shipping 933MHZ P3s in rather short order. Clearly Intel was producing chips that could work faster, they simply didn't bin them higher because there was no need. Game them a way to bring out speed improvement for not cost later. However with AMD's competition, they had to do it sooner.

Yeah, they also had to pay off retailers to keep them loyal.

Things could have been considerably different today, with athlon beating PIII's (in performance and price), as well as netburst being a major fail on release (and arguably during it's entire lifetime). Crooks.

Re:I don't really care for AMD at all

[sznupi]

Intel started doing a bit more than "cut prices, but also released a huge speed bump" (BTW, remember P3 Coppermine 1.13GHz? ;p ), as shown by recent record-breaking fine from the EU and settlement with AMD (both almost $3 billion total? Supposedly Intel cheated the market for at least that much...imagine what AMD could've done with R&D and fabs if they would have the funds which were otherwise illegally funelled to Intel). The company for which you presumably do care about doesn't really share your enthusiasm for competition...the way they fought, it kept costs higher and quality lower on AMD side.

BTW, "Ok but they were still plenty good chips, they performed well enough for what most people used" in regards to P4 wasn't quite the case with first versions, which were much more expensive and slower than P3s they replaced. Plus lots of unsuspecting people of "CPU must be from Intel" type got Willamette Celerons, which were very castrated, cache-starved (as far as Netburst was concerned)...making them very slow, and a horrible deal.

for less than $1000,why not get a 12-core Opteron?

[strstr]

For less than the price of Intel's top desktop chip, you can get an uber-1337 AMD Opteron with 12-cores. Beat that, Intel...

Prices start at $750.

The user ? Or the viruses ?

[DrYak]

Even if you throw in multi-media, including voip and video, I doubt your average user will be able to use all that computing power

...but once you thow into the mix all the dozens of viruses, trojans, spywares and phising systems which the clueless user has collected by clicking open every single e-mail attachment, suddenly you realise that Average Joe's computer has even problems keeping up with simply sitting idle (and spitting tons of SPAM, coordinating DDNS attacks, etc) let alone have enough processing power to run even a browsing session in addition to the rest.

Re:Cores vs performance - VMware

[Lord Kano]

For all us virtualization types more cheaper cores = more better. The future is in virtualization and I think AMD gets this.

Don't doubt for one minute that Intel gets that too. They're just pressing their current advantage over AMD. When AMD catches up again, Intel will start spinning off 8x or 16x core CPUs.

LK

Re:Cores vs performance - VMware

[amcdiarmid]

I am of two minds about this:

Mind 1: Several of the benchmarks essentially state that virtualization is the only place where the AMD 6 core clearly out-benches the Intel 4 core.

Mind 2: Making a whitebox for ESX is such a pain in the ass that it's almost worth more to buy a frigging expensive server to do the job.

-> anyone know a good motherboard to stick a 6 core amd chip in and run with ESX. No: Virtualization on top of a full OS is not the same.

more cores capt'n

[timmarhy]

screw all these whingers bleating about their pathetic use of computer for firefox and email - no one gives a crap that you only need a joke of a computer. just leave this stuff to us REAL MEN who use our large processing power.

I placed an order yesterday for a dual quad core xeon with 24gigs of memory and 10k rpm scsi drives. and that's just to tide me over until my real machine arrives (waiting on availablity of the 6 core systems - my project can't wait).

mind you i'm using this to model the process of extracting resources from about 1000sqr km of dirt.

I DO need the power.....

[Anonymous Coward]

I do live music production, and I can use all the horsepower I can get.

My software is quad-core optimized, at present. On my dual-core machine, I can definitely do more than I could before that optimization. So I'm wondering, when I'm doing a crapload of DSP computations and demanding consistently ultra-low latency performance, where does the processors/threads/cpu speed/ memory speed equation balance out?? It's hard to believe my HD speed matters much, when I've got 8 gb of RAM. I do stream samples from disk, but that stopped being a bottleneck 5 years ago.

No one does these kinds of real world tests in reviews, even if they offer frame-rates for 17 games!!

Re:I DO need the power.....

