Sign in with
Sign up | Sign in
Your question
Closed

AMD vs. Intel

Last response: in CPUs
Share
a b à CPUs
September 26, 2010 5:46:10 PM

If you took the best cpu from AMD and the best cpu from Intel which one would come out on top?

More about : amd intel

a b à CPUs
September 26, 2010 5:52:20 PM

In desktops, Intel, as the 980X is $1000 and 1090T is only $300.

If you include server chips. the 12-core Opteron MIGHT comeout on top in specific apps.
Score
0
a b à CPUs
September 26, 2010 6:42:49 PM

Desktop: Intel (Core i7 980x is the current performance king).

Server: Intel (Xeon 6 Core CPUs trade blows but come out on top, overall, vs AMDs 12 Core Magny cours, while Nehalem-EX, Intel's 8 Core CPUs, have no rivals).
Score
0
Related resources
a b à CPUs
September 26, 2010 10:38:49 PM

Darn I was really hopin' that amd would come out on top. :(  It's also hard to beileive that a 12 core prossecor lost to a 6 core, how did that happen?
Score
0
a b à CPUs
September 26, 2010 10:45:20 PM

HostileDonut said:
Darn I was really hopin' that amd would come out on top. :(  It's also hard to beileive that a 12 core prossecor lost to a 6 core, how did that happen?

Mainly because the 12 core is clocked at 2.16Ghz while the 6 core is 3.33Ghz+Turbo.
Score
0
September 26, 2010 10:51:37 PM

Also, the 6 core has hyper threading, which can help in some circumstances.
Score
0
a b à CPUs
September 27, 2010 3:07:20 AM

What is hyper threading and wouldnt the 12 core have more ghz in the end?
Score
0
September 27, 2010 3:21:28 AM

Hyperthreading is using one core to process two threads at once, giving you about 20% more performance for each extra thread on multi threaded programs. And no, 12 cores won't have more GHz in the end. Clockspeeds dont add up like that. Instead, each core does its own work at its rated speed, like baking 12 cakes, one in its own oven.

And this is to anyone else in general: Does anyone find it absurd that mousemonkey just dug up ten of my old threads randomely to close?
Score
0
a b à CPUs
September 27, 2010 4:31:10 AM

What is a thread and does AMD not have hyper-threading?
Score
0
a b à CPUs
September 27, 2010 6:03:22 AM

If you take up computer science as a subject in your school, you'll get to know exactly what a thread is.When a program is divided into sub-programs and implemented at the same time in parallel, it is known as multithreading.Those sub-programs are called threads.
Score
0
a b à CPUs
September 27, 2010 1:37:29 PM

Okay, does AMD not have hyper-threading?
Score
0
a b à CPUs
September 27, 2010 10:29:01 PM

AMD has CPUs with more cores but no t as high ghz speed, if AMD highered its ghz to around 3 would it ever be as good as intel? If it wouldn't is there somthing else that would make it better besides the low price tag?
Score
0
a b à CPUs
September 27, 2010 10:37:42 PM

Yes but it would push the thermal envelope, which is important especially with sever chips.

The fact that AMD is behind Intel atm is because their architecture, K10 performs worse per Ghz compared to Intels Nehalem, meaning a 2Ghz Phenom performs like 1 1.5Ghz Core i5.

The only way to "solve" that is introduce a better architecture, or find a new process that cuts power-draw dramatically and push the clocks way up.
Score
0
a b à CPUs
September 27, 2010 11:58:38 PM

Which is a better bang for your buck?
Score
0
a b à CPUs
September 28, 2010 12:36:52 AM

Well, desktop wise, for mid-high gaming builds ($800) oe below, AMDs are a better value. High-end ($1200) would be Intel Core i7 territory. Builds ~$800-1200 is a wash, the X4s and i5s care pretty head-on.

Exceptions do exist, like for Low-power HTPC, the i3 wins the Athlons, while for a Design/rendering powerhouses on a budget the Phenom II X6s shine.

For servers, its really more marketing/software.
Score
0
a b à CPUs
September 28, 2010 12:46:28 AM

Timop said:
Well, desktop wise, for mid-high gaming builds ($800) oe below, AMDs are a better value. High-end ($1200) would be Intel Core i7 territory. Builds ~$800-1200 is a wash, the X4s and i5s care pretty head-on.

