AMD Starts Shipping First Bulldozer CPU

MrSeb writes "After an awfully long wait, AMD has finally begun shipment of its Bulldozer-based Interlagos (Opteron 6200) server-oriented CPU. If you believe AMD's PR bots, it is the world's first 16-core x86 processor. Unfortunately, and possibly because of reports that AMD is struggling to clock its Bulldozer cores to speeds that are competitive with Intel's Core i7, there's no word of the 8-core desktop-targeted Zambezi CPU. If AMD doesn't move quickly, Intel's Sandy Bridge-E will beat Zambezi to market and AMD will lose any edge that it might have."

Comments

Sandy Bridge-E

[Anonymous Coward]

If AMD doesn't watch out their mainline $200 processor will be made obsolete by Intel's $1000 EXTREME CPUs!

Re:Sandy Bridge-E

[CajunArson]

You may laugh but think of it this way... if that $1000 Gulftown CPU from March of 2010 can still beat an 8 core Bulldozer that comes out 19 months later, then you would only have to realize a marginal benefit of about $1.75 per day to make it economically worth your while to have bought the "overpriced" Gulftown chip. (that includes the cost from Intel motherboards that tend to be more expensive and the extra RAM for a triple-channel configuration). Nevermind the fact that 6-core chips have been sold for $600 for some time as well. I can think of a bunch of professional applications that can easily show a $1.75 / day benefit from the extra cores. Maybe not for playing games, but for a lot of real applications.

  Bulldozer should beat the consumer-level SB chips at perfectly threaded integer benchmarks, but it remains a very open question if it will be able to beat the almost 2 year old Gulftowns at the same tasks, and it is an almost foregone conclusion that it won't beat the 6 core SB-E chips at those tasks. Factor into account the 315 mm^2 die size of EVERY Bulldozer (not just the 8 core ones, but the cheap 4 core ones too since AMD only has 1 die design) and the immaturity of AMD's 32nm process and things could be expensive for AMD on the desktop. That's why it makes sense to ship the server chips first where AMD has some hope of getting higher ASPs.

Re:Sandy Bridge-E

[aliquis]

If AMD doesn't move quickly, Intel's Sandy Bridge-E will beat Zambezi to market and AMD will lose any edge that it might have.

Ok, so nothing will be lost?

(No, I'm not trolling or flaming, personally I would get a new machine to play Starcraft II, AFAIK SCII only seem to use two cores. AMD themselves has claimed their chip (best consumer chip?) would be similar to 2600k in performance. 2600k is quad core vs octo core for the AMD chip. Considering the work load the 2600k will still outperform the AMD by a lot. Also with socket 2011 we talk quad channel memory instead of dual channel and even back in 1156 vs 1366 days and with SCII only using two cores the 1366 chip beat the 1156 chip. So obviously memory bandwidth was an issue even at two cores. So I'm just stating the obvious. Intel is the leader and will remain the leader for the time being.)

Re:Sandy Bridge-E

[CajunArson]

There's no evidence to indicate that AMD's "mainline" $200 CPU will be much better than the existing "mainline" $200 2500K that's out right now... Just because Intel offers chips at a higher range than AMD doesn't mean that AMD automatically beats Intel at everything below the highest range. When the 2500K first came out it was priced lower than AMD chips that were substantially slower... AMD "corrected" the price to performance ratio by slashing its own prices, which didn't do too much to help its profitability.

Re:Sandy Bridge-E

[Anonymous Coward]

I've (honestly) just been asked what our expected budget requirements are for hardware for the next year. Please inform me where I can go to use the patented Intel time travelling technology so that I can retroactively use things before I decide to purchase them.

Re:Sandy Bridge-E

[Anonymous Coward]

It spells even more trouble for AMD that they cannot charge even $800 for their new top-of-the-line CPU. It is a negative feedback loop for them - by failing to create competitive technology they are forced to keep down their next-generation development budgets, and they need a lot of luck to hit the jackpot on the next round. Still, they are not dead yet, as they can still compete quite well at the low end, and it is not impossible for them to maintain that market position indefinitely even without spending top dollar for best possible engineering designs.

Re:Sandy Bridge-E

[X0563511]

Personally I don't care. I'm still not buying Intel if I can help it.

Re:Sandy Bridge-E

[rbarreira]

There's no evidence to indicate that AMD's "mainline" $200 CPU will be much better than the existing "mainline" $200 2500K that's out right now

There is some, depending on your application of course. If computer chess analysis is your thing, you would see benchmarks results like these, where the $189 Phenom II X6 1100T beats the $219 Intel 2500k.

So AMD already has CPUs which are price-performance competitive, surely Bulldozer shouldn't be worse in terms of price-performance.

Re:Sandy Bridge-E

[aliquis]

It's a free world, or well, market, or well, neither, but you are free to pick the AMD.

Just don't claim it's superior because you'll be wrong and make a fool out of yourself. Or well, feel free to claim it's superior and make a fool out of yourself to :)

Re:Sandy Bridge-E

[CajunArson]

Please inform me where I can go to use the patented Intel time travelling technology Haha an anti-Intel snarky comment on Slashdot... way to speak truth to power. What's funny is that if I actually told you I had a time machine in 2010 that could bring you the fastest CPU available from AMD at the end of 2011 and you only had to pay a few hundred bucks for my service, you'd probably jump at the chance... which is exactly what Intel effectively did with Gulftown.....

Re:Sandy Bridge-E

[beelsebob]

Yeh... Because AMD's $200 X6 1100T isn't already beaten by intel's $190 i5 2400... oh wait, yes it is.

Re:Sandy Bridge-E

[sgt scrub]

the immaturity of AMD's 32nm process and things could be expensive for AMD on the desktop.

That is true, and as you point out, for the desktop. Machines in a data center are cooled so the number of cores is a better measure of functionality. If you build machines that run multiple VM's, which is usually the case, that cheaper 6200 will not only outperform Intel's Gulftown and more likely be preferred when adding more machines to the data center even over the SB-E chips. If AMD can get a better footing in the "cloud" infrastructure they might make enough to move to a die size smaller than 32nm, which is REALLY what they must to do.

Re:Sandy Bridge-E

[bhcompy]

That $200 is still a fuzzy number given the motherboard prices. You can get a very good AMD chipset(880/890) for $100 and have all the latest features(USB3, SATA3, etc) while being forward compatible for quite a while(manufacturer dependent, but Bulldozer(AM3+) is compatible with AM3 chipsets with BIOS updates). With Intel, you're still paying more for the equivalent and next year you'll need to get another motherboard to upgrade that processor because of constant socket changes.

Re:Sandy Bridge-E

[V!NCENT]

Talk to a server farm when Intel is putting money in the development of Coreboot and we'll talk.

For now AMD is superior, because a server reboot requires about 1/100th of the time that it takes an Intel CPU farm to get back up due to horrible BIOSes. The more motherboards you have, the longer it takes due to serialized bootup. Ouch... Massive ouch...

Downtime versus marginable CPU speed... And less cores...

Re:Sandy Bridge-E

[rbarreira]

On what applications? Post benchmarks...

Re:Sandy Bridge-E

[serviscope_minor]

It is a negative feedback loop for them - by failing to create competitive technology

You missed the bit about how they create competitive technology and never get any customers because intel has a huge monopoly.

Also, for cramming flops into U's the quad socket AMD 6100s beat out anything Intel has to offer in the same segment.

Re:Sandy Bridge-E

[Anonymous Coward]

Yeh... Because AMD's $200 X6 1100T isn't already beaten by intel's $190 i5 2400... oh wait, yes it is.

No it's hasn't been.
AMD
Intel

Re:Sandy Bridge-E

[trum4n]

Except Intel's have no pins. This summer, my computer shop repaired over 300 machines. Over 100 of them were Intel boxes, that after a customer bumped their knee on the tower, it didn't work. Re-seating the CPU fixed it, but its a design flaw that i won't pay for. It can cause a BSOD, or serious data loss. Depending how it shifts, it can even blow the motherboard by shorting. Also, paying 800$ for that extra 2% of speed, not even remotely worth it. Considering you can run dual Opterons for the cost, not worth it.

Re:Sandy Bridge-E

[Anonymous Coward]

With Intel, you're still paying more for the equivalent and next year you'll need to get another motherboard to upgrade that processor because of constant socket changes.

Wikipedia:

Ivy Bridge is the codename for the yet-to-be released 22 nm die shrink of Sandy Bridge. Ivy Bridge processors will be backwards-compatible with the Sandy Bridge platform.[6] According to a leaked Intel roadmap, Ivy Bridge processors will be released in March-April 2012.[7]

...
While Ivy Bridge will be compatible with the Cougar Point chipset motherboards associated with Sandy bridge, Intel will also release a new 7-series Panther Point chipset with Ivy bridge.

http://www.overclock.net/hardware-news/988606-tpu-ivy-bridge-1155-compatible-panther.html
http://www.anandtech.com/show/4318/intel-roadmap-ivy-bridge-panther-point-ssds

Re:Sandy Bridge-E

[UnknowingFool]

Yes but that's not really comparing apples to apples. A 6 core 125W processors should beat out a 4 core 95W processor. Yes the AMD is cheaper which is good but it also requires more power which will be offset by electricity costs.

Re:Sandy Bridge-E

[swalve]

That is ridiculous. The processor is locked in. It can't move. Unless you are using the cheapest mainboards in the world whose processor sockets aren't spec.

Re:Sandy Bridge-E

[swalve]

How is it a monopoly if (as purported) AMD chips are both better AND cheaper? Decision makers don't give a crap about the name, they care about the value for the dollar. And Intel has that. It doesn't take too many productivity-sapping weird incompatibilities and spontaneous reboots to sap any dollar advantage AMD might have.

Re:Sandy Bridge-E

[justforgetme]

Yes, they really have to move away from the 32nm process fast. Is the 6200 16 physical cores or just 8 dualthreaded ones? If the cores are physical I can see dual opteron boards become a must for renderfarms/supercomputers and why not virtualization aswell.

Re:Sandy Bridge-E

[Anonymous Coward]

All video games currently in existence. http://www.xbitlabs.com/articles/cpu/display/core-i5-2500-2400-2300_7.html#sect0

1100t wins in applications using 6+ threads. Which is obvious. If you're doing heavy threading then ofc the 1100t is the better CPU. At 2-4 threads, even an 1100t overclocked @ 4.2GHz loses.

Re:Sandy Bridge-E

[Anonymous Coward]

Its like blame Nike that your shoes didn't stop your vehicle because the car manufacturer didn't provide brakes. Ridiculous indeed.

Re:Sandy Bridge-E

[beelsebob]

Here's a much more complete set of benchmarks that don't rely purely on synthetics –oh look, the i5 wins everywhere except for h264 encoding, which the i5 has a hardware unit for.

http://www.anandtech.com/bench/Product/203?vs=363

Re:Sandy Bridge-E

[Anonymous Coward]

Bulldozer(AM3+) is compatible with AM3 chipsets with BIOS updates).