[mkiwi]

HDD speed matters much more than you probably realize. If you get a SSD, which is a very low latency device, you can do more ops / sec and suddenly that 8GB of ram is unnecessary. Plus with some of the more advanced (sampled) audio instruments, even one mic position can take up to 8 GB of space. For example, I have a sampled instrument that is 8GB for each of the mic positions it offers.

You would not believe the performance boost I got when I put an Intel SSD into my music workstation--you don't even need the ones that have high write speeds. An X-25M drive is a fantastic investment because it can read all my files on the fly. It's awfully expensive (and time consuming) to get 64 GB of ram and then have to wait for your mechanical hard drive to load all that stuff into memory, not to mention the stuttering that happens when you push too many I/Os per second through it.

Cores are obviously nice for tone-generated instruments, but IMHO, a faster CPU with a low-latency hard drive works the best if you're into the really high-end stuff.

Re:I DO need the power.....

[Colourspace]

Hi. I'm looking to upgrade my studio rig after 3 years of not making music (after 17 years of actually making music). I would like to pick your brains, as we say in the UK? Still stuck on SX but would like to upgrade. Could do with some SW/HW advice if you have the time.

Re:I DO need the power.....

[Anonymous Coward]

I remember 10years ago .... music stores were quoting that you could only use Intel CPU computers for audio recording, because there was something wrong with AMD PC's ..... I quoted pebkac ... Anywho .... Wake me up when they have 42 ;)

turbo?

[ouachiski]

I thought we got away from this whole turbo thing long long ago with computers. I dont understand, does the turbo run on exhausted gases from the chip? Do these processors have combustion chambers that will benefit from said turbo?

Re:turbo?

[Terrasque]

actually, turbo means that some of the cores are set to powersave, and thus creating less heat (intel's solution can actually power off unused cores completely), thus the remaining cores are able to produce more heat, aka run faster.

Intel's boosting solution is better ATM since those cpu's can power off cores completely, while amd's solution can only set them in powersave mode. And that's why Intel's cpu's still hold an edge in apps that only run on 1-2 cores. While AMD gives you six cores for the price of a 4core intel cpu. If you can use them all, amd wins.

The boosting itself happens automatically depending on load IIRC.

Re:ill pit my i7-920 against any AMD 6 core

[Anonymous Coward]

You lost me at "dark side", sonny boy.

Re:for less than $1000,why not get a 12-core Opter

[this great guy]

And the 8-core Opteron 6128 2.0GHz starts at $280.

Compare to Intel's cheapest 8-core processor, the Xeon x6550 2.0GHz that sells for $2500. ten times the price! When people tell you AMD provides more value per $, here is the proof!

Re:for less than $1000,why not get a 12-core Opter

[Anonymous Coward]

What's this 1337? I can buy 6-gigs of DDR3-1600 RAM for less than $200!

Re:for less than $1000,why not get a 12-core Opter

[randyleepublic]

You and GP must be kidding. 12 cores at 2ghz is about my worst nightmare - high hopes dashed against the rock of reality.
Have mercy and drown me in molasses - the torture would be less prolonged.

Cores and AMD

[yoshi_mon]

To me the new AMD six core is a little bit of "me too!" from AMD. Not that there won't be people who won't find practical uses, and no buying one to OC it so that you can get a higher folding score does not count imo, for them but it's still as many have pointed out hard to find real world scenarios where people need that type of a CPU on their desktop.

AMD not only has to compete with Intel on the technology front but marketing as well. And again I don't want to take anything away from AMD and the idea behind pushing the envelope on new tech. But when it comes to end users they really don't know and or care what is driving what they do with their computer. I see people's eyes glaze over when I even start to talk about what type of hardware I'm going to set them up with. They simply do not care.

However I have seen where people have been brainwashed by the marketing. People have asked me if their system is Intel Inside. I try to explain to them that at most price points AMD is a better buy and the more brainwashed come back to me with some very clueless lines like, "But if I don't have and Intel I won't be able to run what I need to." I even remember back in the early 2000's walking into a local computer shop, I needed a mobo asap, and one of the sales reps told me that AMD CPU's were, "Garbage. We don't even stock any AMD parts."