Exceptions do exist, like for Low-power HTPC, the i3 wins the Athlons, while for a Design/rendering powerhouses on a budget the Phenom II X6s shine.

For servers, its really more marketing/software.


+1
Score
0
a c 131 à CPUs
a b À AMD
September 28, 2010 12:53:33 AM

HostileDonut said:
Which is a better bang for your buck?

Depends on your tasks and what you do. But generally, most current CPUs under $300 are worth their cost and perform accordingly.
Score
0
a b à CPUs
September 28, 2010 2:06:13 AM

Timop said:
Yes but it would push the thermal envelope, which is important especially with sever chips.

The fact that AMD is behind Intel atm is because their architecture, K10 performs worse per Ghz compared to Intels Nehalem, meaning a 2Ghz Phenom performs like 1 1.5Ghz Core i5.

The only way to "solve" that is introduce a better architecture, or find a new process that cuts power-draw dramatically and push the clocks way up.

How does AMD perform half as well as Intel?
Score
0
a b à CPUs
September 28, 2010 2:07:49 AM

Does a core i3 beat a Athlon II X3 440 @3ghz?
Score
0
a b à CPUs
September 28, 2010 2:34:32 AM

HostileDonut said:
Does a core i3 beat a Athlon II X3 440 @3ghz?


Its hard to say a processor "beats" another. Heres your new best friend:

http://www.anandtech.com/bench/CPU/2

See what beats what, but more importantly, in what benchmarks certain things beat others. To answer a few of your question, possibly repeating the info of others, but here it goes. Core count isnt everything. Most games can only currently use 2 or 3 cores. This makes the 4th core on a quad more or less useless besides running background apps. Also, 1 AMD core doesnt equal 1 intel core. The intel cores are quite a bit faster clock for clock. The architecture just makes them this way, without going far in depth. GHz is a standard measurement, so 1 GHz from AMD is 1 GHz from anybody. It just so happens that those AMD 12 cores are clocked very low, while the intel is clocked very high. Back to the gaming example, most software only uses a certain amount of cores. AMD can have a 24 core CPU if they wanted, but using the same architecture, it would very likely perform little to no better, as most software isnt very core efficient. The i3 is a dual core with Hyper Threading. This means for every 1 core, you get 1 thread, for a total of 2 cores, 4 threads. The way intel does it, 1 thread does not equal 1 core. Fact is they share resources, and therefore HT can sometimes give a large increase in performance for apps that are threaded well, but it more generally doesnt add much to real life apps, and sometimes even takes away performance. So one would think, hey a 3 core CPU at 3 GHz should beat a dual core at 2.93? Not really. This goes back to the architecture. Intel's individual cores are more powerful, and therefore they need less cores, and at lower clock speed. Use that anandtech bench thing and pit a 970 vs 1090T. Identical core count, identical clock speed, but notice how the 970 beats it soundly? All has to do with the arch, and likely some help from HT. AMD adds more cores and clocks them higher more out of necessity than because they are faster than intels. Also, i heard you mention that 12 cores at 2.13 GHz has "more GHz" than 6 at 3.33. 2.13x2=4.26 GHz, right? NO. Each core runs at 2.13 GHz, or whatever, not add them all up. Think of it like this. The CPU is a road. First, you have 2 single lane roads, single cores. The speed limit, or clock speed, is say 60 MPH. Now, in this comparison, we have a 12 lane road and a 6 lane road to start. The 12 lane has a speed limit of 40 MPH, and the 6 lane road of 60 MPH. When its 2 AM at night, and people only need 1 lane, the 60 MPH one will go faster. This would represent and single threaded application. Now, the 6 lane road also has much wider lanes. This represents both HT and much faster individual cores. The 12 lane road is moving along at 40 MPH, not bad, but people only need maybe 4, 6, or 8 lanes, and therefore 8, 6, or 4 arent used. This represents cores not being used by software. So now we have a traffic jam on the 40 MPH road only using 6 lanes, but traffic is breezing along at 60 MPH on the nice and wide 6 lanes. This could be made more complex, but thats the best i can explain it.
Score
0
a b à CPUs
a c 114 À AMD
a b å Intel
September 28, 2010 2:43:51 AM

