Slashdot Mirror
AMD Dual-Core Performance Revealed
Timmus writes "In two separate articles, FiringSquad takes a look at the performance of AMD's dual-core Opteron CPU. The first article examines the performance of dual-core in scientific computing applications (MATLAB and LS-DYNA) as well as digital photography, while the second story focuses on the performance of dual-core Opteron paired against Intel's dual-core Pentium Extreme Edition in video encoding, Cinebench, and a few other applications. The performance improvements are pretty impressive in multi-threaded applications that take advantage of the technology."
I am running one right now, which is why I got first post!
The world moves for love. It kneels before it in awe.
here and here.
"I'd rather be a lightning rod than a seismometer." -Ken Kesey
So we have:
scientific computing applications (MATLAB and LS-DYNA)
digital photography
video encoding
Cinebench and
"a few other applications".
So what about the average user? Will the college kid who just needs to type their papers, the parents who want to do their taxes, the gamers who want to play high-end stuff, etc. get any sort of boost from this?
I am scientifically inaccurate.
OK. Anyone have a quick simple explanation of why Dual Core over Dual CPU motherboard? are there inherent advantages to dual CPUs so close together?
Why would you need anything faster? I just dont understand... With gentoo compiled with the -superfast and -fatroll flags my system runs fine.
where I can encode mpeg2 DVD (maybe it will be HD-DVD by then), rip & copy a DVD, Download a huge torrent, and Play UT with a respectable framerate.
Wouldn't a better benchmark be to compare a dual core setup to a similarly configured dual processor workstation?
"What kind of music do pirates listen to?" -Paul Maud'dib
"Yeeeaaarrrrr n' Bee!!" -Stilgar, Leader of Sietch Tabr
Why stop at dual core?
Once a way to link multiple cores of a CPU is firmly implemented scaling the chip to 4, 8, or even 32768 cores should be relatively easy.
With chip dies getting smaller and smaller the only real reasons not to continue this multi-core scaling would be physical space and power usage.
Perhaps they could scale multiple cores vertically instead of just making the chip wider and longer.
And perhaps the cores could only be "turned on" when called for instead of using up juice all the time.
Interesting look at the future of chips.
Sony's Playstation 3 is using a "cell processor" or similar multi-core design that has already been covered here in the past.
Arstechnica article on the cell processor here.
--
NoVA Underground: Fairfax County, Loudoun County, Arlington, Price William chat and local forums
Notice the lack of an Athlon 64 FX version of AMDs dual-core strategy. For the time being, its recognized that games are exclusively written for single-threaded operation and as such run better on single-threaded processors at elevated frequencies. Thus, the FX series marches on at 2.6GHz for now. ... so for games, keep to your single core CPU.
Find Nearby Indie Events
article 1
article 2
...and for those who don't want to flip through pages and pages of flash banner ads:
Scientific Computing
MATLAB: Though the script includes a moderate amount of matrix math, it doesn't seem like much of it is parallelized. Our recommendation from two years ago still stands - for most Matlab users, the fastest performance will come with a single Athlon64 line.
LS-DYNA: I will bench the CPUs using two classic tests, a 3-vehicle collision and a single front-collision. The 3-vehicle collision takes more than 24 hours to complete - we do not have these numbers ready for this round of articles.
Digital Imaging
Capture One: With Capture One only supporting two CPU threads, the dual-core Opteron's lower clockspeed is a disadvantage.
Bibble: It took only 4 minutes to complete with the 2x Dual Core Opteron 275. 4 minutes! That's 4.2MB/sec of processing time - a 2x Dual-Core Opteron 275 can process RAW images about as fast as it takes to copy them from to your computer using a standard-grade USB 2.0 CF card reader!
Noise Ninja: On the slower Opteron 246, the fastest results were had with 4 threads, but on the faster CPUs, 8 threads was better.
Video
After Effects: Since the decoding of WMV-HD does not seem to take advantage of both CPUs, the performance gain from the Dual-Core AMD Opterons is virtually absent.
http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2397
...LONG LIVE COMPETITION!
I wish both AMD and Intel well. All the better for us. Lower prices and better performance.
Paleotechnologist and connoisseur of pretty shiny things.
That's what they did in the first article. Sheesh, does NO ONE read the articles anymore?
Yeah, but it's all windows, windows, windows. Sadly, the same is true for the Tech-Report.
Belief is the currency of delusion.
all we need is some dip, and we've got some heavy duty crunching!