I asked if he knew about the, at the time very high end, computing array that was I think setup at GT that was using AMDs and he started to sputter. "Well, I don't know about that." Of course you don't you idiot I felt like saying, but I just left and have since made it a point to make sure that people that I know and do work for look out to be wary of that place.

My main point is that AMD serves many purposes in what our modern computing landscape is. I personally do like them a lot but as someone who deals with many systems I deal with Intel plenty too. And hell I like a lot of Intel's products. They have top notch R&D and blah blah blah. But we would be a poorer group of computer users without AMD even without all of the other reasons to like them.

Re:Cores and AMD

[Vectormatic]

I even remember back in the early 2000's walking into a local computer shop, I needed a mobo asap, and one of the sales reps told me that AMD CPU's were, "Garbage. We don't even stock any AMD parts."

back in 2003 i ordered a custom built machine at a local shop, they favored intel, but since the northwood 3.0 GHz (only intel chip i cared about at the time) was WAY out of my budget (700 euro cpu, 300 euro mobo), i insisted on an athlon XP. The guy tried to convince me that amd makes unreliable shit and overclocks their own stuff and such, but i insisted.

I got my system, and was happy, but after i while i found out it was running at 100 mhz FSB (as opposed to the specced 166 mhz), they had just upped the multiplier to have the core clock match the specs (yes, my athlon XP 2600+ does not have a multi-lock, none of those chips did until the barton core came about). I asked the guy who built it about this and he claimed that he could not get the system stable at 166 mhz (implicitly blaming AMD). A few years later i found out the stick of ram he had used has errors in it, and doesnt run stable at 166 mhz, causing the instability. Just last weekend i swapped some different ram in there, upped the FSB to spec, and the system is solid as a rock.

moral of the story, people slagging off AMD for stability and such are tools and dont know what they are talking about

Re:Cores and AMD

[sjames]

Things change regularly in the PC world. The guy slagging the AMDs might have been thinking of the long gone K6-3 systems with inadequate cooling back when AMD ran hotter than Intel. I saw plenty of unstable k6-3 systems, but most could be made stable with the proper application of heatsink grease and a properly sized cpu fan. A surprising number of techs in that day were clueless about proper installation of the heatsink.

Not long after that time, there were heat problems with 1U dual Athlon systems, but there were also plenty of 1U Intel systems that spent much of their time triggering the thermal throttling.

Since then AMD has addressed the heat issues nicely and Intel seems to get hotter and hotter (but not so much that they are unstable).

Re:Cores and AMD

[Vectormatic]

well i know there have been heat issues (more recent to my build off course the thunderbird athlons were hotheads..), but there is NO excuse for not just running memtest when a system is unstable, my memtest found the bad ram in just a few minutes

Re:Cores and AMD

[sjames]

I agree fully. It appears he was both a poor diagnostician and failed to keep up with current tech.

3A(Godson 3A) and 3B are coming

[keneng]

Maybe you'll be happy to know there are other alternatives that are very interesting appearing on the CPU manufacturer radar:
1)Godson 3A(4-core) and 3B(8-core).
http://translate.google.ca/translate?js=y&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.gmw.cn%2Fcontent%2F2010-04%2F21%2Fcontent_1099818.htm&sl=zh-CN&tl=en

www.lemote.com will also have a 3A offering in August.

2)Nivida Tegra T20 seems to also be a 4-cores among other cool features.
http://www.clonedinchina.com/2010/01/viewsonic-vtablet-101-android-tablet.html

Re:for less than $1000,why not get a 12-core Opter

[DigiShaman]

Why? That chip is perfect for an MS Terminal Server.

You get more cores dedicated to user sessions, yet each session doesn't need blistering performance. It will provide improved application and session response times however.

Re:for less than $1000,why not get a 12-core Opter

[randyleepublic]

I'm talking about for my workstation. I don't understand why exactly, since I don't do anything, but I need blistering performance. See, here's the problem: what I want is the "instant computer". It sits there doing nothing, making no noise, and hardly using any juice. I click on something. Instantly, what is supposed to happen happens. Instantly! I keep trying, but I'm not there yet. SSDs have had the biggest positive impact by far. I bought a core i7 last time, but effed up and got the conservative board that doesn't overclock. Of course that means that the i7 series processors are the most stable overclockers ever made. Of course. Next is an overclockable board and some wicked fast memory. Add one more ssd to the raid 0 array and that ought to do it.
Anyway, that is what caught my eye about your post, as the high end Intel cpus are sick pricey. So a 12 core for $750? But 2ghz? No way.