Not quite.

AM3+ is compatible with AM3 and AM3+ processor. AM3 is not compatible with AM3+ processors.

Re:Sandy Bridge-E

[beelsebob]

Even at 6 threads, the i5 gives the X6 a very very tough time http://www.anandtech.com/bench/Product/203?vs=363. More very-multithreaded benchmarks go the i5's way than the X6's, and they're all extremely close.

Conclusion... for 6 threads and more they're equally fast, for less than 6 threads the i5 beats it silly. That's the i5 being a faster CPU in my book.

Re:Sandy Bridge-E

[bill_mcgonigle]

Any particular boards you recommend for Coreboot server work? I've transitioned to all AMD over the past 9 months, but I still suffer with legacy BIOS.

Re:Sandy Bridge-E

[Bigbutt]

Funny, that's why I went dual-core vs quad-core when I built my game machine (specifically for Starcraft II).

[John]

Re:Sandy Bridge-E

[serviscope_minor]

How is it a monopoly if (as purported) AMD chips are both better AND cheaper?

Well, Intel provided huge kickbacks provided that they didn't sell AMD processors. AMD lost billions when the Opteron was stomping all over the Pentium 4 on the desktop and in the datacentre. Intel got a big fine, but not big enough. Given the amounts involved, there is no way the whole thing wasn't a net benefit to Intel.

Decision makers don't give a crap about the name, they care about the value for the dollar. And Intel has that.

Not across the board and they certainly didn't when it counted.

It doesn't take too many productivity-sapping weird incompatibilities and spontaneous reboots massive illegal kickbacks to sap any dollar advantage AMD might have.

FTFY

Re:Sandy Bridge-E

[trum4n]

The socket 775 design was a joke, and they keep playing it for some reason. It's not reliable. And if it was designed properly, it doesn't matter who made the motherboard, cause it would just work. And where the HELL did you get the Nike thing? It's like I'm talking to sarah palin. I DONT AGREE, SO ILL MAKE SOMETHING UP! THEN ILL INSULT THEM!

Re:Sandy Bridge-E

[V!NCENT]

http://www.coreboot.org/Supported_Motherboards#Servers

Re:Sandy Bridge-E

[h4rr4r]

So in your universe motherboards are free?
List price of CPU and board.

Re:Sandy Bridge-E

[Smallpond]

"the Opteron 6200 has eight Bulldozer modules, and each module contains two independent integer processors, but only a single FPU and shared fetch/decode/execute units; in other words, it has more than one core, but not quite two."

This is an interesting architecture. Most general processing is just moving data so an FPU per core is overkill. Also, the main part of an algorithm (or a benchmark, anyway) is likely to fit in cache so use of the bus fetch hardware is likely to be in bursts. It should give better performance than 8 HT cores, anyway.

Re:Sandy Bridge-E

[Anonymous Coward]

Yeah, make up scenarios that are a fuckup in themselves.

"Buy our CPUs because when they run you over, your back on your feet faster." - AMD

"Buy our CPUs because they run circles around the competition and don't run you over" - Intel

Hmm, yeah, difficult choice.

Re:Sandy Bridge-E

[beelsebob]

And in the vast majority of tasks the X6 is slower, including some very multithreaded ones:

http://www.anandtech.com/bench/Product/203?vs=363

Re:Sandy Bridge-E

[Anonymous Coward]

It's two full integer units (ALU) per core that share a single 256bit wide FPU that can handle a 128bit operations from each core at the same time. They share the L2 data cache and the L1 instruction cache. Since the Bulldozer Module has full circuitry to run two threads at the same time on full circuitry it is generally considered to be two cores

Intel's hyper threading has only one set of computation circuitry but has the ability to keep state for two threads (known as SMT). It allows Intel CPUs to take advantage of pipeline stalls without the huge cost of switching to another thread in software, it makes up for poor branch-prediction/instruction-reordering. The Intel Atom gains 50% performance through this.

Sun's Niagara Falls chip did 4 thread SMT and had the ability to switch between them with only 1 clock delay. It was a beautiful beast.

Re:Sandy Bridge-E

[Nadaka]

IIRC, all of AMD machines are full multi core, Intel has patents on the partial application.

Re:Sandy Bridge-E

[beelsebob]

Huh? You can get $70-80 intel boards with the feature set you're describing, and next year's Intel CPU (Ivy Bridge) will still use LGA1155. Notably, the whole "AMD boards carry on working with future CPUs" thing is a myth –they work for one generation, at most. Bulldozer chips will not work in AM3 boards, only AM3+. Similarly, the current crop of Phenom IIs will not work in AM2 boards.

Re:Sandy Bridge-E

[neokushan]

A few years ago, AMD had intel by the balls. I mean, seriously they were the king in terms of price, performance AND efficiency. I'm talking about the Athlon and Opteron 64's, the clawhammer/sledgehammer. They had a brilliant chip design, they pushed 64bit computing and all Intel could offer was the Pentium 4, which struggled to go above 3Ghz and could heat your entire home.

Intel bet on pushing more and more Ghz while AMD decided to go down the efficiency route. Sure, they had to create a new way to name processors so people didn't get confused about a 2.2Ghz Athlon 64 beating a 3.06Ghz P4, but none the less Intel hit a brick wall while AMD kept on going. 3000+, 4000+, then dual core. Intel had their own dual core chips, but they were somewhat crude by design, little more than gluing a couple of Prescott P4's together and hoping for the best.

AMD were on a roll, so while the Gravy train was in, they bought ATI with the intention of creating some sort of monster "hybrid" chip that combined the CPU and GPU. Fusion was the future that AMD was going to pioneer, just like they did with 64bit and dual core desktop chips.

What happened? Intel got their shit together, that's what. Intel dumped the P4, went back to the P3 and came up with Conroe, a real performer that didn't overheat the way Prescott did. AMD stagnated, fusion got pushed back again and again, then it got castrated more and more. Within less than a year, AMD/ATI was worth less than what AMD paid for ATI. Fusion has only recently appeared and, surprise surprise, Intel now has graphics embedded in the CPU as well. Whatever advantage AMD was hoping to curtail has gone. AMD needs to step it up several gears, they need something new, another home run like the Athlon 64's and just hope Intel makes another mistake like the P4.

I can't see that happening, though.

*Note: all of this was typed from memory, so some dates/names may be slightly wrong.

Re:Sandy Bridge-E

[beelsebob]

okay then...
X6 1100T - $200
cheap AMD mobo - $50
mid ranged mobo that supports SATA 6000 and USB 3 - $100

i5 2400 - $190
cheap Intel mobo – $50
mid ranged mobo that supports SATA 6000 and USB 3 - $65

Re:Sandy Bridge-E

[Anonymous Coward]

Furthermore, AM3+ is a single-generation socket. Bulldozer is the final AM3+ CPU. For once, AMD did worse at socket compatibility than Intel.

Re:Sandy Bridge-E

[Anonymous Coward]

Opterons are LGA (i.e. no pins) as well.

Re:Sandy Bridge-E

[Targon]

Could be, could be, but there is no information out there how production Bulldozer processors will perform once they start pumping out consumer chips. AMD may have been forced to wait this long to release Bulldozer just to deal with those 32nm process issues. I am not saying that Bulldozer will beat SB, but on the flip side, AMD machines have been selling well in that $500 and under range to this point, so Bulldozer SHOULD help.

The average consumer doesn't need a LOT of processing power, so if the Athlon 2 and Phenom 2 has been enough for THEM, then Bulldozer will be better yet. You also have to look at how far behind AMD has been to this point, and how much of an improvement Bulldozer is. In a Marathon, if Intel was 3 miles ahead, and the release of Bulldozer closes the gap to Intel being only 1 mile ahead, Intel is still in the lead, but AMD is looking better in comparison.

Re:Sandy Bridge-E

[trum4n]

Really? Noted. Last one i used was socket. Been using Phenom II's since they came out. Id still use Opterons. They cost that much less for the same performance.

Re:Sandy Bridge-E

[Anthony Mouse]

The reason for the pinless chips is obviously to reduce returns. With a chip with 1000+ pins, it's easy for one to get bent during transit and then you have customers returning them. On top of that, they pretty much have to use gold conductors for the pins, and have you seen the price of gold lately?

You put the pins on the motherboard and you shift the liability and cost to the motherboard manufacturer. And since Intel designs the socket, that's what they did.

Re:Sandy Bridge-E

[0123456]

On what applications? Post benchmarks...

If he posts benchmarks, the AMD fanboys will say 'oh, but that's just synthetic, my X6 feels much faster than any Intel CPU'.

If he posts game benchmarks, the AMD fanboys will say, 'yes, the Intel chip is twice as fast at 1024x768 when the game is CPU-limited, but when you run at 4096x2048 with the game GPU-limited, the CPU makes no difference, so you should buy a slower CPU so long as it's the most expensive AMD chip.'

Re:Sandy Bridge-E

[Antisyzygy]

At 3 universities I have been to, not one of them had Intel clusters for scientific computations. Some of them have Nvidia tesla, etc., but no Intel. The biggest one I went to was a Tier 1 research University, meaning they get shitloads of funding to buy whatever they want.

Re:Sandy Bridge-E

[Joce640k]

Most programs can't saturate the FPU anyway.

Intel/AMD FPUs can accept data on every clock cycle, the data bubbles through the system and a result pops out N clock cycles later. It's a rare program that can keep putting data in on every cycle so it makes sense to have one FPU per two cores.

You can easily see the effect, try this:

float array[arraySize];
float result=0;
for (i=0; iarraySize; ++i) {
    result += array[i];
}
print(result);

float array[arraySize];
float result1=0, result2=0;
for (i=0; iarraySize; i+=2) {
    result1 += array[i];
    result2 += array[i+1];
}
print(result1+result2);

If your compiler's any good the second loop will run about twice as fast as the first because it keeps the FPU busier. The effect is much more noticeable when the code has lots of divisions and square roots.

Re:Sandy Bridge-E

[afidel]

It really depends. Even with lots of VM's single core performance can matter. We have a lot of varied workloads but I'm much more likely to have a single core in one VM pegged than I am to have any host be processor bound (my entire production cluster averages ~50k MHz of Gufltown processor time at peak load over the last month, this is out of almost 250k available).

Re:Sandy Bridge-E

[V!NCENT]

You can start working for Microsoft. I heared they were looking for good PR in case Windows 8 throws a Vista...

Re:Sandy Bridge-E

[Anthony Mouse]

Well what do you expect? That they would make a Bulldozer with no memory controller and a 100MHz bus so you can drop it into Socket 7 motherboards? The idea is that you can buy a motherboard, then a year or two later get the next generation processor and drop it in. After that it doesn't even make any sense -- the bottleneck stops being the CPU and becomes the fact that the older socket is using two channels of DDR2 instead of four channels of DDR3 etc., which is the whole reason that sockets change in the first place.

