Sign in with
Sign up | Sign in
Your question
Closed

Tom's CPU Architecture Shootout: 16 CPUs, One Core Each, And 3 GHz

Tags:
Last response: in Reviews comments
Share
July 26, 2011 5:43:52 AM

This was a pretty nifty article. Hope you guys revisit it after Bulldozer and Ivy Bridge drop. Thanks!
Score
31
July 26, 2011 5:45:07 AM

Wow, this has got to be one of your biggest comparos EVER. Didn't finish reading it yet, but it looks like this one will be quite a doozy
Score
2
Related resources
a b à CPUs
July 26, 2011 5:47:59 AM

Well AMD hasn't released a new architecture in a long time.. what you expect?
Score
4
July 26, 2011 5:51:38 AM

Thing is, Intel will already have their updated Sandy Bridge processors by the time Bulldozer comes out - which will probably maintain the gap. AMD would need a MASSIVE effort to catch up or even pass Intel at this point. Bulldozer's going to have to try really hard to win back the high-end enthusiast market. Intel clearly has had much better direction from Intel Core up to now.
Score
21
July 26, 2011 5:55:49 AM

Every single comment posted before Darkerson (there were quite a few) seem to have mysteriously vanished... strange.
Score
11
July 26, 2011 6:12:32 AM

dragonsqrrlEvery single comment posted before Darkerson (there were quite a few) seem to have mysteriously vanished... strange.


How many more were there? That's not something I've seen happen before.

Best,
Chris
Score
8
July 26, 2011 6:14:59 AM

One of the most interesting articles I've seen from toms. I already had an idea of the standings between intel and AMD per core per clock by how game minimum specs usually say something like "Minimum Intel Core 2 Duo 1.8Ghz or AMD Athlon x2 2.4Ghz", but it was nice seeing it put into perspective.

Hope you revist it with bulldozer/ivy bridge.
Score
5
a b à CPUs
July 26, 2011 6:18:31 AM

Great review, long live my i7 2600k :) 
Score
2
a b à CPUs
July 26, 2011 6:39:45 AM

Looks like Bulldozer is going to combat SB by moderate increasing of IPC AND increase in frequencies. Purely by IPC it will lose to SB but if it can deliver high frequencies at the same time, we may have a decent competition.
Score
4
July 26, 2011 6:44:44 AM

cangeliniHow many more were there? That's not something I've seen happen before.Best,Chris

I don't know for certain, maybe around 6 or 8. I'm only aware of this because I was one of the people who posted a comment.
Score
1
Anonymous
a b à CPUs
July 26, 2011 6:48:04 AM

Now this was a great article! Good work guys!
Score
4
July 26, 2011 6:51:38 AM

... you couldn't wait until the bulldozer comes out?
Score
-18
a b à CPUs
July 26, 2011 7:00:53 AM

DjEaZy... you couldn't wait until the bulldozer comes out?

I think it may be a bit difficult to test bulldozer this way because configuration of cores within modules is rather funky. That is they are not full fledged cores within a module.
Score
9
a b à CPUs
July 26, 2011 7:06:53 AM

Also, clock for clock testing is not always fair. For example, in P4 vs. Athlon 64 days clock for clock testing would not be fair because while P4 was much worse in IPC, it was designed for higher frequencies. Sure, Athlon still bit it in most cases, but gap was much narrower that clock for clock tests would suggest. We may have a similar situation (in reverse) in SB vs. Bulldozer.
Score
-10
a b à CPUs
July 26, 2011 7:25:07 AM

Pretty cool article.

Thanks.

That must have been one tedious sob to pull off. I wouldn't have had the discipline to even finish the P4 tests lol...

I'm gonna go stroke my kentsfield... I think it knows when the new system's up, it's going to be subjected to 1.5+ volts ala Frankenstein's Monster... bwuahahahaha
Score
14
a b à CPUs
July 26, 2011 7:44:33 AM

yyk71200Also, clock for clock testing is not always fair. For example, in P4 vs. Athlon 64 days clock for clock testing would not be fair because while P4 was much worse in IPC, it was designed for higher frequencies. Sure, Athlon still bit it in most cases, but gap was much narrower that clock for clock tests would suggest. We may have a similar situation (in reverse) in SB vs. Bulldozer.