Re:Scientific Computing

[MrWookie]

I'm in a computational biology lab. I extensively benchmarked intel vs. AMD. We run different kinds of calculations, some of which will use a whole node's worth of processors and others that run just on a single core, but we run tons of them at the same time. AMD is a tremendous value in either scenario. Intel processors are interesting if you can only have a certain small number of nodes and are only trying to maximize processing power rather than cost effectiveness, but AMD's approach of just giving you a bunch of cores on the cheap has been great for us.

planning for the future

[j1m 5n0w]

Secondly - unless you are ... calculating eigen values of huge matrices ...

Actually, that is what I've been using my home computer for recently. I think you're right though about few "normal" applications being able to use many cores effectively. However, I think you should keep in mind that much of the software that will be run on current hardware hasn't been written yet. Of the computers that are being bought new now that will still be used regularly three or four or eight years from now (why buy a new computer if the one we have is fast enough), they will probably be running much more heavily threaded applications.

Even if the vast majority of applications in the not-too-distant future remain single-threaded, it is likely that the few that are CPU-bound will be optimized and properly threaded. You only have to fix the 1% of the code that takes 99% of the time - the rest doesn't matter.

Re:Cores vs performance - VMware

[symbolset]

Lots of cores are good. A good amount of RAM helps too. And when you put a bunch of servers that do real services in one box, storage and network bandwidth and latency are also important. At the moment memory seems to be the sticking point rather than cores.

For many virtualization scenarios right now with VMWare there is the VMWare licensing to consider. The 8GB DIMMS are still spendy but so is the VMWare licensing so there's tension between server density with expensive DIMMs to minimize rackspace and licensing cost, or twice as many servers with the cheaper DIMMS and paying the increased licensing.

HP and Cisco both have interesting propositions in this area, with blades that do 10GbE and FC for good bandwidth, have dual 6-core processors and support 192GB of RAM or more. When better processors and cheaper memory come out the Cisco UCS solution may have challenges because the architecture may become I/O bound with only 2 10Gbps links for both network and storage per half-width blade, and sharing at least half of that in the chassis uplink. The HP blade solution supports full line rate between servers and an insane amount of uplink - and it's denser than the UCS so it takes up less rack space. The UCS solution uses an ASIC to more than double the number of memory sockets, so for example a full-width server supports up to 192GB using the cheaper 4GB DIMMS and they claim they all work at 1333MHz. I don't know what IBM and Dell are doing here but I know they have products too.

All of the basic virtual environments are basically free (except Microsoft's Hyper-V, of course). Microsoft software is practically free in education environments so they're making inroads there. But in the enterprise the high availability and reliability features of the advanced commercial packages are compelling. There's something awesome about asking an admin management type to evacuate a server so you can work on it (because the local IT support is out today), and watching her migrate the VMs off in a few seconds so you can take it down.

One of the really neat things about VM consolidation is that 20 physical servers with 4x 1Gbps Ethernet don't actually use it so by consolidating them you eliminate waste. Not only that, but by moving to the VM host with 10GbE you get virtual servers that have multiple 1Gbps connections but each has a 0.2ms ping to the external gateway and each other. This makes many things work faster like databases, websites and such. The downside is the downside of sharing: if you don't plan carefully and get a storm load, the servers will contend for bandwidth and knock each other offline.

Now that DDR3 is becoming cheaper than DDR2, I'm glad to see AMD adopt it. I like their 8-core server chips for workstation stuff - a coworker and I are building out dual-8 core boxes for virtual machines and such.

Re:Cores vs performance - VMware

[seifried]

Mind 2: huh? You can buy AMD based servers you know. Sun/Oracle, Dell, IBM, pretty much every other company that sells servers.

Re:Cores vs performance - VMware

[haruchai]

You're totally right about ESX/ESXi and whiteboxes but have you considered Citrix's XenServer? If you have experience installing ESX, then XenServer is quite straightforward and it doesn't snub commodity hardware like ESX does.