Re:Sandy Bridge-E

[MyCookie]

You would be correct until now, but reports indicate that next year AMD will move all of it's processors to it's new FM2 chipset.

Re:Sandy Bridge-E

[Billhead]

It's the required car analogy, what more can you ask for?

Re:Sandy Bridge-E

[afidel]

They are 16 physical cores sharing 8 fetch/decoder units, 8 L2 pools, and 8 dual pipline FPU units. It's basically halfway between a HT setup and a "pure" SMP setup. It's AMD making more efficient use of silicon to make up for Intel's process lead. Intel may end up copying the basic design principal because it's just better engineering, much like they did when they copied hypertransport to move away from the front side bus.

Re:Sandy Bridge-E

[Sloppy]

Is the 6200 16 physical cores or just 8 dualthreaded ones?

Part of what makes Bulldozer interesting, is that there is no quick straight-forward soundbite answer to that question. It's 16 less-than-complete cores, or 8 more-than-merely-dual-threaded cores.

I look forward to the day when some profiler nerd figures out that most cost-effective ratio is to make a chip with 3 memory busses, 17 integer units and 11 floating point units, with 23 sets of registers. /proc/cpuinfo will say it's a 23-core machine so that'll be the soundbite answer, and then some math twit will whine that it's really only an 11 core machine. And everyone will be both wrong and right simultaneously.

Re:Sandy Bridge-E

[Anonymous Coward]

1100t has a lot more overclocking room. i5 2400 doesn't compare to 6 K10 cores @ 4.0 GHz in a thread-bound app.

I'm dubious about those benchmarks too. Which of them are core-bound and scaleable? I recognize 7-zip, x264-second-pass, WinRAR... are there more? IIRC, x264's first pass doesn't fully utilize all available cores.

Re:Sandy Bridge-E

[Mad Merlin]

Actually, if your compiler is any good it would rewrite the first loop to look like the second (or another even faster variant) at compile time.

Re:Sandy Bridge-E

[Anthony Mouse]

The real problem for AMD is that they didn't expect Intel to turn it around so quickly with Conroe. They had an excellent design and they expected to continue extracting reasonable margins from it in order to fund Fusion development, but when Core 2 came around they lost that.

The thing is, Bulldozer is a great direction for them. It will not beat Intel's best at single-thread performance, but it isn't supposed to. What it's supposed to do is offer better performance per watt and per rack unit for common server workloads, which it very much does. And that gets them back into the server room with reasonable margins and keeps AMD alive to continue developing new products.

The real interesting thing is that I'm not sure Intel has a good way to respond to it. SB is designed to be the fastest processor for single threaded performance, and it does that. The trouble is that the server rack is where the high margins are and most of those people don't care about single thread performance, they care about the things that AMD has designed for. So Intel ends up with a bad choice: Choice one is that they can try to adapt SB to be more like Bulldozer, i.e. SMT with more execution units so that you can get nearly two cores worth of performance out of substantially less than two cores worth of silicon and power budget, with the consequent hit to single threaded performance. Or choice two is to continue developing SB on its existing goals for the desktop and for the relatively small number of non-threadable server applications and then dump a billion dollars into developing an entirely new processor to meet Bulldozer for the lucrative parallel server market.

Re:Sandy Bridge-E

[Anonymous Coward]

Um...nearly ALL of them?
Benchmarks

Re:Sandy Bridge-E

[Anonymous Coward]

What are you trying to tell me?

Your argument was basically that AMD CPUs are superior because their motherboards boot faster? Seriously? Boot time vs. run time is what? 1:500000? You sound like a Linux moron: "Oh yeah, it's great we optimized this 0,0002% corner case. Who give a fuck that the whole thing runs like molasses the rest of the time."

And now all you have is this? Poor AMD apologist.

Re:Sandy Bridge-E

[serviscope_minor]

What happened? Intel got their shit together, ... and gave massive bribes to vendors to prevent AMD getting any significant increase market share. This cost AMD enormous amounts of revenue and seriously hurt their ability to compete. This is something that Intel have been found guilty of in court. The damage, however has been done.

Re:Sandy Bridge-E

[beelsebob]

The i5 isn't overclockable at all, but its barely more expensive bigger brother the 2500k has *way* more overclocking headroom than the Phenom. They've got happily to 4Ghz on the stock cooler, and 4.5Ghz on better air coolers.

Re:Sandy Bridge-E

[beelsebob]

Agreed, it doesn't make much sense to keep the same socket for many years. This is one of many reasons why the above "zomg, intel changes sockets all the time" mantra is bollocks.

Intel releases a chip on one socket, they then do a die shrink on the same socket. They then move to a new architecture with a new socket. This is not very much different to what AMD is doing, and gets you the same guarantee of being able to upgrade your CPU in a year or two if you really want to on your now-a-bit-out-of-date board.

Note – AMD are even dropping this level of compatibility. Bulldozer will work on AM3+ and AM3+ only.

Re:Sandy Bridge-E

[0123456]

You put the pins on the motherboard and you shift the liability and cost to the motherboard manufacturer. And since Intel designs the socket, that's what they did.

Bending pins on a CPU is easy. Bending pins on the motherboard is hard if you even remotely follow the instructions.

Re:Sandy Bridge-E

[neokushan]

Except that all came about when AMD came around with the Athlon 64's, because people were asking why the likes of Dell refused to sell them.

Re:Sandy Bridge-E

[UnknowingFool]

My understanding is that AMD is cheaper but normally requires more power and cores to an Intel chip. Electricity costs then offset any price difference. Where AMD has an advantage is their motherboards are easier to upgrade the CPU in the future. I however tend to buy a new computer rather than upgrade my old ones so for me that's not an advantage.

Re:Sandy Bridge-E

[serviscope_minor]

Except that all came about when AMD came around with the Athlon 64's,

Well yes, and that really wrecked AMDs revenue. The were number on on terms of absolute performance and price/performance, but a distant two in sales.

The massive market fixing done by Intel meant that AMD was unable to put nearly as much as they could have done to develop future processors. Perhaps as a result (we will never know), Intel was able to catch up. It takes several years to develop a new microprocessor architecture.

That was also not the first time that Intel had to pay AMD because of illegal market fixing.

Re:Sandy Bridge-E

[Anonymous Coward]

The idea is that you can buy a motherboard, then a year or two later get the next generation processor and drop it in.

But that's not true for AM3+. AM3 was already owned by most AMD owners. Some owners upgraded early to AM3+ for Bulldozer. Only a few % hopped onto the AMD bandwagon late straight into AM3+. So there is no vaunted 'processor upgrade' path here. Bulldozer is effectively the first and last CPU for AM3+. It's not comparable with the 1155 SB-to-IB situation at all.

Re:Sandy Bridge-E

[Anonymous Coward]

Funny that I easily found AMD mobos with both SATA 6Gb and USB3 for $69 on new egg.. with dual 16x PCIe 2.0's, 4 memory slots, and sorrted other additional goodies.

If Intel is so good, why do you need to be dishonest? Your dishonesty here is evidence that you are a fanboy.

Re:Sandy Bridge-E

[UnknowingFool]

Well that depends on how often each event is in your typical operating scenario. If you have to factor in server reboot, I'd have to say there is something seriously wrong. Yes, shit happens but if you have to reboot your servers enough time to where you have base purchase decisions on which processor boots faster rather than performs faster/costs less/uses less power, I'd say you have other more pressing problems.

Re:Sandy Bridge-E

[Anonymous Coward]

The i5 isn't overclockable at all

Non-K CPUs can be overclocked... only with a limited multiplier... up to 41-42x for an extra +400mhz... plus more (200+mhz?) for risky BCLK overclocks.

but its barely more expensive bigger brother the 2500k has *way* more overclocking headroom than the Phenom.

I know that, I wasn't contesting the performance of the 2600 or 2500k/2600k :). I have an i5 2500 arriving in the mail shortly, FWIW.

They've got happily to 4Ghz on the stock cooler, and 4.5Ghz on better air coolers.

I've read some reports of 4.5 OC's with the stock cooler... but those were probably with good TIM and airflow.

Re:Sandy Bridge-E

[kimvette]

Boot time vs. run time is what? 1:500000?

You obviously don't run Windows, or at least don't install the patches. ;)

Re:Sandy Bridge-E

[m.dillon]

You can't overclock an 1100t to run ALL 6 cores at high speed at once without water cooling and without it drawing such an extreme amount of power that it doesn't make sense to do it in the first place. I've done this myself.

You can overclock an Intel i7 to almost 5 GHz with water cooling and it will use half the power of an overclocked phenom in a similar configuration.

There's no comparison.

The only place where the phenom x 6 might win running all 6 cores is with floating point, because the 4 core x 2 hyperthreaded intel i7 only has 4 fp cores. But even there AMD tends to lose because a lot of programs written for e.g. windows don't even bother using AMD's advanced floating point features, whereas every single one of them are optimized for Intel's.

-Matt

Re:Sandy Bridge-E

[UnknowingFool]

So what you are saying even in the best AMD scenario the MBs are priced nearly the same and the AMD chip is $10 more expensive while offering less performance.

Re:Sandy Bridge-E

[Anonymous Coward]

m.dillon, the discussion was about the i5 2400... if it was about the K-series/HT CPUs then I would've agreed ;).

Re:Sandy Bridge-E

[robthebloke]

Most programs can't saturate the FPU anyway.

Well yes. No one uses the FPU anymore.

It's a rare program that can keep putting data in on every cycle so it makes sense to have one FPU per two cores.

No, it's actually very easy. Avoid dependency chains.

If your compiler's any good the second loop will run about twice as fast as the first because it keeps the FPU busier

Uhm. Wut? It may run faster, but not for the reasons you state.....
1. If your compiler is any good, they will run at the same speed. Compilers aren't stupid - a good one would move the computation on the SSE/AVX regs. A semi-good compiler would find the second version harder to optimise. KIS-KIS!
2. Continually assigning a value to the same variable is a big performance no no!! (which you are doing for both i and the total).
3. The second version is a classic case of a naive optimisation that causes a buffer overrun. eg, #define arraySize 1
4. Neither version makes efficient usage of the FPU. "Load -> Add -> Store" is bad. "Load -> Do Tonnes of Stuff -> Store" is always better. The time spent reading/writing that float (from the array) eclipses the time spent adding
5. If arraySize is sufficiently large, the only thing that would improve performance would be pre-fetching/streaming (or striping if you are really desperate).
6. By unrolling the loop you have halved the number of integer comparisons. This has *nothing* to do with the FPU at all.

The effect is much more noticeable when the code has lots of divisions and square roots.

7. Incorrect. If you are doing lots to that single float within a tight loop, the saving of the integer comparisons becomes less and less noticeable.
8. As the number of instructions increases (in each iteration), so too does the size requirements for the instruction cache. If by unrolling the loop the instruction cache is exceeded, your performance will drop like a stone....