On other benchmarks the AMD dual core gets 10-20% better performance! SiSoft Sandra is an exception, where there is a mixed bag between the two processors.
This pretty much verifies for me that Intel did a seriously rushed cludge to get this thing out the door. The only reason I can think of to target this to gamers is that no OEMs would want to buy them for server or desktop use, so you have to target people who like the latest technology even if it isn't that great.
AMD on the other hand seems to have a pretty good product here. I can't wait until the desktop versions come out.
Now, it's struck me as very peculiar that the benchmarks where the dual dual core setup from AMD really shines leave out any comparison whatsoever to the Intel dual-core offering. This begs the question whether the person doing this review is a journalist or a marketing represenatative of AMD.
"We did not have time to evaluate the Intel platform with the Intel MKL, the P4 3.0GHz is an older reference measurement." is a very cheap excuse and indicates either lazyness or bribes on the side of AMD... I hate hardware review sites!
Dr. Dobbs last month had an item regarding threading in real-world environments. The authur said that while multi-threaded applications run a lot faster than single-threaded applications, that always isn't so. In addition, there are some significant issues in running in a multi-tasking, multi-threaded environment, not solved with the use of mutexes and semaphores.
Multi-threading and mult-cores are definitely the way the industry needs to go, but the current development methodologies and application architectures (as well as computing theory) may need to catch up a bit.
Dell still won't sell servers with them ....
> This begs the question...
No it doesn't. If anything, it raises the question.
Check out the article at Anandtech linked a little bit above, which does have comparisons with the Pentium lines. The results are still clearly in AMD's favor.
You can never go home again... but I guess you can shop there.
"The first dual core CPU on the market was IBM's Power4 chip in 2001, but this was reserved for true enterprise grade servers and sites willing to shell out over a half a million dollars for a single server: by Christmas 2005, many gamers and PC enthusiasts will likely own a dual core CPU."
These people never heard of SPARC?
For a much better review, check out ANANDTECH.
www.anandtech.com (not affiliated, just love their coverage)
Anandtech has an AMD dual core Opteron and Athlon64 X2 article that might compliment the original poster's story pretty well. It has a sh*tload of benchmarks:
? i=2397
http://www.anandtech.com/cpuchipsets/showdoc.aspx
I really wish they wouldn't do gaming benchmarks with an Opteron in stories like these. Just because the Opteron used has similar specs to the dekstop processor that hasn't been released doesn't necessarily mean that the gaming benchmarks are all that useful. Just my 2 cents.
It'll be interesting to see how soon prices fall for these AMD processors (server and desktop) when they go mainstream. Read the cost comparisons for these badboys in the article.
Finally, I'm glad that Anand decided to demonstrate that the new AMDs will be backwards compatible with Socket 939 motherboards WITH BIOS revisions. Intel's dual core processors don't offer that luxury, from what I read in the article.
IronChefMorimoto
Okay, let's think about two things:
the limited three dimensions of space, and the limited number of layers you can put on the chip wafer as it is fabricated. The limited number of layers on the wafer is a simple concept to get. The limited 3-d space to work in also limits interconnects between multiple processors in clusters by limiting the topology which the cluster can form.
The uber-cluster concept was in the Thinking Machine (TMI) something-or-other which had 1024 processors linked together in (effectively) a ten-dimensional-hypercube. When you've got clusters currently, they've got to be networked and have an efficient and rapid method for passing messages and data. If you go mebbe to four processors, I can see the interconnects fitting on the wafers tightly. Maybe even to eight...
But beyond that, you start having difficulties maintaining direct interconnects between processors. The Thinking MAchines supercomputer effectively implemented 10 interconnects for each processor by having each processor (defined / labelled as a 10-bit address) connect to each processor whose address differed from its own address at exactly one bit. So a message could pass from any of the 1024 processors to any other of the 1024 processors in at most ten steps. And like the internet fabric, there were redundant pathways for the message to take. (fairly cute bit flipping algoritms for it too).
But I don't see compressing the network connections as easily for a hypercube on a wafer.
But maybe a simple linked line of processors.
hmmm...
Less heat, less space, less energy requirements, eventually less money because there is only one chip.
Two cores inside one chip may make less heat than two separate chips, but I betcha you'll surely need a substantially better heatsink for a dual-core chip than you'd need for a comparable sized single-core chip.
Now, it's struck me as very peculiar that the benchmarks where the dual dual core setup from AMD really shines leave out any comparison whatsoever to the Intel dual-core offering.
