How much is AMD behind.

[Obsoleet]

Perhaps not making fun of customers, but a highly educated discussion about AMD's epic, unforgivable, vomit inducing shortcomings can only benefit everyone.

Oh, and to the OP, AMD is not at all behind, and I truly mean that. In order for a company to be behind, they have to be a participant in the race.

^ This is one of those strange, perverse posts that read like AMD raped his mom back in 2001. I think there's some hard feelings on the Intel side. Very unusually strong feelings over bashing a company that has done wonders throughout time considering its resources, and I like their products and keep up on them not because I like technology but because most, if not all of their stuff is great to decent.

 

[Hypertag]

Still doesn't warrant how rude members of this forum are to each other. I honestly can't take most of this forum seriously with all the childish bickering. Bulldozer's performance has been made abundantly clear by the benchmarks and discussion that took place shortly after release. I think we can all agree that it wasn't exactly what AMD was hoping for. Do we really need an entire thread devoted to making fun of AMD and their customers for being behind Intel?

That isn't what this thread is though. Its literally a thread in which everyone is claiming that AMD is barely behind, and will shortly regain a performance advantage over Intel with piledriver. So there isn't really anyone making fun of AMD customers here because this thread contains the only persons in the world that are convinced that the AMD is going to shortly reign supreme over the tyrannical Intel monopoly.

 

[Ferzerp]

That isn't what this thread is though. Its literally a thread in which everyone is claiming that AMD is barely behind, and will shortly regain a performance advantage over Intel with piledriver. So there isn't really anyone making fun of AMD customers here because this thread contains the only persons in the world that are convinced that the AMD is going to shortly reign supreme over the tyrannical Intel monopoly.

Not everyone. The same 5 or 6 people dogpile any thread that touches on AMD. The more sane people vacate because these 5 or 6 people have shown time and again that intellegent discourse will not happen once they appear in a thread like this. Then those same 5 people try to create an appearance of a favorable towards AMD consensus once the thread devolves in to an echo chamber where all that is left are the AMD fanboys (or whatever we should label them as...) and one or two people who feel like arguing with a wall. Luckily, no one buys it, but it happens over and over.

 

[taltamir]

Generally people are really nice in this forum, because if they aren't they get a vacation from the mods.
Report any post you see that is violating forum rules.

 

[moonbogg]

^ This is one of those strange, perverse posts that read like AMD raped his mom back in 2001.

You didn't have to go there with me, did you? You seem smart enough to make a point about my postings without resorting to language or resorting to insulting attacks like this. I'm just having fun in the forum.

 

[TuxDave]

Fine then, on the high end there is the use of lower quality integrated heat spreaders on IVB then on SNB. As well as the ever decreasing multiplier and power consumption on the desktop parts.

I'm not speaking on behalf of any company but this is what I observe. Lower power consumption for desktop parts is market driven and is taken into consideration when defining the project. Bad engineers make products that no one wants.

I said they hobble PERFORMANCE not EFFICIENCY.
They have been sacrificing performance for power efficiency and lowered costs (smaller die, cheaper IHS, etc).

And its not engineers who come up with those kind of decisions.

Which is exactly why they hobble performance.

Efficiency and performance goes hand in hand. Better efficiency leads to better turbo which leads to better performance. If you want a high performance server chip, there will be one. If you want a low power mobile chip, there will be that too. Don't look at a low power chip and claim "look how they gave up on performance!" As for who makes the decision. Guess what, die size is a debate between design and the fab engineers. So go figure that out.

No engineer that I know has the attitude that they want to look better on the next project by doing a crappy job on the current project. And lowering costs is an engineering goal whether you beleive it or not. I don't want to make a chip with craptacular yields, extra-expensive process, giant die size and costs a ton because that's just bad design.

I don't know if you are coming with some bad experience with working with terrible design teams but from my experience, I see a lot of ambitious people who come up with features that have good performance/power ratios. I probably said enough in this thread and probably have to abstain from any further comments.

 

[taltamir]

Efficiency and performance goes hand in hand. Better efficiency leads to better turbo which leads to better performance.

Except turbo is much much lower then what it can safely be run as. Intel OC headroom has been steadily increasing with every CPU since the core architecture first came out because they are simply lowering the clocks due to lack of AMD competition so its just slightly faster then last years model.

And because people wised up to it intel went and hobbled OCing (then starting selling unlocked K units for extra cash).

No engineer that I know has the attitude that they want to look better on the next project by doing a crappy job on the current project

Which part of "this is not the engineer's decision" did you not understand?
This is a MARKETING decision overruling engineering decisions.