Re:Sandy Bridge-E

[TheLink]

Shut up dude.

We need AMD alive (I hope I don't have to explain why). But I'm not going to buy inferior CPUs. So someone _else_ has to buy them.

So you should stop going around trying to convince the fanboys that AMD is inferior.

Re:Sandy Bridge-E

[TheLink]

http://slashdot.org/comments.pl?sid=2416784&cid=37329250

Re:Sandy Bridge-E

[UnknowingFool]

Your response to people who point out inconvenient facts is to tell them to shut up? AMD once offered the best performance and price. That is no longer true. While I think we need AMD around to keep Intel competitive, ignoring facts does not help AMD or the consumers in any way.

Re:Sandy Bridge-E

[ustolemyname]

Read what he said about chained boot for a cluster. He said with intel, the cluster had to be booted *one at a time*, which is obviously a problem.

Re:Sandy Bridge-E

[Score Whore]

a) It was called Niagara.
b) It had no delay when switching.
c) It was slow as shit.

Re:Sandy Bridge-E

[Score Whore]

No, head said his entire cluster has to be booted one at a time. So AMD is better because it uses coreboot rather than UEFI.

If you have an environment that requires that everything goes down together and then comes back up one at a time in order, you've got bigger problems that what code executes at poweron.

The situation described is highly bizarre and would likely benefit from someone with some application architectural expertise.

Re:Sandy Bridge-E

[Anonymous Coward]

The dirt cheapest pos intel motherboard that supports both an i5 -2400k(LGA 1155) processor, sata6 and usb3 on newegg is $64.99. Enjoy starving your processor with ddr3-1066. If you want something that won't burn out in 6 months and actually lets the i5 beat a 1100T in any benchmark you'll be running in the $100 to $170 range.

For amd the dirt cheap pos board that supports a X6 1100T (AM3), sata6 and usb3 on newegg runs $59.99, a whopping $5 cheaper. Again if you want something that's not total crap, you'll run about $70 to $105.

Note: all research on boards you'd actually want was conducted on newegg with the following requirements: More than 20 reviews, scores higher than 4 eggs, supports the desired processor, at least 1 sata6 connection, and at least 1 usb3 connection. Low range was calculated by the lowest price I could find that met requirements. High range met this and an added DDR3 1866 (without that OC bs) support requirement.

Re:Sandy Bridge-E

[TheLink]

Uh you're ignoring this fact: if those fanboys ignore the facts and continue buying AMD it certainly helps AMD more than if those fanboys started buying Intel instead.

I don't plan to buy an AMD CPU for my future machine (unless AMD really pulls something amazing out of the bag - yeah right), but someone has to. If it's AMD fanboys doing the buying, what's your big problem with that?

My concern is if AMD goes poof, Intel might not be so focused on making CPUs faster.

Intel are already doing stuff like this: http://hardware.slashdot.org/story/11/08/14/1515221/Intel-To-Offer-CPU-Upgrades-Via-Software
Basically Intel can now mess about while waiting for AMD to catch up.

Re:Sandy Bridge-E

[Nimatek]

Having recently assembled a socket 1155 system, I can tell you there is no way for the CPU to shift one bit (see what I did there). The amount of force required to lock it in place is actually scary.

Re:Sandy Bridge-E

[Anonymous Coward]

Very interesting test results. For the tests where the X6 is faster, I'm thinking: "Thats exactly the things I would want to run faster!". For the tests where Intel is faster I'm thinking: "Ok, you found a benchmark that I didn't even know existed."

Re:Sandy Bridge-E

[Anonymous Coward]

Agreed, it doesn't make much sense to keep the same socket for many years. This is one of many reasons why the above "zomg, intel changes sockets all the time" mantra is bollocks.

Bollocks to you sir. I've upgraded a 3+ year, dual core Athlon X2 3800+ on AM2 socket to Phenom X4 820, keeping the AM2 ASUS motherboard and 4GB DDR2 667MHz. I sure am running this CPU somewhat below its capabilities, but I dropped $100 and what was a bottleneck of a system became very much faster and usable. I saved at least $100 over traditional mobo+RAM+CPU upgrade path - 50% savings.

The 2x core increase is great even if I didn't get 50% per core speedup and lower power usage at idle. The extra performance I could get from AM3 and DDR3 is not worth the $100 to me.

I never thought I would use this upgrade path either, until I did. I'm very happy with AMD and ASUS that I could do this.

Re:Sandy Bridge-E

[beelsebob]

You didn't know that Photoshop, DivX, h264, Windows Media, Cinebench, Blender, MS Excel, WinRar, dragon age, dawn of war, WoW, starcraft or power consumption existed? What are you doing on slashdot?

Re:Sandy Bridge-E

[Chris Burke]

Sure I know those things existed, but I just don't see what a photography supply store, a failed DVD pseudo-rental scheme, the planet the xenomorphs were found on in Alien, a food processor, an exclamation of surprise or amazement, and that other crap has to do with processor performance!

Re:Sandy Bridge-E

[V!NCENT]

I'm trying to tell you two things:
1. CPU's statistically mean shit when it comes to uptime; it's the OS. And Windows and Mac OS X and any other OS on the planet crashes every now and then;
2. You can spin shit around to make it sound like it's awesome, while it is not.

So you'd make a great PR guy, because you can make shit believable while it's not.

Re:Sandy Bridge-E

[Sycraft-fu]

http://www.anandtech.com/bench/Product/203?vs=363

Let's see here, according to this:

--All Sysmark tests
--Photoshop CS4
--DivX encoding
--x264's first pass
--Media Encoder 9
--3dsmax r9
--Cinebench single and multi-threaded
--Blender
--Ever single game tested

Not what one would call a trivial list. Oh and it uses less power while doing it.

Re:Sandy Bridge-E

[Anonymous Coward]

That was pretty fail.
X6 1100T - 189$ http://www.newegg.com/Product/Product.aspx?Item=N82E16819103913
i5 2400 - 189$ http://www.newegg.com/Product/Product.aspx?Item=N82E16819115074

amd board - 60$ http://www.newegg.com/Product/Product.aspx?Item=N82E16813138324
intel board - 65$ http://www.newegg.com/Product/Product.aspx?Item=N82E16813186211

This is the internet.. learn to link things.

Re:Sandy Bridge-E

[Rockoon]

A known liar, for fanboy reasons, may have been lying about other things too. For example, the price of th AMD CPU' is actually:

AMD Phenom II x6 1100T for $189

So he lied about both the cost of the AMD CPU and the cost of the AMD motherboard.

What else did he lie about? I cant find a single SATA 6Gb + USB 3.0 motherboard for Sandy Bridge for under $89. Hard to prove that he lied here.. absence of evidence and all that, but we know that he is a liar so we cannot accept what he says....

As for you. We know that when confronted with the fact the at least one piece of data from the liar was incorrect, you still worked off the rest of his data unquestionably. That makes you a fanboy too. When I said "we" cannot accept what he says, that obviously didnt include you, who happily did because you agreed with the conclusion.

Re:Sandy Bridge-E

[Guspaz]

Extreme or not, Intel's price range has really gone down with Sandy Bridge; their highest priced chip (and quite possibly the single fastest consumer CPU on the market) is about $300ish. And I believe their highest price Sandy Bridge Xeon is only $600ish.

Re:Sandy Bridge-E

[Chris Burke]

Well yes, and that really wrecked AMDs revenue. The were number on on terms of absolute performance and price/performance, but a distant two in sales.

It did more than that. It wrecked their entire financial structure.

For a brief period, AMD's market share was fab limited. So AMD built a huge new fab, acquiring several billion dollars in debt in the process. No big deal, given the possible revenue the extra capacity could supply. Fabs are expensive. It's part of the game.

Once the fab came online, though, AMD's market share barely moved. Their market share was now Intel-limited. This is bad, because while their fabs were formerly full and thus efficient and profitable, now they were largely empty. An underutilized fab is a money sink fab, above and beyond the debt from building it and the depreciation of the equipment.

This more or less directly led to them spinning off their manufacturing division as Global Foundries. Their debt structure no longer matched their prospects, and it was going to be very expensive to finance the next fab which they would have to build to stay in the game.

There were certainly missteps by AMD, too. They did underestimate how quickly Intel would adapt. They had been developing their own from-scratch high-frequency K9 architecture, but it was canceled late in development. Instead they just tweaked K8 and doubled the cores to four to create Phenom, which had big enough problems that when rev C fixed them they re-launched the brand as Phenom II.

It is hard to say what would have happened had Intel not engaged in market manipulation. One thing I think can be said -- Intel, too, has had their missteps, but they're less damaging because they have the financial stability to absorb them. Maybe AMD would have made the same mistakes, and still been behind Intel just as much, but if they'd been able to attain the the market share they could have, they would have had the money to ride out the problems, and been able to keep their fabs, and a lot of engineers that were laid off.

Re:Sandy Bridge-E

[QuantumRiff]

Don't forget that those cheaper AMD motherboards also usually include a decent (not great) video card.. That can run complete circles around any embedded intel card.. (for cheap PC's that you might want to play some games on)

Re:Sandy Bridge-E

[Anonymous Coward]

The problem is, 99% of all games now and well into the future will struggle to use more than three threads. Really advanced game engines might use four. In the real world, for the average user, and including power gamers, having more than four cores is a waste (even account for three game threads + VoIP client). That's reality.

Re:Sandy Bridge-E

[BlackSnake112]

I am betting that your customer hit that machine very, very hard to move the cpu. I watched a tower fall down a flight of stairs and he CPU did not move. The hard drive was doa, but the rest of the computer was fine. The case was banged up, a lot, but the computer worked after replacing the hard drive. Many people did not like the 775 or the other pin-less sockets. They do seem to work fine. I have seen more complaints about the push pin intel cpu fans then the lack of pins. I have both AMD and intel machines. I want a cpu that works well for what I am using it for.

Re:Sandy Bridge-E

[beelsebob]

So, where are the USB3 ports on it?

Re:Sandy Bridge-E

[m.dillon]

You're missing the point, though. The i7 is not a $1000 cpu. Intel differentiates their chipsets and locks the multiplier on many of them BECAUSE THEY HAVE THE PRICING POWER TO DO SO. This is why their lowest-end cpus don't have virtualization and why their highest-end consumer cpus don't have ECC, and why they have *three* different lines of SandyBridge consumer cpus (i3, i5, i7) instead of one.

In anycase, the i5 beats the crap out of AMD's high-end offering too. Just because you can overclock an unlocked AMD phenom doesn't mean that it's actually a good idea to do so. The phenoms generate an enormous amount of heat just idling when overclocked to the point where they can compete with even just an i5, while the entire i5 system (cpu, mobo, the works) is idling at 45W.

There's just no comparison. This is why Intel has the pricing power to charge people for relatively minor differences and AMD does not.