This question is basically pointless. It's like asking what is faster, Lambourghini's fastest car or Toyota's. Sure the Lambourghini Murcielago is the king between the two companies but there are few situations where the Murcielago would be noticably different than the Supra Turbo. Of course one of those situations would be the sticker shock of $500,000 vs. $70,000. The 980X is just a geeky penis extension. My Phenom II X4 940 is not top of the line by today's standards but I am still giddy about the speed it has even 18 months after buying it. Anyone who wants to pay $1000 for a CPU that will feel the same as the Phenom II X6 is either really stupid or is trying to compensate for something. I don't care if the person is richer than Midas, it's still a stupid waste of money.

As for the multi-cores, a slow-clocked 12 core like the Opteron is like a powerful diesel engine while the 8-core Xeon is like a powerful gas engine. The Xeon is faster off the line and has a higher top speed but it's far easier to overload. The Opteron doesn't go as fast, but it doesn't slow down at all no matter how much you throw at it.
Score
0
a b à CPUs
September 28, 2010 2:54:48 AM

Avro Arrow said:
This question is basically pointless. It's like asking what is faster, Lambourghini's fastest car or Toyota's. Sure the Lambourghini Murcielago is the king between the two companies but there are few situations where the Murcielago would be noticably different than the Supra Turbo. Of course one of those situations would be the sticker shock of $500,000 vs. $70,000. The 980X is just a geeky penis extension. My Phenom II X4 940 is not top of the line by today's standards but I am still giddy about the speed it has even 18 months after buying it. Anyone who wants to pay $1000 for a CPU that will feel the same as the Phenom II X6 is either really stupid or is trying to compensate for something. I don't care if the person is richer than Midas, it's still a stupid waste of money.

As for the multi-cores, a slow-clocked 12 core like the Opteron is like a powerful diesel engine while the 8-core Xeon is like a powerful gas engine. The Xeon is faster off the line and has a higher top speed but it's far easier to overload. The Opteron doesn't go as fast, but it doesn't slow down at all no matter how much you throw at it.

Why does it say at the bottem of your post im not a amd fanboi im just a intel hating plz give me reasons why amd is better.
Score
0
a b à CPUs
September 28, 2010 2:55:46 AM

ares1214 said:
Its hard to say a processor "beats" another. Heres your new best friend:

http://www.anandtech.com/bench/CPU/2

See what beats what, but more importantly, in what benchmarks certain things beat others. To answer a few of your question, possibly repeating the info of others, but here it goes. Core count isnt everything. Most games can only currently use 2 or 3 cores. This makes the 4th core on a quad more or less useless besides running background apps. Also, 1 AMD core doesnt equal 1 intel core. The intel cores are quite a bit faster clock for clock. The architecture just makes them this way, without going far in depth. GHz is a standard measurement, so 1 GHz from AMD is 1 GHz from anybody. It just so happens that those AMD 12 cores are clocked very low, while the intel is clocked very high. Back to the gaming example, most software only uses a certain amount of cores. AMD can have a 24 core CPU if they wanted, but using the same architecture, it would very likely perform little to no better, as most software isnt very core efficient. The i3 is a dual core with Hyper Threading. This means for every 1 core, you get 1 thread, for a total of 2 cores, 4 threads. The way intel does it, 1 thread does not equal 1 core. Fact is they share resources, and therefore HT can sometimes give a large increase in performance for apps that are threaded well, but it more generally doesnt add much to real life apps, and sometimes even takes away performance. So one would think, hey a 3 core CPU at 3 GHz should beat a dual core at 2.93? Not really. This goes back to the architecture. Intel's individual cores are more powerful, and therefore they need less cores, and at lower clock speed. Use that anandtech bench thing and pit a 970 vs 1090T. Identical core count, identical clock speed, but notice how the 970 beats it soundly? All has to do with the arch, and likely some help from HT. AMD adds more cores and clocks them higher more out of necessity than because they are faster than intels. Also, i heard you mention that 12 cores at 2.13 GHz has "more GHz" than 6 at 3.33. 2.13x2=4.26 GHz, right? NO. Each core runs at 2.13 GHz, or whatever, not add them all up. Think of it like this. The CPU is a road. First, you have 2 single lane roads, single cores. The speed limit, or clock speed, is say 60 MPH. Now, in this comparison, we have a 12 lane road and a 6 lane road to start. The 12 lane has a speed limit of 40 MPH, and the 6 lane road of 60 MPH. When its 2 AM at night, and people only need 1 lane, the 60 MPH one will go faster. This would represent and single threaded application. Now, the 6 lane road also has much wider lanes. This represents both HT and much faster individual cores. The 12 lane road is moving along at 40 MPH, not bad, but people only need maybe 4, 6, or 8 lanes, and therefore 8, 6, or 4 arent used. This represents cores not being used by software. So now we have a traffic jam on the 40 MPH road only using 6 lanes, but traffic is breezing along at 60 MPH on the nice and wide 6 lanes. This could be made more complex, but thats the best i can explain it.
i looked and the core i3 which is dual beat the tri core athlon in games, how?
Score
0
a b à CPUs
September 28, 2010 2:57:32 AM