There are 3 things intel is doing
1. Ever increasing OC headroom combined with lowering power consumption ON THE DESKTOP indicate that they are hobbling clockspeed due to lack of competition.
1b. Overclocking blocked in non K parts as people wisened up to the above.
2. Shrinking die size (physical) indicates hobbling due to lack of competition.
3. Lower quality accessory components (Integrated heat spreader) lowers cost and is no big deal due to massive OC room.
4. Removal of features (combined with that "upgrade CPU bought from OEM by typing a code to unloick components"). Again, all due to lack of competition.

Intel is NOT and never has had engineers just decide on their own to sabotage the chips POTENTIAL by not implementing efficiency improving designs.
What intel MANAGEMENT and MARKETING is doing is ordering that those chips be modified in the above 4 ways. Most of which (lower default clocks / cheaper IHS) are trivial. Others (no OC without K parts) require specialized engineering.

If AMD suddenly released a massively faster part, intel will be able to start production of a higher power higher clocked IVB within hours.

 

[baverdt]

I love how people are throwing sysmark numbers around and talking nonsense all the time. Since I don't like repeating myself over and over again here is the post I made about FX8150 vs Westmere 980x:

Ok again,some facts instead of fiction. First of all,do you consider a 980x Westmere to be a fast modern multicore chip that does good in both single and MT workloads? I bet you do. Westmere is still a fast desktop chip. Has 6 cores/12T and has a high Turbo clock @ stock.

Now let's see how FX8150 @ stock compares to Westmere 980x @ stock.
Numbers can be found here. This is what they used in their tests:
Single-Threaded Efficiency Run:
980x - 9:32 or 572s
FX8150-11:13 or 673s

980x is : 572/673=0.85 or 15% faster in pure single thread workloads (all time values added up;results are total run time and lower is better naturally).
So 15% higher single thread performance of Westmere is a "huge performance gap" somehow? Didn't think so.

Move on to MT runtime.
Multi-Threaded Efficiency Run:
980x - 13:49 or 829s
FX8150-16:31 or 991s

980x is : 829/991=0.836 or 16.4% faster in multi-threaded workloads.
We have a "fat" 6 core chip that supports SMT(12T) versus a "slim" 8 core chip with only 4 FP units(8T) that should be weak in single thread workloads and somewhat strong in MT workloads. What we get is 15% higher single thread performance and 16.4% higher MT performance for "fat" core. FX is not that slow after all.At least if you compare it to Westmere.

One more comparison using hardware.fr numbers is here .
Compared to top IB 3770K,in applications FX8150 is 183.7/150.7=1.22 or 22% slower ,stock vs stock. This is not bad at all since we have "slow" 32nm FX "8core"(which actually has 4 floating point,8 threaded subunits) versus 4C/8T IB @ 22nm. 22% slower in desktop workloads is peanuts gap and can be closed with Vishera with little to no problems.

As can be seen from above,facts are facts and fanboy fiction is fanboy fiction. Let's stick with the facts and this forum may be regarded(one day) as a serious tech forum,not a fanboy lounge .

What is a fat core? Google thinks you are making this term up, since the only link that seemed to be related led me back here.

 

[taltamir]

What is a fat core? Google thinks you are making this term up, since the only link that seemed to be related led me back here.

he did make up the term fat core, but its not making up info.
Fat core obviously means a physically larger die. The larger a die is, the more expensive it is to produce and the higher the waste (random defects occur per waffer and the larger each die / core is the larger the damage per defect)

 

[IntelUser2000]

Fine then, on the high end there is the use of lower quality integrated heat spreaders on IVB then on SNB. As well as the ever decreasing multiplier and power consumption on the desktop parts.

Thermal interface material that's less capable of handling higher power usage = Correct

Lower quality thermal interface material used = incorrect


Intel takes significant majority of CPU development cycle to make sure that it works reliably, and without error for long periods of time. You may or may not have seen the research from Intel that states possible issues arising from soldered TIM interface they have been using for years.

The problem may be compared to a situation in construction when everything is made to a perfect situation. Ideally you can leave absolutely no space between two not flexible objects and you'd have no problem. But in real world, there's humidity, different temperature and weather conditions that makes materials expand and contract(even metals!) that will cause serious problems over periods of time.

 

[carop]

If we turn to gaming the Hardware.fr basically shows that the FX8150 needs to perform 70% more to reach IB.

Your 70% estimate is close to the SPEC CPU2006 "speed" benchmarks.

There are several advantages about using standard benchmarks for performance evaluation. First, you can trust the data. Second, it is run according to benchmark kits developed by a committee. Third, it is audited.

SPEC CPU2006 consists of "speed" and "throughput" benchmarks for integer and floating point performance.