I really do love AMD, which is why nearly every machine I've bought prior to this year was AMD-based. It was cheap to upgrade, everything cost less, and you got competitive performance out of it for the life cycle of the cpus (which tended to be less than 2 years).

That changed in 2011. All of those benefits no longer apply. AM3+ is a joke when there's no point upgrading to get the best performance out of AMDs new line of chips (which aren't even out yet for the consumer). At this point in the product cycle Intel's power consumption is so low that it easily warrants the cost of a new mobo.

-Matt

Re:Sandy Bridge-E

[marnues]

Thanks for a fine example of a straw-man!

Re:Sandy Bridge-E

[Anonymous Coward]

They are 16 physical cores sharing 8 fetch/decoder units, 8 L2 pools, and 8 dual pipline FPU units. It's basically halfway between a HT setup and a "pure" SMP setup. It's AMD making more efficient use of silicon to make up for Intel's process lead. Intel may end up copying the basic design principal because it's just better engineering, much like they did when they copied hypertransport to move away from the front side bus.

Unlike the shift from FSB to on-die memory controller, it's not clear that this is a universally better and/or inevitable development.

In fact, even with regard to integrated memory controllers, Intel didn't lag because they needed AMD to inspire them. Kind of the opposite actually: Intel was too soon. It's often forgotten because it didn't get brought to market, but Intel designed, validated, and then cancelled what would've been the first IMC x86 product. It got the axe because it was a consumer CPU with a RDRAM interface, and RDRAM prices stayed artificially high due to collusion among the DRAM manufacturers, so a consumer product which could only interface to expensive RAM was pointless to try to sell. After eating the cost of an entire CPU project with $0 revenue to show for it, Intel was wary of getting burned by committing to any given DRAM interface years before they knew it would be a commercial success, and therefore stuck with external memory controllers (which have much shorter design turnaround times and lower development costs than CPUs) much longer than the sell-by date.

Anyways... Bulldozer will be a good match for some applications (highly threaded integer workloads where low performance per thread doesn't matter) and a poor match for others. Intel is pretty heavily invested in high single-thread performance, and that approach has been working quite well for them since it's useful across a broader range of workloads, so AMD's new approach will have to be quite disruptive if they're going to copy it.

(Also I think an equally valid way of describing Bulldozer is that it's 8 physical cores, where each core is 2-way hyperthreaded and has 2 clusters of integer units, one per thread. AMD itself has often been using the term "module" rather than "core" because 1 module is more than 1 core but less than 2 -- whether you're looking at HW resources or realized performance.)

Re:Sandy Bridge-E

[Lennie]

Yes, that is why it is important that people still buy AMD.

Even if AMD have to work within the limits of their abilities, this forces them to be creative and makes sure Intel does not slow down their development as happends with any monopoly. But AMD isn't completely lost and thus even Intel can take their ideas and hopefully the customers win.

Re:Sandy Bridge-E

[HappyPsycho]

Have never heard that, other than the Atoms Intel chips almost always consume more power and as a side effect tend to run warmer.

Re:Sandy Bridge-E

[UnknowingFool]

A known liar, for fanboy reasons, may have been lying about other things too. For example, the price of th AMD CPU' is actually:

First of all beelsebob did not name the source of his pricing information, you assumed it was newegg as the AC found a cheaper MB on newegg. It was not. After some research I speculate that his pricing came from buy.com as the Phenom and Core i5 prices match. If you had done your research before jumping to the conclusion that he was a fanboy, you would found this out.

What else did he lie about? I cant find a single SATA 6Gb + USB 3.0 motherboard for Sandy Bridge for under $89. [newegg.com] Hard to prove that he lied here.. absence of evidence and all that, but we know that he is a liar so we cannot accept what he says....

I cannot find this MB on buy.com but I have to admit that the website does not make it easy to search MB by feature like newegg does. But since we can't find it, I can't assume he lied.

As for you. We know that when confronted with the fact the at least one piece of data from the liar was incorrect, you still worked off the rest of his data unquestionably. That makes you a fanboy too. When I said "we" cannot accept what he says, that obviously didnt include you, who happily did because you agreed with the conclusion.

It appears that you were the one who has made conclusions on incorrect data. At the worst I may have accepted his data but that does not make me a fanboy for I have no preference for either company and have owned both. Most of my purchasing decisions are based on sales. I however did research his data and it appears he is correct if my speculation about his source is correct.

Re:Sandy Bridge-E

[Rockoon]

You wont find any USB 3.0 ports on cheap Sandy Bridge motherboards because Intel does not support USB 3.0 in its Sandy Bridge chipsets.

This fact seems to be lost of the Intel fanboys, that do not realize that any Sandy Bridge board that supports USB 3.0 (or more than 2 SATA 3.0's) must use a 3rd party support chipset..

The fanboys are delusional when they claim $65 Sandy mobos with SATA 3.0 and USB 3.0 .. they don't exist.

Re:Sandy Bridge-E

[Rockoon]

Oh wow... when faced with figures that do not allow you to continue to believe the thing you decided upon without any research... you data dredged the web and found figures that give you an "escape" and you continue to believe what you believed before you had any data...

....even though you are already faced with hard data that disprove that preconceived fanboy notion of yours...

Re:Sandy Bridge-E

[toddestan]

For some reason though, they are still selling that LGA 1156 Gulftown Extreme Edition for $1000. I have no idea who would buy it at this point at that price, but I guess people are.

Re:Sandy Bridge-E

[Anonymous Coward]

You're right of course. It costs $90-100 for a Cougar Point (SB-capable) mobo w/ USB3.0 and SATA3. Sixty dollars usually buys a shitty H61/H67 mobo meant for HTPCs. Cheap mobos often lack USB3.0 headers too.

I'm buying a $180 Gene-Z soon (for x8/x8 lanes) so maybe I shouldn't argue about outrageous prices....

Re:Sandy Bridge-E

[UnknowingFool]

Oh wow... when faced with figures that do not allow you to continue to believe the thing you decided upon without any research... you data dredged the web and found figures that give you an "escape" and you continue to believe what you believed before you had any data...

What part is incorrect: (1)That he did not list where he got his price? (2)Or that on buy.com the prices for the processors are exactly as he named? (3)Or that you did not actually look on buy.com before calling him a fanboy? (4) Or that you assumed he got the prices from newegg and that he was lying? As for dredging the web, it took like 2 seconds to found out where he may have gotten his price information.

....even though you are already faced with hard data that disprove that preconceived fanboy notion of yours...

Please, you jumped to conclusions as soon as you read his post. You based your assertions on assumptions which turned out to be wrong. Hard data is that he is correct according to buy.com. Now he didn't use the cheapest price for all parts but then again he didn't say he found the cheapest price.

I see when faced with the possibility you may have jumped to conclusions, you resort to name calling instead of accepting the fact that you might be wrong. The other thing is that you don't know anything about what processors I prefer or how many AMD or Intel computers I have. The fact of the matter is I don't have a preference; the last chip I got was an Intel because the chip and MB were on sale making them cheaper than a comparable AMD and MB.

Re:Sandy Bridge-E

[Daniel Phillips]

the immaturity of AMD's 32nm process and things could be expensive for AMD on the desktop...

True, however do bear in mind that AMD's fab strategy pits TSMC and friends against Intel. In the long run, I'll bet on the former.

Re:Sandy Bridge-E

[Daniel Phillips]

And by the way, I'm running AMD on the desktop right now, way more machine for my money than Intel, cool and quiet too. The way things are shaping up it seems likely to be AMD on the desktop for the next generation as well. Lots and lots of cores, just serve 'em up please :-)

The previous four generations before Phenom II were all Intel for me so it's not like I'm an AMD fanboy.

Re:Sandy Bridge-E

[Daniel Phillips]

My understanding is that AMD is cheaper but normally requires more power and cores to an Intel chip. Electricity costs then offset any price difference...

OK, that does it. I shut down my 4 way Phenom II box and put my killawatt[tm] on it. Right now, posting to Slashdot under the latest and greatest KDE... 79 watts from the wall. I challenge you to make the same measurement with your four way Intel box.

See, AMD and Intel both claim similar power draw. From what I have seen, only one of them tells the truth. Please correct me if your actual measurements indicate otherwise.

Re:Sandy Bridge-E

[Anonymous Coward]

Your words:

For now AMD is superior, because a server reboot requires about 1/100th of the time that it takes an Intel CPU farm to get back up due to horrible BIOSes.

We are in a discussion about server CPUs. So what fucking relevance has boot time to CPU superiority when reboot time is 1:hundredsofthousands to run-time? And besides, you rather run servers with hackish BIOSes to save some seconds a few times a year that also run slower the other time? I hope you're just some basement dweller with no real responsibility.

Also: quoth coreboot.org:

Fast boot times (3 seconds to Linux console)

Are you seriously trying to tell us that Intel boards take 3s*100 = 5 minutes to boot to a Linux console?

You fail at arguing and instead you're pulling some random crap out of your ass in order to look credible or whatever. Funny man you are.

Re:Sandy Bridge-E

[Phoghat]

Well, if Intel's CPU is more expensive, it MUST be better ?

Right ? Errr...right !

Re:Sandy Bridge-E

[Life2Death]

I'd like to know what you're smoking.

$1200 for the 6-Core Xeons still, best I could find.

Re:Sandy Bridge-E

[toddestan]

This is how you get a Gulftown processor for $1000: http://www.newegg.com/Product/Product.aspx?Item=N82E16819115079

Knowing Intel I doubt you could use it in a Xeon motherboard (in any case I don't think it does ECC). Since it's Socket 1156 you're stuck putting it into a last-generation motherboard.

Re:Sandy Bridge-E

[Anonymous Coward]

Over 100 of them were Intel boxes, that after a customer bumped their knee on the tower, it didn't work.

You serious? You're actually claiming that ONE IN THREE computers that fails was just bumped? I've been in computer repair since before the LGA775 came out and I have seen NO changes in CPU-related failures compared to the older designs or to current AMDs. I've seen no news items to make me think any different, either.

If you're seriously seeing that many CPU mounting failures in your own work then I'll make sure never to buy from you as you're either using sub-substandard boards or doing it horribly wrong.

Re:Sandy Bridge-E

[cynyr]

What x264 settings and source material were used? I'd like to try it on my computer. I bet my 1055T (undervolted to be ~95w) can do 95FPS on 480P video using the x264-preset-fastfirstpass profile.

Also I like how the "x4 980 BE" beats out the "x6 1090T" Seems the benches don't make good use of memory bandwidth and more than 4 cores.

x264 with the lossless_slower ffmpeg preset is about the only thing that will cause my cpu fan to spin up here. ohh and compiling chromium (the browser).

Re:Sandy Bridge-E

[cynyr]

I don't see any "server" like loads in there. I also see no benches that put the 6core 3.3 GHZ over the 3.8GHZ 4 cores. This suggests that those benchmarks do not scale above 2-4 cores. which by the way is about the max (resonable afordable cores) you can get from intel. Can't you buy a 12 core opteron these days?