Someone said that a intel 1.15 ghz is as fast as amd 2ghz how is that possible?!
Score
0
a b à CPUs
September 28, 2010 3:01:21 AM

Its because the AMD architecture does less work per Ghz than Intel. Like How a 3Ghz P4 didn't match a Athlon 64 @2Ghz, its the other way around now.

But its more like 1.5Ghz not 1.15Ghz.
Score
0
a b à CPUs
September 28, 2010 3:23:08 AM

HostileDonut said:
Someone said that a intel 1.15 ghz is as fast as amd 2ghz how is that possible?!


Im not sure what cpus from intel and amd there comparing each other with but if there modern cpu, those numbers seam a bit big.... could be just me though. :/ 

Anyways, how is a higher ghz cpu the same performance as a lower ghz cpu, well, it has to do with a few things but mainly with IPC (intruction per clock). Basically IPC meaning the amount of work you can get done per clock.

To make this as simple as i can get it: if say intel cpu can send out 2 apple pies per second while AMD cpu can send out 1 apple pie per second, even though there at the same ghz, which one works faster? Intel as it can get more pies out than AMD. Meaning Intel IPC is higher than AMD

Now there more to it than that but that the best example that i could explain IPC. Maybe someone else can give a better detail than I but IPC is what makes the difference in a cpu when compairing a core to core of difference cpu's.



Now a example of this IPC difference is if you go look at the past of cpu history, you'll find stuff around 6years ago about AMD Athlon 64 vs Intel Pentium 4, You find that the Athlon 64 was faster on it's IPC than a Pentium 4 by a big margin.

When you see a Athlon 64, it will have a 4 digit number as the model number and that represented the Pentium 4 equivalent speed in MHz.

So an AMD Athlon 64 3200+ (which was at 2.2GHz) would equal 3.2 GHz Pentium 4.

Hope this cleared up some questions.
Score
0
September 28, 2010 3:27:04 AM

Technically, Amd 1090T wins slightly from i5 870 AND i7 930. BUT i7 950 wins slightly from AMD 1090T.

Now that's the border you want to know.
Score
0
a b à CPUs
September 28, 2010 4:09:07 AM

So how does my AMD tri-core @ 3ghz loose to a core i3 dual core at 3ghz or like 3.2? Even tho the dual has a little more power and whatever the ICP or PCI is higher was wouldnt the extra core make up for it? Will AMD ever be better than Intel ever again?
Score
0
a b à CPUs
September 28, 2010 8:14:53 AM

AMD's next real chance to be genuinely better than Intel will be when Bulldozer comes out (which should happen sometime around next summer or fall). That should be interesting. Before that, Intel should hold the lead quite solidly, perhaps even extending it a bit with Sandy Bridge this coming Jan.
Score
0
September 28, 2010 10:20:14 AM

".... ;DELETE * FROM THREAD WHERE TITLE LIKE "Amd vs Intel""

hope this works.
Score
0
a b à CPUs
September 28, 2010 11:01:35 AM

HostileDonut said:
So how does my AMD tri-core @ 3ghz loose to a core i3 dual core at 3ghz or like 3.2? Even tho the dual has a little more power and whatever the ICP or PCI is higher was wouldnt the extra core make up for it? Will AMD ever be better than Intel ever again?