Citrix has been offering free downloads of XenServer for a while but a limited trial of their extras as well.

Re:for less than $1000,why not get a 12-core Opter

[j1m 5n0w]

I noticed you can get a 2Ghz 8-core Opteron 6128 for $300 from newegg, which might not be a bad deal for a desktop, assuming you have an application that can use that many cores.

Some of us CAN use those extra cores.

[Anonymous Coward]

Anyone who knows a bit about cgi (I mean ray tracing, not the the real time rasterization you see in games) knows how much of a pain in the ass it is to have to wait 48 hours to render a scene just to be able to see if the light scatters correctly on that orc's forehead. Ray tracing is one of the few practical applications that parallelizes really well, and you will actually use close to 100% of your six core grunt. Since I got an AMD Phenom II 810 quad core, the difference in rendering time for blender has been, well, phenomenal.

Lets not be too quick to jump on the "omg moar cores is teh suxorz" bandwagon, some people actually need it.

Re:Cores vs performance - VMware

[Vectormatic]

Don't doubt for one minute that Intel gets that too.

That's why intel disables their VT instructions on certain CPU's

Re:Cores vs performance - VMware

[Lennie]

I wouldn't be surprised if they do so, because it cheaper to just produce the same thing instead of different productionlines and disable some features or disable things on when production fails on a series and sell it for less.

Re:Cores vs performance - VMware

[mlts]

Don't forget the other advantages of VMs. Snapshots come to mind, and with a backup program that is able to deduplicate, backing up a VM server box hosting a bunch of VMs isn't a daunting task.

I even have seen IDS systems that had both a part that ran in a VM, and a part that watched VM traffic. If either one detected significant intrusion, the VM would be snapshotted for forensic purposes, then a last known good snapshot would be reloaded. Of course, this was mainly for systems which were caching items or Web render farms, rather than ones where persistant data was critical.

Of course the biggest advantage to a VM is the fact that the client does not need any modifications at all when the VM server gets a hardware upgrade. AMD or Intel, OS X, Linux, or Windows? If a variant of VMWare can run it, it works.

Re:ill pit my i7-920 against any AMD 6 core

[mlts]

I have found that it doesn't matter for computer life if you have Genuine Intel, or Authentic AMD.

Want to know the stuff that actually matters?

1: First and foremost a decent PSU. This means the difference between a well made, stable machine versus one where components die of mysterious causes every few months. I have had great luck with Corsairs and Antecs. Just make sure to get one that has significant headroom between the estimated wattage and its rated wattage. Mainly because one of the most reliable ways to feed USB devices power is through hanging them from your machine, and a couple USB PCI cards. When in doubt, find a quiet 1000 watt power supply, and call it done.

2: A decent enclosure. Yes, people talk that cases are cases, so go with the cheapest. However, I've had enclosures last multiple motherboards. So might as well get a case that has rolled metal edges to minimize getting cut, a solid rack for hard disks, etc. Cooling is important, so it can't hurt to get a case that can support multiple fans.

3: Cooling. What can kill a machine is not enough air blowing through it. Getting a case and decent fan setup can make a machine last a long time.

4: Motherboard. Yes, it might be easy to go cheap on this, but this is what is controlling your CPU, and going cheap on this may mean major headaches in the future, especially if something partially works. You want to get a decent name brand motherboard, because the better ones actually have true hardware RAID on board (not hardware-assisted), which allows you to use two drives for your OS.

5: RAM. I've seen people buy pulls and then wonder later on why they keep having subtle problems, until I fire up a RAMtest utility and find areas with problems.

6: This is one thing that is important as everything else, but something not often checked after. A UPS. Unexpected power cut outs are horrific for equipment. Not just software with unexpected downtime, but hardware. Putting your machines on a solid UPS will easily prolong their lives a number of years.

Re:ill pit my i7-920 against any AMD 6 core

[sqkybeaver]

as if this was not stuff i did not know! sure there is a better price with the AMD products, over all that is their biggest benefit by far. everything else is minute in comparison. you chould look into the market share for the last 20 years, Intel has been the leader and will continue for many more. it is Intel's generosity is the 80s and early 90s that allowed AMD to even enter the X86 market. look at the Itanium, it was out before any AMD 64 platform and largely influenced current intel micro architecture.