Re:Sandy Bridge-E

[cynyr]

how many cores are in the xbox360? there are only 2 main cores in the PS3 as well. the WII has a dual core in it as well AFAIK. hmm interesting that games sem to be built for the "lowest common denominator".

Fun experament i ran after getting a X6 1055T, fired up WoW in Stormwind, looked at my FPS while i rode about on the ground(Cata), then i fired up a chromium emerge with a make -j4 options. Sure enough it started to use up 4 cores, but i had 2 left to play wow and my FPS moved down by a few from ~50 to ~45 or something like that. There was a bit of a fight for my 7200 RPM HDD, but that was to be expected.

Fire up a 4 threaded x264 encode, and then your game on both CPUs, let me know which is more playable.

Re:Sandy Bridge-E

[cynyr]

I usually point out the "threaded" benches that put the 3.6GHz X4 ahead of the 3.3GHz x6, and explain that the bench is not scaling well.

Re:Sandy Bridge-E

[Anonymous Coward]

> a failed DVD pseudo-rental scheme

It's not that DivX. Turn in your geek card without delay.

Re:Sandy Bridge-E

[Anonymous Coward]

Whoosh... Read the other examples, then turn in your joke detector without delay!

The price/performance ratio.

[Chrisq]

What will be interesting is the price/performance ratio compared to the Intel chips. This chip will be typically used in server farms, and this will be at least as important as the raw power - though obviously there is an overhead in running more servers. AMD has usually been ahead of Intel, and it still is on most mid-range and low-end chips, but it has started to fall behind at the high end.

Re:The price/performance ratio.

[aliquis]

AMD has usually been ahead of Intel, and it still is on most mid-range and low-end chips, but it has started to fall behind at the high end.

What? AMD got mid-range chips now?

Now that is trolling. It's funny trolling though :D

Re:The price/performance ratio.

[aliquis]

Shouldn't post this but whatever:

AMD has usually been ahead of Intel

Also "usually" = During P4?

I mean the K6 or whatever it was called vs the MMX both had their merits. P4 was probably worse than Athlon XP/Athlon 64 (64 bit + integrated memory controller = short lead for AMD). Pentium-M was a step in the right direction and since Core 2?

I would buy AMD if they where competitive because everyone benefit from more players and Intel is huge as is.

But to claim AMD is usually ahead of Intel? Some people would probably say the same about ATI but Nvidia had the lead with TNT, TNT2, Geforce-Geforce4 and still have it if nothing else for alternative OS support (Linux, FreeBSD, Solaris) and PhysX. Was it really ATIs superiority which took the lead over the FX-series or was it because the FX-series was a bad product which gave them the chance for a short while?

The future is open for both of them and both can become the leaders. But as of right now I would still pick Intel + Nvidia.

Re:The price/performance ratio.

[sco_robinso]

AMD has usually been ahead of Intel?? At the low-end price-point maybe. For under $100, AMD has usually offered the more competitive options. But rarely has AMD really been ahead of intel in any other market segment.

The Athlon XP was more or less on par with the P4, P4A, and P4B. I bought into AMDs architecture at the time because it was a bit more bang for the buck, and excellent companion chipsets for gamers and enthusiasts (nForce and nForce 2). They couldn't stand up top the P4C's, so in comes the Athlon64. The Athlon64 compared to the later P4's / PD's were indeed competitive. Hence, AMD did really well during this time, and most would argue this was really the only time when AMD was truly "ahead" of intel. Their financials reflected as much.

Post-core has been another story. Since Core2 (Conroe), Intel has pretty much dominated. The only times I really see AMD doing well is, again, the low price point segment. This is partially because they always shove another core in, so 3-core chips end up getting compared to 2-core chips.

AMD is certainly competitive, always has been... They've often offered slightly better bang for the buck, but saying AMD has 'usually' been ahead of intel is a big, big strech.

Re:The price/performance ratio.

[aliquis]

I wasn't an Intel fanboy back when I thought the P4 was worse.

And I'm not now.

I just buy what's best for my purpose.

AFAIK Bulldozer will show up in the desktop to.

I honestly don't even know what's the major difference of AMD and Intels server vs desktop chips. As long as we speak Opteron and Xeon atleast.

Re:The price/performance ratio.

[afidel]

From 2005 to 2010 in real world server applications, yes AMD beat Intel on price/performance. Then Nehalem was release and for most workloads AMD got beat like a redheaded step child. They battled back for large multithreaded server workloads by adding more and more cores but they still get beat silly for anything where single thread performance matters (a whole bunch of things).

Re:The price/performance ratio.

[afidel]

Only if you're looking at the desktop market, in the server market Intel really couldn't compete till Nehalem with the introduction of QPI, at least for most of my big workloads.

Re:The price/performance ratio.

[aztracker1]

I would go so far as to say, especially since the AMD buyout, and use in integrated video chipsets, that they have been ahead for low/id range desktops and laptops where integrated graphics are used. Even now, the AMD chipsets for graphics are better than the Intel gfx glued into some i3/i5 solutions. It really depends on your use, and how you look at them as a whole. For a mid-range graphics workhorse for gaming, you can get in a bit cheaper with a high-end AMD CPU, and a mid-range mobo. Since Intel's low-mid range motherboards have limited PCIe lanes, you generally can't run dual video cards (since one of the x16 slots, if there are two, are locked to x4). Which means you have to spend a lot more, when you are mainly wanting more GPU performance. Right now, I think a higher clocked brother to the E-350 would be really nice, though the E-350 laptops have jumped in price a little (as they're really good performers), I was really happy to pass a few along at $350 for the 15" ones, now around $400. Though I think with laptops you go as cheap as you really need, and will do what you want. Desktops, I'm inclined to spend a bit more. Though my laptop is a core2 based mbp, and my desktop is an i7. I'm spoiled and have a bit more than I *really* need.

Re:The price/performance ratio.

[aliquis]

Wikipedia:
"Nehalem (pronounced /nÉ(TM)ËheÉlÉ(TM)m/[1]) is the codename for an Intel processor microarchitecture, successor to the Core microarchitecture.[2] Nehalem processors use the 45 nm process. A preview system with two Nehalem processors was shown at Intel Developer Forum in 2007. The first processor released with the Nehalem architecture was the desktop Core i7,[3] which was released in November 2008."

Re:The price/performance ratio.

[afidel]

And the server part was released in late 2009 with OEM system availability in very late Q4/early 2010, what's your point?

Re:The price/performance ratio.

[trilliwig]

Er, no. Gainestown (the 2-socket version of Nehalem) was released in March 2009, with system availability by the summer.

First to market is relevant, but not everything

[drinkypoo]

If AMD can kick out a believable press release stating that they'll have the 8-way chips out in a reasonable amount of time, then they'll be fine.

I had such high hopes for Bulldozer

[Anonymous Coward]

Especially after Intel started shipping hardware DRM and remote killswitches in Sandy Bridge..

What is the FPU performance of these things...?

[serviscope_minor]

Does anyone know the FPU performance of these things?

So comparing a 16 "core" 'dozer to a 12 core magny-cours:

The number of parallel integer (and memory addressing) threads has gone up from 12 to 16.

The number of FPUs has dropped from 12 to 8.

The new FPUs are now twice as wide with the AVX instructions.

So, two threads share one wider FPU now. If it's hard keeping an FPU full, then this should make better use of the hardware. It seems that if your code does well for parallel, scalar FPU work already, then there may be a performance drop.

If you have trouble filling the FPU for scalar work, then this should give better utilisation of less hardware. There's a possible performance increase if your utilisation is currently under 67%. Since the two core units can feed the FPU independently, there is a little latency hiding now. This could help even if there are two completely independent processes using the FPU at the same time.

I suppsose the reasoning is that there is often fine-grained parallelism to be had, and the problem of fine grained parallelism and keeping the FPU full are often independent. So AVX will improve performance there.

So, it seems that the peak FPU performance has increased in the ratio of 16/12.

The actual performance could be all over the place. It will be interesting to see.

The other thing is that these are now single chips with 8 bulldozer units on and 16ish cores. Perhaps AMD will go and make more MCMs like before, giving 32ish cores per socket :)

Re:What is the FPU performance of these things...?

[lorax]

There's a whole lot of variables in play and I don't think we'll know until people really start using this. Intel's first viersion of AVX didn't extend all SSE instructions to 256 bits, I don't know if AMD's does so you might give up performance in some workloads there. I think for desktop use this will be a real win since you rarely run two applications that both use floating point.

Re:What is the FPU performance of these things...?

[Anonymous Coward]

it is a MCM ; the single chip is 8 cores, and the MCM is 8 modules, so I guess you're mixing them up :)

you can still build a mildly affordable 64 core box with 256GB ram that way.

Re:What is the FPU performance of these things...?

[Cajun Hell]

Does anyone know the FPU performance of these things?

No.

Or at least nobody who can talk.

Next question?

:-) Sorry, but what kind of answer did you really expect? :-)

Re:What is the FPU performance of these things...?

[Anonymous Coward]

"...what we do when we need that extra push over the cliff, you know what we do?"
"You put it on 11?"
"Eleven. Exactly. One louder."

Re:What is the FPU performance of these things...?

[marcosdumay]

Yep, that is the old problem of chossing processors that did never realy go away.

You simply can't know what you are buying before you go out and buy one of each to test. And, of course, if you are buying just one or two machines there is no reason at all to test anything.

You can read the specs and go with the processor that is more likely to fit your work, but any detail could change things.

Re:What is the FPU performance of these things...?

[Anonymous Coward]

Generally every time the FPU instruction set has been widened, the actual execution resources has lagged behind for a couple of generations. (makes sense: doubling the resources when 0.001% of software can take the advantage doesn't really pay off)

But this is actually a refreshing idea in that regard. This time the wide instructions really should do give you twice the performance as long as the other thread doesn't do FPU work. Aaaaand if you don't use AVX stuff, you should get the full normal FPU performance from both threads.

Well, in theory at least. There are several ways of botching it. For instance if there's only a single 256-bit scheduler one thread may get odd phantom stalls due to the other thread's dependencies even in non-AVX-code. Two 128-bit FPUs wouldn't have that problem, but when you throw AVX to both of them simultaneously it might hurt the other thread more than your gain is in the other. Finally IDK if it's feasible to design a resource-overkill 128-bit FPU that does AVX with instruction pairs. It wouldn't have any of those nonobvious scheduling issues but the pipeline would be pretty scary..

Re:What is the FPU performance of these things...?

[Anonymous Coward]

Keep in mind that the second integer unit only adds 10% to the module size while increasing integer performance by 80% (compared to a 5-8% increase per core for Hyperthreading (SMT) which yields 30% increase at most, but can decrease performance for certain loads).