Like i said, most games dont use more than 2 cores. Also, gaming is much less dependent on CPU, and much more dependent on GPU. This just goes back to currently, intel has faster cores right now. They can clock them lower, have less of them, but they still come out on top. AMD will get back in the game when they stop putting more cores and clocking them higher on a clearly beaten architecture. This is where Bulldozer comes in, but we shall see...

@Dipankar, actually, this thread has been rather flame retardent IMO...
Score
0
September 28, 2010 12:54:44 PM

Avro Arrow said:
The 980X is just a geeky penis extension.


Win.
Score
0
September 28, 2010 1:22:21 PM

If you put the best Intel against the best AMD, the Intel will win. Intel's are faster than AMD but you pay for it. If you are building a system, how much speed do you really need? Personally, I'm going AMD on my next build. You can't beat AMD's price to performance ratio. The money saved from buying an AMD processor can go into getting a better GPU.
Score
0
a b à CPUs
September 28, 2010 1:44:32 PM

cjl said:
AMD's next real chance to be genuinely better than Intel will be when Bulldozer comes out (which should happen sometime around next summer or fall). That should be interesting. Before that, Intel should hold the lead quite solidly, perhaps even extending it a bit with Sandy Bridge this coming Jan.


That's true - JF already discounted the Theo Valich story about BD being delayed until the 4th quarter of next year, which is good for me since I don't want to wait that long.

The only thing that could prove to be a problem is this :

Quote:
SAN JOSE, Calif. - It's official. After much speculation, Advanced Micro Devices Inc. (AMD) has lowered its forecast.

AMD announced that it expects revenue for the quarter ending Sept. 25, to be in the range of down 1-to-4 percent, as compared to revenue of $1.65 billion for the quarter ended June 26. Originally, AMD expected revenue to be up seasonally for the third quarter of 2010.

The sequential decrease is due to weaker than expected demand, particularly in the consumer notebook market in Western Europe and North America, according to AMD.

In its most recent quarter, AMD recently announced revenue for the second quarter of 2010 of $1.65 billion, a net loss of $43 million, or $0.06 per share, and operating income of $125 million.

Rival Intel Corp. also recently lowered its forecast for similar reasons. On Wednesday, one analyst lowered his forecast for AMD.

Due to weaker than expected PC demand through the back-to-school selling season and an imminent sub-seasonal holiday build period, Gleacher & Co. analyst Doug Freedman is lowering his estimates for AMD.


If AMD suffers yet another >$100M loss quarter that could mean more layoffs and delays.

OTOH, there's a front-page news story here quoting JF that sales of Magny Cours are good, and that is the high-profit end of the product lineup.
Score
0
a c 172 à CPUs
a b å Intel
September 28, 2010 3:08:41 PM

HostileDonut said:
Why does it say at the bottem of your post im not a amd fanboi im just a intel hating plz give me reasons why amd is better.

He is an AMD fanboi using hatred for Intel as justification. You see it in his signature under his avatar.

It's not a case of Intel being technically better. It's a case of Intel being found guilty of some predatory business practices.

There are two things we can give AMD credit for: first, just for being there. Competition has a downward pressure on prices.

Second, we owe the X86-64 64 bit instructions to AMD. AMD was first to market with an X86 compatible 64 bit instruction set. What Intel had in mind was something different. ANd while I do not know for certain, I heard that Microsoft told Intel that they were not going to support two different 64 bit instruction sets.


Score
0
a c 131 à CPUs
a b À AMD
September 28, 2010 4:10:30 PM

Quote:
@ares Mate no matter how much you try you just cannot make an Amd fanboy understand Intel i series is better.

Every time I have seen a post of yours, it's been use saying something about "amd fanboys" or how intel is better at everything.
Are you a troll or do you just not have time to actually say something useful?
Score
0
a b à CPUs
September 28, 2010 4:24:33 PM

I agree with jsc.Without AMD it'll be just lame.