The components are at the following URLs:

http://www.spec.org/cpu2006/CINT2006/

http://www.spec.org/cpu2006/CFP2006/

The following are some "speed" and "throughput" benchmark results for Sandy Bridge and Bulldozer processors at the same clock rate:

AMD Opteron 6274 2200 MHz 32 cores, 2 chips, 16 cores/chip SPECint2006 = 28.7
http://www.spec.org/cpu2006/results/res2012q1/cpu2006-20120116-19346.html

Intel Xeon E5-2660 2200 MHz 16 cores, 2 chips, 8 cores/chip, 2 threads/core SPECint2006 = 48.3
http://www.spec.org/cpu2006/results/res2012q3/cpu2006-20120613-22875.html

The integer "speed" of Xeon E5-2660 (Sandy Bridge-EP) is 168% of Opteron 6274 (Bulldozer).

AMD Opteron 6274 2200 MHz 32 cores, 2 chips, 16 cores/chip SPECfp2006 = 42.8
http://www.spec.org/cpu2006/results/res2012q1/cpu2006-20120116-19345.html

Intel Xeon E5-2660 2200 MHz 16 cores, 2 chips, 8 cores/chip, 2 threads/core SPECfp2006 = 80.7
http://www.spec.org/cpu2006/results/res2012q3/cpu2006-20120613-22877.html

The floating point "speed" of Xeon E5-2660 is 188% of Opteron 6274.

AMD used to have better "throughput" results. However, Intel is better now.

AMD Opteron 6274 2200 MHz 32 cores, 2 chips, 16 cores/chip SPECint_rate2006 = 464
http://www.spec.org/cpu2006/results/res2012q1/cpu2006-20111219-19201.html

Intel Xeon E5-2660 2200 MHz 16 cores, 2 chips, 8 cores/chip, 2 threads/core SPECint_rate2006 = 597
http://www.spec.org/cpu2006/results/res2012q3/cpu2006-20120613-22882.html

The integer "throughput" of Xeon E5-2660 is 128% of Opteron 6274.

AMD Opteron 6274 2200 MHz 32 cores, 2 chips, 16 cores/chip SPECfp_rate2006 = 353
http://www.spec.org/cpu2006/results/res2012q1/cpu2006-20111219-19200.html

Intel Xeon E5-2660 2200 MHz 16 cores, 2 chips, 8 cores/chip, 2 threads/core SPECfp_rate2006 = 460
http://www.spec.org/cpu2006/results/res2012q3/cpu2006-20120613-22878.html

The floating point "throughput" of Xeon E5-2660 is 130% of Opteron 6274.

The most innovate thing in AMD's 2012 - 2013 road map is Jaguar, the Bobcat derivative. They are always late, and it seems they need to act more quickly.

http://www.eetimes.com/electronics-news/4389907/AMD-warns-of-Q2-sales-shortfall

Cheers.

 

[tulx]

AMD really is way behind

in high-end CPUs. And way ahead in APUs/iGPUs.

AMD has made it abundantly clear that high-end CPUs aren't their priority anymore.
They concentrate on APUs and beat Intel by a significant margin there.

 

[Dribble]

I'd be surprised if AMD was still around in a few years.
-They are not competitive in their core market of desktop and laptop processors.
-They are about break even in their core graphics market.
-In all those markets they are very weak in the real profit areas: servers for cpu's and professional for gpu's.

On top of that future is pretty clear, traditional laptop and desktops will loose market share to phones and tablets. The other future market is the high end server/compute one.
-AMD is not in the phone/tablet market.
-AMD is not in the gpu compute market. Yes I know their current gpu can compute but the software/drivers and capabilities required for serious (super computer) compute are so far behind their competitor that they have no hope of breaking in.

This can all change IF AMD have enough money to invest in R+D and buying companies to get into new markets, but they have no money only debt. Where does that leave them - stuck fighting in a declining market against bigger stronger competitors they can't afford to compete with. Logic dictates they are sc***d.

 

[inf64]

That isn't what this thread is though. Its literally a thread in which everyone is claiming that AMD is barely behind, and will shortly regain a performance advantage over Intel with piledriver. So there isn't really anyone making fun of AMD customers here because this thread contains the only persons in the world that are convinced that the AMD is going to shortly reign supreme over the tyrannical Intel monopoly.

Are you serious? Who said that? We are comparing PD based Vishera with 3770K,not SB-E. Piledriver won't destroy any intel chip,it will just make AMD somewhat more competitive in mainstream performance segment. If you somehow missed this then re-read the topic please.Thank you.

What is a fat core? Google thinks you are making this term up, since the only link that seemed to be related led me back here.

If you had read the topic you would see I had already explained what I meant by this. No,you just registered account to troll away.

 

[Sweepr]

That isn't what this thread is though. Its literally a thread in which everyone is claiming that AMD is barely behind, and will shortly regain a performance advantage over Intel with piledriver. So there isn't really anyone making fun of AMD customers here because this thread contains the only persons in the world that are convinced that the AMD is going to shortly reign supreme over the tyrannical Intel monopoly.