You talking stock or overclocked? And if overclocked, then is it a 24/7 clock with what cooler? My point is that we don't really have any idea how Bulldozer will compare in clock speeds – it could be very well at stock but very poorly when overclocked (24/7 clock, but once again I'm not specifying with what cooling (could be low-end air to high end water or even LN2!) – but same on each), for all we know.
Score
4
July 26, 2011 8:05:52 AM

Where are those bots i buy all my cloths from?
Score
22
a b à CPUs
July 26, 2011 8:09:39 AM

Excellent work reviewers!
It is great to see a solid, in depth, tech article again =D

@yyk7120, the entire point of this article was to show the relative IPC performance between the different architectures.
In a battle of pure IPC performance, nominalizing all variables is not being unfair, it is entirely necessary.
Think of it more as an in depth look into the underlying architectures rather than another AMD vs Intel @ price point that most are so used to.
Score
9
a b à CPUs
July 26, 2011 8:16:02 AM

This is about as close as you can get to real raw core efficiency. I love this article and was wondering the same thing as well.
Score
14
Anonymous
a b à CPUs
July 26, 2011 8:25:24 AM

I would LOVE to see one additional page - 'performance per transistor'

Basically - take the performance data, and divide the results by the amount of transistors per core (eg 1 billion transistors for 8 core -> 125 mil transistors for core). There would be a bias towards CPUs with huge L3 cache (as they can use all of it), but it would still be very interesting thing to see

I would really want to see how the efficiency of the processors increased compared to the amount of transistors they use
Score
20
July 26, 2011 8:42:23 AM

pschmid said:
The CPU landscape is really complex. Both AMD and Intel offer tons of different models. But how would today’s processors perform if they didn't have multiple cores? We take 16 different CPUs and compare them all using a single core running at 3 GHz.

Tom's CPU Architecture Shootout: 16 CPUs, One Core Each, And 3 GHz : Read more

wow I really didnt expect that clock for clock amd was still competing with conroe
Score
2
July 26, 2011 9:02:17 AM

looks like toms forgot to include first gen phenom in this article....
pick the dual core kuma instead of agena, since its easier to oc an athlon 7850 to 3 ghz than an agena
Score
2
July 26, 2011 9:22:43 AM

yyk71200I think it may be a bit difficult to test bulldozer this way because configuration of cores within modules is rather funky. That is they are not full fledged cores within a module.


I will want to see a module treated like a (big) core in this test. But in a max core test, you may see a 2 module vs. a 4 core Sandy Bridge (about same price & watts). If AMD can make 4 module chip in the same class as a 4 core SB, then that 4 module bulldozer will do very well against a 4 core SB.
Score
1
July 26, 2011 9:45:35 AM

Would be nice to see benchmarks on Bulldozer AMD CPUs, it should show that per core performance gain...
Score
3
July 26, 2011 9:54:25 AM

interesting... i likey this one
Score
1
July 26, 2011 9:59:37 AM

Wow, almost makes me ashamed of my 1055T :(  AMD has got some serious architecture work to do!!
Score
6
July 26, 2011 10:38:18 AM

I still see no reason to upgrade from my Lynnfield.
Score
0
July 26, 2011 10:40:16 AM

Can I assume HT was left on as no mention of HT being disabled was mentioned in the article.
Score
-5
a b à CPUs
July 26, 2011 10:50:24 AM

tranzz said:
Can I assume HT was left on as no mention of HT being disabled was mentioned in the article.

If you look at the CPU-Z screenshots, you would notice that all CPUs show 1 Core 1 Thread.
This indicates that SMT has been deactivated for the tests.
If SMT where active, CPUs that support it would show 1 Core 2 Threads.
Score
4
July 26, 2011 11:30:55 AM

yyk71200I think it may be a bit difficult to test bulldozer this way because configuration of cores within modules is rather funky. That is they are not full fledged cores within a module.

Well testing a Bulldozer core this way would provide the best case numbers : When one module is combining the ressources of two cores to run one thread (this is the kind of performance you can expect when a Bulldozer is running same or less number of threads then it has modules).

Then you would need to come up with a way to test the worst case numbers : When one module is splitting the ressources of two cores between two equally demanding threads (which is unlikely to happen unless you are putting some specific effort into making it artificially happen).

But choosing an average performance per core between those two set of numbers (best and worse) can only be the result of an arbitrary decision.

If anything, it will be fun to watch how far some biased people(Intel, AMD, benchmark designers, reviewers, fans...) will go out of their way, to twist benchmarks and benchmark results one side or the other in the most ridiculous ways, to achieve the kind of arbitrary decision that suits their own purpose :) 
Score
5
July 26, 2011 12:04:33 PM

i hope Bulldozer grabs the crown again. Those were fun days where AMD had the efficency and intel could only compete by upping the clocks to meltdown levels.
Score
6
July 26, 2011 12:30:55 PM

Great article

Hope you compare Bulldozer and the Ivy Bridge to this
Score
3
July 26, 2011 1:10:43 PM