Bulldozer will be 315 mm^2 as compared to 346mm^2 for Thuban (two Thuban's to the magny cours) and Bulldozer will also have much better power management (ability to set TDP and much better clock gating). The power savings (and fine-grained TDP tweaking) alone is huge, add on a 50% performance increase for 30% more cores (while multi-threading in a typical (read: not ideal)) server and any company has cause to be excited.

As for the FPU, why is it always AVX?

AMD is also adding SSE4.1, SSE4.2, AES, CLMUL, XOP, and FMA4 (Intel lacks the latter two FPU instruction sets) all of these sets combines should give one of the single greatest FPU performance boosts in AMD's history.

It is all about the die size

[sgt scrub]

Unfortunately, and possibly because of reports that AMD is struggling to clock its Bulldozer cores to speeds that are competitive with Intel's Core i7, there's no word of the 8-core desktop-targeted Zambezi CPU.

If you increase the clock on the CPU you have to cool it. Reducing the die reduces the amount of cooling that needs to be done. AMD is not able to shrink their die. Yet.

Re:It is all about the die size

[Anonymous Coward]

Yeah, Intel's vast capital reserves mean they typically have a generational lead in process tech, and get the increased efficiency / decreased temps of a die shrink "for free". AMD have to out-innovate them just to produce an equivalent CPU, let alone a better one. Unfair, but that's how (near-monopolistic) business works I guess.

Re:It is all about the die size

[grimmjeeper]

It's more complicated than that. As you shrink die size, you have to fight all sorts of things that bleed current through paths its not intended to go. As you increase frequency, you have to increase voltage to make it work. As you increase voltage, you increase the bleed through. The better you are at fabbing a particular die size, the less bleed through you have and the more you can crank up the voltage and frequency. That's what they were talking about when they discuss AMD's problems getting their cores up to competitive speeds.

As they get more experience in 32nm, AMD will get better yields. Trouble is, they're behind in the race. While they're messing around with 32nm, Intel is forging ahead with their work, constantly staying ahead. But this isn't all bad. Without AMD nipping at Intel's heels, they wouldn't have any incentive to keep up the pace in development and we'd be stuck with slow, hot Itaniums.

Re:It is all about the die size

[Anonymous Coward]

Unless you remove some cores along with the shrinking die size, a smaller die will mean more cooling, since you are dissipating the same energy in a smaller area.

Re:It is all about the die size

[grimmjeeper]

Not necessarily. The smaller die usually means shorter paths through the chip (less resistance so less heat). Also, you can generally get higher speeds with less voltage (within reason). Less voltage generates less heat. And that's the whole point to smaller die sizes. Faster speeds, less power draw, lower heat generation.

Re:It is all about the die size

[trickyD1ck]

AMD could outsource production to those who have competitive process tech.

Re:It is all about the die size

[0123456]

AMD could outsource production to those who have competitive process tech.

Maybe they could get Intel to make their chips.

Re:It is all about the die size

[scheme]

Not necessarily. The smaller die usually means shorter paths through the chip (less resistance so less heat). Also, you can generally get higher speeds with less voltage (within reason). Less voltage generates less heat. And that's the whole point to smaller die sizes. Faster speeds, less power draw, lower heat generation.

That hasn't been true for a while. Nowadays, reduced sizes mean that the heat generation doesn't go down because the gates and chip features are small enough that voltage leaks through and generates waste heat. Intel and AMD have done a lot to avoid or mitigate this through better materials and different process technologies. However, it's still there since the TDP for processors have basically been stagnant for a while.

Re:It is all about the die size

[grimmjeeper]

TDP has remained constant but actual throughput and power has increased dramatically. The 6 core I'm running now has the same TDP as the dual core I bought years ago. I guarantee my 6 core runs circles around the dual core. So I think the rule still holds true.

Re:It is all about the die size

[0123456]

However, it's still there since the TDP for processors have basically been stagnant for a while.

Except the new-generation i5 have very low power consumption for their performance. Leakage current used to give you a CPU that could fry eggs when idling, whereas the i5s use very little power when idle these days and (unless overclocked) only seem to peak at around 50-60W under full load.

That's actual measurements, not TDP.

Re:It is all about the die size

[Osgeld]

well they could just go the intel route and tell you to find your own cooling solution, so when it blows up they don't have to warranty it

Re:It is all about the die size

[Anonymous Coward]

They are already doing this. AMD doesn't own any fabs anymore. Radeons are made by TSMC and the rest by GlobalFoundry. Doesn't seem to really help it now, does it?

Re:It is all about the die size

[shugah]

Well it certainly helps GlobalFoundry to have more than one client. Apparently GF is still scaling up (and down) it's sub 32 nm processes and facilities and will be able to give TSMC some serious competition soon. AMD can only benefit from this competition.

For the future, I think This is what we are going to see from AMD. An integrated, heterogeneous MPSoC architecture with open (or at least standardized) on chip interfaces that might allow a mix and match CPU Core, GPU, PIO and Memory. Sort of taking Fusion one step further.

Re:It is all about the die size

[Chris Burke]

AMD could outsource production to those who have competitive process tech.

Intel is the undisputed front-runner in process tech. The closest thing is Global Foundries, formerly AMD's own fabs, which it now outsources production to.

real time core

[roman_mir]

I am now using a lot of parallel programming for what I do in report generating (funny?), it's not an easy thing to do, because I am using aggregate functions upon data, so splitting data retrieval and then recombining it once it's back (and having to poll for all of the DB requests to return) is tricky to say the least.

The reports make dozens or even hundreds of calls to the DB (PostgreSQL), and that DB doesn't run SQLs in parallel on separate processors (maybe some day, but not now), so I split the data retrieval into multiple segments and then recombine the data. To me the advantage of multiple cores at this point is only found in having completely independent processes running in separate cores, and nothing else on the DB side. The project itself is in Java, so it's using pthreads, so I have more control on the front than in the DB.

So basically to be honest, until the software like databases, takes advantage of more cores, there is no value in paying to switch to the new CPUs. However I would be interested in having a CPU with multiple cores, where some of the cores could have different architecture, to guarantee real time processing.

Wouldn't it be interesting to be able to have one CPU with normal applications running, but also with an app or part of apps running in the same CPU that could do real time processing - guarantee real time response with a simpler OS running on that core?

Having more than one OS running on the same CPU, one being a real time OS, but this would require real changes to the environment around the CPU - at the minimum there would have to be a dedicated memory bus.

Of-course the same can be achieved with just separate computers, so maybe there is no point in it.

Maybe AMD needs to look into helping some projects to switch to their architecture, to fix existing code base to use multiple CPUs at the same time, maybe that would make some sense. There is definitely a need for tools that would help switching existing applications to multi-processor/multi-core environments actually to use this stuff.

Re:real time core

[Anonymous Coward]

You should take a gardening class. Gardening is excellent. Pay is not that bad and you get to work the soil, away from the sterile environment of an office. Or maybe a cooking class? Working as a chief de cusine is great and always impress the chicks!

Just stay away from IT, please. You don't have a fucking clue about what's going on.

Re:real time core

[Anonymous Coward]

Postgres uses a new backend process for each connection. Each of these will be scheduled independently by the kernel, meaning that it does take advantage of multiple cores at the level of the connection.

Since you're already threading your java, try not using a shared connection for all of the threads, and instead open a new connection in each. See how that goes.

Re:real time core

[serviscope_minor]

However I would be interested in having a CPU with multiple cores, where some of the cores could have different architecture, to guarantee real time processing.

I don't think CPU architecture has much to do with that.

Wouldn't it be interesting to be able to have one CPU with normal applications running, but also with an app or part of apps running in the same CPU that could do real time processing - guarantee real time response with a simpler OS running on that core?

Just get yourself an RTOS. These days a stock kernel with CONFIG_PREEMPT_RT defined dopes a pretty good job. There are other options (RTLinux, RTAI) which I don't know much about.

Just avoid getting a motherboard with a terrible implementation of SMBus routines. And good luck finding one of those without trial and error...

Having more than one OS running on the same CPU, one being a real time OS, but this would require real changes to the environment around the CPU - at the minimum there would have to be a dedicated memory bus

That's how RTAI and RTLinux work (or so I gather). They are essentially a micro-OS which can run a number of processes. Linux is one of those processes. Other non-linux processes communicate with Linux and each other using FIFOs provided by the micro-OS. When Linux runs, it runs whatever of its processes need running.

Re:real time core

[roman_mir]

PostreSQL can use a separate process for separate SQL executions actually, but it cannot use parallel processing for one SQL execution.

I am using a connection pool, so this is not even the question.

The problem right now is that PostreSQL does not split one SQL request into parallel pthreads or processes, so one query executes on one processor sequentially.

PostgreSQL is wonderful, I am using it for all my projects, but to overcome this problem I have to do much more on the front to speed up execution for large requests by splitting the requests into sub-requests that can be ran in parallel and that make sense to run in parallel because the data that comes back can be merged back together meaningfully.

This would be less of an issue of the database itself had ability to split the query into multiple parallel processes. PostgreSQL actually used to have this feature long ago, it came from Ingres I think but then it was pulled out due to some patenting issue.

Re:real time core

[roman_mir]

You don't think there is an issue with the common 86 architecture for real time processing? I put together an Atmel based 3d printer controller back in 2003. If the bus for that controller was shared with another processor somehow, then there would be conflicts created in resource sharing.

When I said architecture, I should have been saying: bus architecture. How do you have an RTOS running on the same bus with another OS without having to rely on messaging and sacrificing actual real time?

Re:real time core

[serviscope_minor]

You don't think there is an issue with the common 86 architecture for real time processing?

Well, not per se, but there can be a lot of strangenesses of particular implementations. Like the SMBus horror, for instance.

I put together an Atmel based 3d printer controller back in 2003. If the bus for that controller was shared with another processor somehow, then there would be conflicts created in resource sharing

I guess that's the problem with shared busses, though probably not limited to x86. Modern x86 multiple CPU architectures don't share a FSB any more, they use a high speed message passing bus (HT or QP).

When I said architecture, I should have been saying: bus architecture. How do you have an RTOS running on the same bus with another OS without having to rely on messaging and sacrificing actual real time?

Well, if the non-RTOS is pure software (of course it isn't), then it will be pre-empted happily. I suppose that the thing to do is to avoid having shared busses. This might not be as hard as it seems: most motherboards have several independent PCI and PCIe busses, though you have to dig into the device tree to find out which is which.

DMA transfers on one bus shouldn't affect the others. PIO stuff should be preempted happily unless you have a stupid piece of hardware (sadly not unusual).

But yes, sometimes it is easier to make two separate computers on a single board and communicate over some noddy serial link to avoid any contention.

Re:real time core

[Anonymous Coward]

Try SQL Server or Oracle ... ducks

Seriously, who cares about politics if you have a job to do.

Didn't just yesterday you said

[Anonymous Coward]

the next generation is on hold?

Re:Still no benefits to multi-core CPUs...