You're talking about the same guy who was predicting BD would run circles around i7 2600K, beating even SB-E depending on launch frequency short before the first benches appeared.

http://crazyworldofchips.blogspot.com.br/2011/07/20112012-clash-of-cpu-titans-sbsb.html

He's always overly optimistic about future AMD chips, trying to downplay the competition, its been like this for years now....

"K10 will beat Core 2 by 20-30% clock per clock"

http://www.xtremesystems.org/forums...Mark06-(Inq)&p=2399934&viewfull=1#post2399934
http://www.xtremesystems.org/forums...elona-Thread&p=2194270&viewfull=1#post2194270

@ Thread: Both companies have very talented engineers, but Intel has much more money to burn on R/D of new chips and process nodes, so it should have an advantage. AMD actually does some very impressive things for such a smaller company, they dont need high-end x86 CPU dominance to be competitive.

 

[ShintaiDK]

Ye, its just a rerun of the same rubbish. And the same lame excuse of "just wait a little more".

 

[inf64]

Nice to have you intel "experts" to educate the masses here .

 

[AtenRa]

On top of that future is pretty clear, traditional laptop and desktops will loose market share to phones and tablets.

Actually, Desktop and Laptop market will continue to grow

www.idc.com/getdoc.jsp?containerId=prUS23371512

-AMD is not in the phone/tablet market.

http://us.acer.com/ac/en/US/content/models/iconia-tab-w

ICONIA TAB W Series
W500-BZ467

Genuine Windows® 7 Home Premium - 32-bit version - AMD Dual-Core Processor C-50 (1.0GHz, 1MB L2 Cache) - 2GB DDR3 Dual-Channel memory - 32GB solid state drive - Acer® 10.1" HD Multi-Touch LED-backlit TFT LCD (1280 x 800) - ATI Radeon™ HD 6250 Graphics - AMD A50M Fusion™ chipset - 802.11b/g/n WLAN - dual webcams - Dolby® Advanced Audio® v2 Audio Enhancement - Bluetooth® - HDMI® - USB ports - Acer® FineTip keyboard - 2-in-1 card reader - 3-cell battery - 1-year limited warranty

 

[Rvenger]

What is a fat core? Google thinks you are making this term up, since the only link that seemed to be related led me back here.

Goes to show that even the members that just read this forum are getting upset of all the garbage that gets spewed here. I agree with you baverdt, it seems that you either get flamed here or go home anymore.

 

[Ferzerp]

Actually, Desktop and Laptop market will continue to grow

Correct. Despite the journalists' and phone obsessed's dreams, the mobile market is, so far, a totally discrete market from the traditional PC market. PC sales are increasing, and the boom in mobile sales seems to stand completely alone from it (Of course sales in the mobile space are increasing faster, it's a relatively new, unsaturated market).

Will most media report that? No, but the sales volumes show a story that doesn't align with the story presented by to those who fantasize about a "post-pc" world. As of yet, mobile devices (i.e. phones/tablets) have supplimented, but not replaced traditional PCs. In fact, I'd suggest that anyone who can actually drop in a tablet and replace their entire computing experience was treating the PC as a modern television, and never really using it to begin with.

 

[StrangerGuy]

@ Thread: Both companies have very talented engineers, but Intel has much more money to burn on R/D of new chips and process nodes, so it should have an advantage. AMD actually does some very impressive things for such a smaller company, they dont need high-end x86 CPU dominance to be competitive.

Very romantic. And totally irrelevant in consumer terms.

The only reason AMD surviving is now is Intel letting them live to avoid antitrust suits.

 

[Haserath]

I was wondering... Does anybody really care about Intel's strong-arm tactics in the not so distant past as long as they deliver better products than AMD?

 

[KompuKare]

I was wondering... Does anybody really care about Intel's strong-arm tactics in the not so distant past as long as they deliver better products than AMD?

Probably not. But if past form is any indication of what they may do again if they become even more dominant. Still I am fairly sure that Intel will never again make a NetBust size mistake so any strong-arm tactics would probably be against somebody other than AMD.

Incidentally, for those who think Intel is their best friend: why exactly did i5-2500K launch at $216 and i5-3570K at $235? Maybe the lack of competition from AMD this year was even less than last year?

 

[Remobz]

Very romantic. And totally irrelevant in consumer terms.

The only reason AMD surviving is now is Intel letting them live to avoid antitrust suits.

The only reason for AMD to exist is for competition sakes.

I use to be a huge AMD fanboy until 6 years ago.

 

[StrangerGuy]

I was wondering... Does anybody really care about Intel's strong-arm tactics in the not so distant past as long as they deliver better products than AMD?

In the "not so distant past" CPUs didn't last ~6 years for Windows and basic computing stuff and everything revolved around x86. Times have changed enough for Intel to make strong-arming tactics not work.