Great Article!!! Assuming scaling for clock and core count I can now tell that my upgrade from a x2 6000+ to Phenom2 x3 740 (both OC'd) ended up roughly doubling my performance in something that takes full advantage of multi-thread or about 35% if it uses one core.
Score
1
July 26, 2011 1:34:52 PM

results are to be as expected..
Score
-2
July 26, 2011 1:37:10 PM

no i3?
Score
-6
July 26, 2011 1:42:01 PM

I know they included 3d mark, but I wished you would include a few games as well.
Score
5
July 26, 2011 1:42:33 PM

Pentium 4 wasn't initially released at 1.3 GHz, so the Pentium 4 did not rise from 1.3 GHz. When the lowly 1.3 GHz Pentium 4 was released, it was a lower cost alternative to the already present 1.4 GHz and 1.5 GHz processors.
Score
4
a b à CPUs
July 26, 2011 1:47:14 PM

The omission of the llano chips that you have reviewed is frustrating. I have heard they are roughly the same architecture as the Phenoms, but hasn't it recieved a process shrink?

It is the most recent offering from AMD and yet you omit it while including intels most recent chips. Seems a bit crappy. Devalues the article.
Score
-4
July 26, 2011 1:59:47 PM

Great article!

I too would really like to see Bulldozer take the crown again, but somehow I doubt it. What's the deal with those trigate transistors? They seem like a pretty important breakthrough - will other chipmakers be able to use them or is it an Intel-only thing? If the latter, methinks AMD is doomed...
Score
4
a c 172 à CPUs
July 26, 2011 2:30:14 PM

Looks like Intel really got it right when they moved from P4 to Core2. I will be able to skip another generation and wait for Ivy Bridge.
Score
5
July 26, 2011 2:31:10 PM

Wow. This was a fun read from a theoretical perspective. Really highlights how far behind AMD has been in terms of raw core efficiency for years now. It makes you appreciate how amazing AMD's business model is given the massive differences in computing power per clock. They have really done an amazing job filling in nitch segments and adding cores at given price points to compensate; staying in business considering the theoretical "handicap" of their chips has been no accident...well done AMD.

I miss the days when these companies traded blows with every architectural update tho...
Score
10
July 26, 2011 2:56:00 PM

the P4 is like IE8. outclassed and way off the charts.
Score
2
July 26, 2011 3:23:22 PM

This is the type of article that keeps TH in my favorites, just wanted to say thanks :) 
Score
8
July 26, 2011 3:27:40 PM

This is the ultimate "you get what you paid for" article.
Score
-4
July 26, 2011 4:02:58 PM

To include the next gen of CPU's like the BD and the IB this test should be limited to 2 cores each and 3Ghz each, and the Llano and Atoms should be included for the sake of completion.
Score
3
a b à CPUs
July 26, 2011 4:11:16 PM

outlw6669Excellent work reviewers!It is great to see a solid, in depth, tech article again =D@yyk7120, the entire point of this article was to show the relative IPC performance between the different architectures.In a battle of pure IPC performance, nominalizing all variables is not being unfair, it is entirely necessary.Think of it more as an in depth look into the underlying architectures rather than another AMD vs Intel @ price point that most are so used to.

Oh, I am not talking that IPC is not important. It is very important. I am just telling that a company may choose combat a CPU of another company by designing a CPU with higher frequencies in mind but lower IPC. For example, if SB could deliver only 1GHz in frequency, it would have been a flop regardless of superior IPC. Phenom II are designed to run at similar frequencies as modern Intel processors (default). Therefore, it is perfectly fair to test them clock for clock. In fact, current CPUs frequencies are similar to P4 frequencies, so it is fair as well. I am talking about a very situation, when chips are meant to run at very different frequencies like Athlin 64 vs. P4 because frequencies is a part of performance equation just like IPC. You can say P4 high frequencies were part of the chip's architecture just like IPC together producing the result.
Score
2
July 26, 2011 4:16:51 PM

Where's the douche bag that continually tells me that Phenom II's don't are not comparable to Core 2?? That's weird, cause in the majority of these tests, clock for clock the Core 2 cores still outperform Phenom II cores.

OOOOOWWWWWNNNNNEEEEDDDDD
Score
1
a c 143 à CPUs
July 26, 2011 4:19:26 PM

Can Atom even be overclocked? With the fastest Atom at what, 1.8GHz...trying to run one at 3GHz sounds like a quick way to see some magic blue smoke.

Might have been interesting to see a Celeron or two in the mix; sort of like the Athlon II which suffers for lack of L3.
Score
6
July 26, 2011 4:26:57 PM

A lot of wasted work finding the right motherboards when it is so easy to get Windows to boot on only 1 CPU.
Score
0
Related resources
!