I just want to give a hint of Bulldozer performance.AMD said that the 16 core Interlagos Bulldozer will be 50% faster than a 12 core Magny Cours.If you consider that the Magny Cours scores 100 based on the tests you perform, then the Interlagos will score 150.So, core for core Bulldozer is 13% faster than the current AMDs.So considering the Phenom II x6 1090T scores 100, a 3.2 GHz eight core bulldozer(Orochi) will score 150.Now I calculated that the i7 970 is 29.5 % faster than the X6 1090t( from the AnandTech bench, neglecting the benches in which there is negligible difference between the two.)If I round it off to 30%, the i7 scores 130.When you calculate score/ core it turns out to be 16.7 for the 1090 T, 18.75 for Bulldozer and 21.7 for the Core i7 970, considering same clock speeds.

So at best Bulldozer will match the current Core i7s, while Sandy Bridge will add at least 10% over the Nehalems.
Score
0
a b à CPUs
September 28, 2010 4:28:46 PM

Quote:
Looks like only threads which gets importance here in Tom are threads with Intel Vs amd thread.

Why should it not be given importance?It is importance to people who are new to building PCs.
Score
0
a b à CPUs
September 28, 2010 4:29:58 PM

So next Jan. when bouldozer comes or whatever AMD might be better? I think with price AMD is better, plus AMD has 6 core CPU's which beat i5 quads foe less than 200 and for real super computers 12 core CPUs.
Score
0
a b à CPUs
September 28, 2010 4:31:29 PM

HostileDonut said:
So next Jan. when bouldozer comes or whatever AMD might be better? I think with price AMD is better, plus AMD has 6 core CPU's which beat i5 quads foe less than 200 and for real super computers 12 core CPUs.

They do not beat the i5s by that much(the 1090T perhaps, but not definitely the 1055T).
Score
0
a c 131 à CPUs
a b À AMD
September 28, 2010 4:35:36 PM

Tamz_msc said:
I agree with jsc.Without AMD it'll be just lame.

I just want to give a hint of Bulldozer performance.AMD said that the 16 core Interlagos Bulldozer will be 50% faster than a 12 core Magny Cours.If you consider that the Magny Cours scores 100 based on the tests you perform, then the Interlagos will score 150.So, core for core Bulldozer is 13% faster than the current AMDs.So considering the Phenom II x6 1090T scores 100, a 3.2 GHz eight core bulldozer(Orochi) will score 150.Now I calculated that the i7 970 is 29.5 % faster than the X6 1090t( from the AnandTech bench, neglecting the benches in which there is negligible difference between the two.)If I round it off to 30%, the i7 scores 130.When you calculate score/ core it turns out to be 16.7 for the 1090 T, 18.75 for Bulldozer and 21.7 for the Core i7 970, considering same clock speeds.

So at best Bulldozer will match the current Core i7s, while Sandy Bridge will add at least 10% over the Nehalems.

You forgot a couple things.

This "per core" value is when threads are being used in both cores in a single module. Each thread there would perform 90% of the performance of one thread on one core of the module with the other core idle (no shared resources), I believe. I remember john frueh saying something about two threads: on the same module, 180 units throughput, if on different modules with one core idle on each module, 200 units throughput.

You again also forgot that the Bulldozer for desktop will have a turbo core, unlike the server variation.

You also neglected to calculate properly. You can't compare the desktop and server counterparts directly, the server part will sacrifice per core performance for more cores.

I intend to go through all of the blog posts and comments from John's blog, he actually reveals quite a bit. When I get the time, I will attempt to calculate my expectations of improvement in the desktop variants.
Score
0
a b à CPUs
September 28, 2010 4:41:04 PM

enzo matrix said:
You forgot a couple things.

This "per core" value is when threads are being used in both cores in a single module. Each thread there would perform 90% of the performance of one thread on one core of the module with the other core idle (no shared resources), I believe. I remember john frueh saying something about two threads: on the same module, 180 units throughput, if on different modules with one core idle on each module, 200 units throughput.

You again also forgot that the Bulldozer for desktop will have a turbo core, unlike the server variation.

You also neglected to calculate properly. You can't compare the desktop and server counterparts directly, the server part will sacrifice per core performance for more cores.

I intend to go through all of the blog posts and comments from John's blog, he actually reveals quite a bit. When I get the time, I will attempt to calculate my expectations of improvement in the desktop variants.