[Anonymous Coward]

If only I knew someone who had a use for browsers and operating systems...

and amd has more MB choice will more pci-e lanes

[Joe_Dragon]

and amd has more MB choice will more pci-e lanes on some of the boards. While on intel you have to take some of the x16 lanes for video or use the x4 DMI bus to fit in USB 3.0 sata and other stuff.

Re:and amd has more MB choice will more pci-e lane

[beelsebob]

Except that USB3 and SATA 6Gb/s are all on the chipset, so you're bullshitting. What are you, as a desktop user (note, server boards have QPI links and hence more PCIe bandwidth) actually going to do with that extra PCI bandwidth?

Re:AMD SERVER CPUS??

[Anonymous Coward]

Cray?

InterLAGOS CPU Family?

[billstewart]

Great - My CPU will start generating kernel messages telling me it that it's the SON OF THE TECHNOLOGY MINISTER OF NIGERIA and that it has 25 GIGAFLOPS of PROCESSING that it needs help smuggling out from behind seven proxies and I can have 25% of it in return for overclocking the CPU by 10%.

Re:AMD SERVER CPUS??

[Anonymous Coward]

AHAHAHAHAHHAHAHA. AHHAHAHAHAHAHAAH. Who in their right mind would use AMD chips for servers???? They're crash prone junk. Always have been, always will be.

Uh...AMD has been leading the server market for a really long time. Intel has solidified desktop dominance since the core duo came out, but they're far behind on server farms.

hogwash!

[Anonymous Coward]

Show me Benchmarks, or it didnt happen.

Re:and amd has more MB choice will more pci-e lane

[Rockoon]

Wrong. There is absolutely no USB 3.0 support on P67 and H67 (Sandy Bridge) chipsets, and it only supports two SATA 6Gb ports.

Furthermore, only one PCIe x16 2.0 is supported by this generation chipset.

The fact that you are not aware of how stunted the Sandy Bridge chipset is with regards to I/O, is telling... fanboy much?

Is high performance really an issue?

[gman003]

I've never really considered AMD the manufacturer to look towards when looking for high-performance stuff. In my mind, at least, they're the "dirt cheap and good enough" side - I bought a triple-core Phenom for about the price of a low-end Core 2 Duo a year or two back. They've always had the best performance per dollar. Sometimes, yeah, they did even have the best absolute performance, but Intel's back in the lead again.

High performance just isn't a very profitable market segment. Gamers and high-end servers will buy it, but that's not where the big market is. The big market is desktops and laptops - stuff where a 4gHz sextuple-core processor is overkill. A business machine will work fine with half that - and with AMD's price advantage, they've been moving in on business and desktops. Supercomputers might also be enough to sustain the company - they buy by the thousands, and AMD's power efficiency and multi-core design has usually been attractive to the few in that business. There, performance per core isn't nearly as important as cores per watt.

That said, I'm not surprised that AMD is (supposedly) having issues meeting their targeted clock rates. Pre-release info pegged the top desktop processor at 4.2gHz - a record for an x86 processor. The last to get close to that was the last few Pentium IV HTs at 3.8gHz. AMD's top processor to date only reached 3.7ghz (Phenom II X4 980BE), and that was after years of refining their process. AMD set their sights too high, and is having problems for it.

Re:Is high performance really an issue?

[adamchou]

I don't know the actual numbers when it comes to sales of servers vs. sales of consumer pc's but the profit margin on server chips is definitely way higher than it is on consumer chips.

what a silly one-sided summary

[Anonymous Coward]

This is an announcement about a new AMD product with makes AMD "best" for the time being. They deserve some praise for the achievement seeing how they are the underdog. So this is the best article selected by slashdot, where the entire article consists of AMD bashing ?

Here, let me summarize the post for you "These slow-pokes at AMD managed to put out some supposed 16-core CPU, but we will discuss what they have not released yet which will certainly be beaten to a pulp by Intel's greatest and also unreleased CPU which will be released faster".

Seriously ?

Re:and amd has more MB choice will more pci-e lane

[Anonymous Coward]

Anything I want to, what kind of question is that? really, what am I, as a desktop going to do with ANY extra bandwidth be it PCIe, RAM, disk, internet connectivity. It doesn't matter what I do with it, I PAY for that availability.

wrong

[unity100]

AMD didnt have intel by the balls then. by then, intel just paid money to pc makers to use their chips. they got fined for it, but not enough. they got a net benefit from the entire affair. and negated amds competition by such whoring.

Re:Still no benefits to multi-core CPUs...

[slater.jay]

Games are increasingly becoming thread-aware. There's a video from Gamescom showing DiRT3 running on a prerelease Zambezi chip, using four cores. Rise of Flight is heavily multithreaded (to the point where (until the last patch) it was bouncing up against a problem with the Windows scheduler on dual-core processors. There are probably more examples, but I can't think of them off the top of my head.

Re:and amd has more MB choice will more pci-e lane

[aztracker1]

The issue seems to be that on Intels mid-low end CPUs the motherboards are limited in number of PCIe lanes available, and often the second PCIe x16 slot is locked to 4x, meaning that dual GPUs aren't a good option... so you have to spend a lot more on the Intel side of the fence. IMHO a faster single-GPU is probably better spent on the Intel platform at the moment. In either case, we've already hit the "fast enough" stage for most people... I've recommended E-350 based laptops to a lot of people, as they really are good values at their power usage. Which is nowhere near as fast as either in the comparisons we are making here. One more generation of GPUs and a single card may well run a Triple 30" setup pretty well for gaming. I just don't see the need to push things much more than that. What this means to me, is that the really high end CPUs in 2 more generations will likely be more expensive and used mostly in servers. That's just where I see all of this going.

FSB, FSB, FSB!

[Anonymous Coward]

Just want to remind people that, since 1996, AMD still uses motherboards with over TWICE the FSB speed of Intel boards. For this one simple reason, even though Intel consistently produces faster processors, AMD systems consistently out perform Intel boxes overall. In 15 years, this has always been the case, and nothing about any announcements this week change that.

Re:FSB, FSB, FSB!

[Anonymous Coward]

Just want to remind people that, since 1996, AMD still uses motherboards with over TWICE the FSB speed of Intel boards. For this one simple reason, even though Intel consistently produces faster processors, AMD systems consistently out perform Intel boxes overall. In 15 years, this has always been the case, and nothing about any announcements this week change that.

I hope you're joking. Neither Intel's nor AMD's processors use FSBs any more. At all. Haven't done for years.

Also, deep in the mists of time when both of them were using FSBs, I don't think it's true at all that AMD's FSB was clocked twice as fast.

Re:and amd has more MB choice will more pci-e lane

[Joe_Dragon]

but on the low end boards you have to cut into video pci-e just to fit USB 3.0 in and even if it moves to the chip set the x4 link for cpu to chip set is will have to have network, sound, sata, usb, x1 pci-e slots all running over it and to get more pci-e lanes you have to get a high end i7 cpu.

With amd you can get a low-end to mid range cpu and get boards with lot's of pci-e io and even the low end chip sets has.

on a intel board useing a x4 cable card tuner eat's up a lot of the pci-e io.

Re:AMD SERVER CPUS??

[elbonia]

How about the number 3 super computer in the world?

(DOE/SC/Oak Ridge National Laboratory)

http://top500.org/system/10184

Not only that

[Sycraft-fu]

But let's stop pretending that Intel's highest end chip is the only one to talk about. Yes, Intel has, for a long time, had a chip for people with more money than sense. They make an ultra high end chip for $1000 that is only a tiny improvement over the one below it. It is for sale to people who buy for bragging rights, more than anything else. That would be the i7-990X right now. 3.46GHz, 6 cores.

However right below that is a chip with near the performance but around half the cost. Right now that is the i7-980. 3.33GHz, 6 cores but only $600. 96% of the speed, 60% of the cost. They scale down quickly from there too.

Trying to compare Intel's ultra high end to something more reasonably priced says only that you are trying to stack the results.

Re:Not only that

[HappyPsycho]

Not if you are picking the top of the line of both companies.

Re:Not only that

[Sycraft-fu]

Sorry, but that is fanboy logic. "Oh we get to pick the highest end because that makes the comparison look bad!" If your concern is only "the highest end" then price goes out the window. It is valid to say "Who has the fastest chip, price is not an issue." However you can't insist on a highest end comparison and then start whining about price.

If price is a factor the question is then who offers the best performance at a given price point. So say you take AMD's high end desktop right now, which is the 1100T. That's $190, so it's price point competitor from Intel is the i5-2400. Well having a look at some benchmarks (http://www.anandtech.com/bench/Product/203?vs=363) it is indeed a faster chip.

Now Intel makes faster chips still. The i5-2500 and -7-2600 are a bit faster and for highly multi-threaded tasks the i7-970 and i7-990X can be faster still since they have more cores, though the efficiency of the Sandy Bridge chips narrows that. So if you need higher end performance in the desktop space, Intel is your only option at present, AMD caps out at about the $200 mark.

So, that's where you compare things, if a real comparison is what you are interested in in terms of value for money. Trying to skew things to make your chosen platform look good is not useful, it is just fanboy silliness.

And that's the real rub

[Sycraft-fu]

It isn't a question of if you can find a benchmark that the AMD processor does better, it is a question of how it does overall and the AMD processor and Intel processor compare. That Anad benchmark shows the answer is not well. For example the i5 is ahead in x264 encoding. Ok well that is a completely parallel activity, it will use all cores 100%. So being that the AMD processor has a slight clockspeed advantage and 2 more cores is should stomp the i7, be 50% faster. Instead the i7 bests it slightly.

Also another problem you run in to is not everything goes for tons of cores. You'll find apps that can only use a few threads effectively. Games are often this way. So going from 4 to 6 cores gains you nothing for them. That means if the individual cores aren't powerful, having more is not the answer. You need better performance per core and per clock, not just more cores.

Re:And that's the real rub

[Anonymous Coward]

For example the i5 is ahead in x264 encoding.

I'd re-examine that claim. 1st-pass for x264 used to run at ~50% load (2010). 2nd-pass did the actual work. Benchmarks above correlate with that discrepancy.

I wouldn't argue

[Anonymous Coward]

I found that going from a hard drive to a fast SSD made far more difference than going from dual core to quad core ... I'm running a heavy desktop load now, Thunderbird plus Chrome with 30 tabs open. (Kubuntu 11.04) The CPU is running 20%. I've consided going to a hex core, but I'm really unmotivated about it.

Re:AMD SERVER CPUS??

[trilliwig]

What? AMD is hovering around 6% market share in the server market, with Intel making up the other 94%! Intel has been dominating there ever since they were the first to introduce a quad-core server processor (Clovertown,) back in November 2006. Intel's average selling price (ASP) is also 33% higher. AMD's chips just do not perform anywhere near as well as Intel's on most server tasks. AMD really needs Bulldozer to improve their competitive position if they want to remain a player in the server processor market.