Questionable car analogy: not quite turbo . . .

[wrencherd]

"Turbo" seems the wrong (perfunctory) auto analogy.

Boosting clock-speed on some of the cores when others are idle sounds more like "variable displacement technology" than it does turbo-charging.

Re:for less than $1000,why not get a 12-core Opter

[r_jensen11]

Screw 12-cores, we're turning this up to 11-...er, nothing to see here....

Homebrew != piracy

[tepples]

ca65, and FCE Ultra.

Ahh, so you are playing PIRATED game ROMs in FCE Ultra.

I guess you missed the "ca65" part. It's a Free assembler targeting MOS Technology 6502, the CPU architecture used in the Nintendo Entertainment System. I use it to make my own ROMs for other people to play. (GIMP is for editing the 128x128 pixel sprite sheets that the NES uses.) Or are you calling Concentration Room and LJ65 pirated?

Re:Cores vs performance - VMware

[Caetel]

No doubt forcing you to buy higher price chips if you want VT. That seems like getting it to me.

This is a Great Chip for Power Users

[foxalopex]

For anyone that plans on using this CPU as a workstation or light server chip, this is the best way to go. I recently priced an Asus M4A785TD-V EVO motherboard and it's only an amazing $120. (Comes complete with a built in low powered graphics card too) Pair that with this Phenom X6 and ECC ram and you have an amazingly great value Virtualization or Parallel rendering system. This chip is probably overkill for consumers and gamers but for the folks who can use it, it's an amazing steal. :) In any event, I work for a small company and so far AMD's proven to be the best value for light servers. Intel's primary best designs and strengths are in the Laptop market where they make advanced chips but on the Desktop I still find AMD to be great.

its nice, but

[nimbius]

until these multicore monsters start using less electricity, i dont know if ill commit to anything more than the 45 watt dualcore i have now.

Yay....

[Anonymous Coward]

....for Moore's Law!

Simulated "turbo" mode on existing BE's

[bored]

If you have a Phenom II black edition, you can have functionality similar to Intel's turbo feature using this (http://sourceforge.net/projects/clockboost/) little utility which monitors core temp, and overclocks the processor if the temp is below a set threshold.

Re:ill pit my i7-920 against any AMD 6 core

[DDLKermit007]

lol that's fine you went over to the dark side. My system cost half the price of yours for what amounts to a 10%-15% gain.

Re:Cores vs performance - VMware

[Anonymous Coward]

Yes, we just got bit by this at work. We have run a mixed enviromnet of AMD and Intel, and we just found out as we depolyed virtualization for our developers that our intel quads can't do hardware virtualization and thus is locked out of 64bit vms. While all our AMD processors that fit AM2 or better can do hardware virtualization. So new company policy, buy AMD

So what?

[Anonymous Coward]

If you look at the reviews they barely can match the 920/930/etc. line w/only 4 cores. (We'll ignore the Intel hexa-core as it's WAY over $200 and i7 930s can be had from MicroCenter for $200 as 920s before them could also be had, or for a little more from most online retailers.) i.e. they scale horribly.

AMD needs to SPEND MONEY on R&D. Their per-core performance is pretty pathetic which is why they are forced into the value level of the market, unlike up until 2006(core) they could charge on parity(and above) what Intel was charging since their CPUs ATTM were actually superior performers. Now they're just running the same old arch with a new memory controllers (DDR2 & 3) and tweaking HT a little bit.

Personally, I'm waiting next year for Sandy Bridge & Bulldozer. We already know that SB will have a new socket, LGA2011(only one worth looking at IMNHO) and I strongly suspect that Bulldozer will have a new socket as well, although I don't recall reading any articles mentioning it yet, but it seems likely.

What about power consumption?

[brucmack]

One thing I'd love to see is a price/performance analysis that includes the cost of running the system, so I could compare the purchase price to the total price after running 4 hours a day for a year, for example. AMD has always been good at offering similar performance to Intel at any given price point they release to, but lately their products have been much less efficient than Intel's, as they require more cores to achieve similar performance and are built on a larger process.