Yes, these are just the tentative results which can be expected, unless they improve upon it.I'm not saying that these are the actual results when they're going to be released.But i still believe that AMD can match the Nehalems with Bulldozer at best.
Score
0
a c 131 à CPUs
a b À AMD
September 28, 2010 4:49:09 PM

Tamz_msc said:
Yes, these are just the tentative results which can be expected, unless they improve upon it.I'm not saying that these are the actual results when they're going to be released.But i still believe that AMD can match the Nehalems with Bulldozer at best.

I sure hope so, for the sake of technology moving ahead. Intel didn't release anything other than the i7 920 and 940 in november 2008. It took nearly a year before any other i7s and the i5 were released.
Score
0
September 28, 2010 5:41:16 PM

The fanboys can have their fun. I'm a brand 'ho. I go wherever the function offers me the best price. I've switched back and forth between Nvidia and ATI (AMD) more times than I can remember. I do the same with CPUs. I currently own one Intel and one AMD chip in my desktop PCs. I will admit that Intel has peeved me with all these socket changes of late. I hate swapping MBs! In that respect, I like that AMD has stuck with a socket (or backwards compatible cpu) for awhile now. That could all change with Bull..err dozer though. I suppose I will always have a small epeen. But I'll have more toys in the garage...
Score
0
a b à CPUs
September 28, 2010 6:07:22 PM

So AMD an Intel are about head on head?
Score
0
a b à CPUs
September 28, 2010 6:22:34 PM

enzo matrix said:
I intend to go through all of the blog posts and comments from John's blog, he actually reveals quite a bit. When I get the time, I will attempt to calculate my expectations of improvement in the desktop variants.


Heh, good luck with that - JF has been notoriously close-mouthed about performance figures other that what AMD has officially annouced, which is the "BD 50% faster with 33% more cores compared to MC"..

BD will have a 4-issue decoder compared to the K8 (& K10 & K10.5) 3-issue, and various other enhancements to increase IPC, but where I really expect it to shine in comparison to Sandy Bridge will be in apps with a large number of heavy-usage threads (i.e., not many slack clock cycles). The SMT that Nehalem & later generations use only shows significant improvement when there are lots of unused clock cycles available in both threads so that the core can switch from one to the other and make use of those slack cycles.
Score
0
a b à CPUs
September 28, 2010 6:39:11 PM

Tamz_msc said:
Yes, these are just the tentative results which can be expected, unless they improve upon it.I'm not saying that these are the actual results when they're going to be released.But i still believe that AMD can match the Nehalems with Bulldozer at best.


I did the same math, but JF said it really doesnt work that way. We also have to consider that 16 "core" really means 8 modules. 8 modules might be cheaper to make than 12 cores. So we just need to wait. Im not so sure about performance, but IMO, price to performance will be undeniable with BD, and it should hopefully have the performance to back it up a little. But they might not take the crown from i7+10-20%. It should be close. Smaller features like lack of overclocking, new sockets, heat, temps, and so forth might make the difference.
Score
0
a b à CPUs
September 28, 2010 7:09:39 PM

David Kanter of Realworldtech has an interesting comparison of Sandy Bridge vs. Bulldozer:

Quote:
The Sandy Bridge CPU cores can truly be described as a brand new microarchitecture that is a synthesis of the P6 and some elements of the P4. Although Sandy Bridge most strongly resembles the P6 line, it is an utterly different microarchitecture. Nearly every aspect of the core has been substantially improved over the previous generation Nehalem. Many of these changes, such as the uop cache or physical register files, are drawn from aspects of or concepts behind the P4 microarchitecture. While the P4 was ultimately a flawed implementation, it embodied many good ideas – ideas that are reappearing across the industry, and in Sandy Bridge. The underlying philosophy of Intel’s approach to CPU design is to focus on maximizing per-core performance and efficiency. This is a contrast to AMD’s Bulldozer, which backs off slightly from per-core performance to emphasize aggregate throughput. This article will explore the microarchitecture of Sandy Bridge – a 64-bit, quad-core, dual threaded, 4 issue, out-of-order microprocessor with the new 3 and 4 operand AVX instruction set extension, implemented in Intel’s 32nm process.

...

Sandy Bridge is a fundamentally new microarchitecture for Intel. While it outwardly resembles Nehalem and the P6, it is internally far different. The essence of an out-of-order microarchitecture is tracking, re-ordering, renaming and dynamically scheduling operations to achieve the limit of data flow. Nehalem and Westmere rely on the same mechanisms that date back to the original P6. Sandy Bridge changes the underlying out-of-order engine and uses the more efficient approach taken by the EV6 and P4. That one change alone qualifies Sandy Bridge as a different breed entirely from the P6. But, there are changes in almost every other aspect of the design. The uop cache is a huge improvement for the front-end, largely by eliminating many of the vagaries of x86 fetch and decode. The implementation is quite clever and achieves many of the aims of the P4’s trace cache, but in a far more efficient and reliable manner. AVX improves execution throughput and most importantly, the more flexible memory pipelines benefit almost all workloads.

In the coming year, three new microarchitectures will grace the x86 world. This abundance of new designs is exciting; especially since each one embodies a different philosophy. At the high-end, Sandy Bridge focuses on efficient per-core performance, while Bulldozer explicitly trades away some per-core performance for higher aggregate throughput. AMD’s Bobcat takes an entirely different road, emphasizing low-power, but retaining performance. In contrast, Intel’s Atom is truly intended to reach the most power sensitive applications. The two high-end microarchitectures, Sandy Bridge and Bulldozer, are shown below in Figure 7. Note that each Bulldozer module would include two integer cores while sharing the front-end and floating point cluster. Also, the floating point cluster in Bulldozer does not directly access memory, instead it uses the memory pipelines in the two attached cores, which then forward results to the FP cluster.


and

Quote:
A Sandy Bridge core should have substantially higher performance than a Bulldozer module across the board for single threaded or lightly threaded code. It will also have an additional advantage for floating point workloads that use AVX, (e.g. numerical analysis for finance, engineering). With AVX, each Sandy Bridge core can have up to 2X the FLOP/cycle of a Bulldozer module, although they would be at parity if the code is compiled to use AMD’s FMA4 (e.g. via OpenCL). FMA4 will be relatively rare because, while elegant, it is likely to be a historical footnote for x86, supplanted by Intel’s FMA3. For software still relying on SSE, the difference between the two should be minimal. In comparison, Bulldozer will favor heavily multi-threaded software. Each module has twice the memory pipelines and slightly more resources (e.g. retirement queue/ROB entries, memory buffers) than a single Sandy Bridge core with two threads, so Bulldozer should do very well in many highly parallel integer workloads that exercise the memory pipelines.

In many ways, the strengths of Sandy Bridge reflect the intentions of the architects. Sandy Bridge is first and foremost a client microprocessor – which requires single threaded performance. Bulldozer is firmly aimed at the server market, where sacrificing single threaded performance for aggregate throughput is an acceptable decision in some cases. Perhaps in future articles, we can examine the components of performance in greater detail (e.g. frequency, IPC, etc.), but for now, high level guidance seems appropriate – given the level of disclosure from both vendors.



Score
0
a b à CPUs
a c 114 À AMD
a b å Intel
September 28, 2010 8:04:55 PM

HostileDonut said:
Why does it say at the bottem of your post im not a amd fanboi im just a intel hating plz give me reasons why amd is better.

Honestly, AMD isn't better per se, it's just a more ethical company than Intel. Intel's anti-competitive practices have cost us all a ton of money and have held back CPU development by at least 2 generations. To me, that's not a company whose products I want to buy even though I've owned 15 Intel CPUs all the way back to the original 8088 in the IBM PC. In addition to that, nVidia has proven itself to be a rather dishonest company as well (rebranding of the 8800 series as the 9600GT, 9800GT, 9800GTX+ and now the GTS 250) so since DAMMIT (ATi/AMD) hasn't been pulling dishonest marketing crap or trying to damage competition in the marketplace, I'll support them as long as their products do what I need them to. I don't love them, I just hate Intel and nVidia for the crap that they've pulled. It's about ethics with me, not technology. The fact that ATi and AMD consistently give much better value for the money and are therefore more consumer-friendly than Intel and nVidia are another reason I took the side I did.
Score
0
!