They couldn't test a dual core multiprocessor chip from Intel because one doesn't exist yet. They've only released single processor dual core chips so far.
AMD introduced dual core on their multiprocessor server chips first, with desktop chips coming later on. Intel introduced dual core on their single CPU desktop chips first, with server chips coming later on this year, or in early 2006.
The problem is that you can easily run a single multiprocessor-capable CPU in a system, while you cannot run two single-processor-only chips in a system since they lack that capability.
Given ASUS has not updated the socket 940 SK8V BIOS in over 6 months, and hasn't even had a new beta BIOS (yuck) since December, what are the odds of any motherboard really supporting these CPUs? How many companies rush to support the older boards?
In fact, the SK8V is suddenly gone from both the motherboard page and the retired products page on the asus website. Hmmm...
With comments like:
"Even grandmothers own 8-megapixel consumer digital cameras now"
I really have to question the intellegence of this poor guy. I don't know many grandma's that drop $700-$1000 on digital camera's.
Unstable Apps: Our Android Apps Don't Suck
On my Gentoo AMD64 -O3 compiled system, running Gnome, Rhythmbox, Evolution, Epiphany, Liferea and GAIM, it is using 365 MB of RAM, not including buffers or disk cache.
In contrast, Windows XP running a similar set of applications was only using 230 MB.
this enthusiat hardware web sites are simply ridiculous. how can you make a decent test with desktop applications and server processors and systems and reach conclusions? evet their DB tests suck.
>With that comparison, my Emacs session is 6MB.
:)
.
Oh, come on. It took more than that 20 years ago
More seriously, I just launched one, and it immediately was using 9.5Mb on FreeBSD. It then did a couple of things on its own (???), displaying a notes buffer, and hopped up to 9.5Mb.
I shudder to think of what it will do if I actually type anything in its window . .
hawk
I'm an AMD fan boy I was thinking it would smoke the Intel too hard.
The dual core is mostly tested in encoding an area where Intel has traditionally dominated. And it still kicks intel's ass.
Well mine's so fast I can't get a word in edgewise.
Did you just count those processes, or all of the shared DLLs that are loaded, explorer, etc that is necessary to run them?
My beliefs do not require that you agree with them.
How can anyone take an article seriously when the very first sentence just screams, "AMATEUR!!" like this one does:
Intel may very well go down in history as the first processor manufacturer with a dual-core solution, if only by three days.
IBM Power4, Power5
HP PA-8800
Sun Sparc IV
All full-fledged dual-core processors shipping long before Intel -- HP's been shipping for over a year and IBM's already well in to their 2nd generation of dual core processors with Power5.
Sure, you can excuse the author with some hand-waving about x86 context only or whatever. But if they really knew what they were talking about, they would have said it that way - or at least a competent editor would have corrected it. If these guys can't even get the trivial stuff right, how can anyone trust them to get the real technical details right?
When information is power, privacy is freedom.
Actually, no it isn't. The moron kid's friend is the father, so its his DNA.
The world moves for love. It kneels before it in awe.
Seriously. This is the level of usage that I'm trying to achieve, not UT (World of Warcraft). But I do want to be able to queue up a few torrents, rip and encode my movies (for a media centre setup) and play a game at the same time while running Itunes (Decoding lossless).
? i=2397Anandtech
? i=2388 ? i=2389
Back in the day, and I have posted this before, when I had a friend running an Abit BP6 (Dual Celeron board) he could burn a CD, encode MP3s, host a dedicated UT server and then join it. Dual Cores / Dual CPU setups have always been for us hardcore users and this is a remarkable step forward.
Here is the Anandtech AMD Dual-Core Review (I prefer it slightly to the FiringSquad articles)
http://www.anandtech.com/cpuchipsets/showdoc.aspx
Here is Part I of the Intel Dual Core Article
http://www.anandtech.com/cpuchipsets/showdoc.aspx
Here is Part II of the Intel Dual Core Article
http://www.anandtech.com/cpuchipsets/showdoc.aspx
That should provide some useful information.
It strikes me odd that AMD is pricing their Athlon x2's between $500-$1000. This seems particularly high when compared to the prices expected for Intel's offerings. The Pentium D's are supposed to be between $250-$500. I cannot see AMD getting that much higher of a price for similar processors.
even when I game I tend to have background tasks going (downloads, background services, etc)... so in such a case wouldn't it balance the load of the game on one and the background apps on another?
http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2397
As usual.
AMD looks like it has come out with a nice product in the Opteron 875/275/175 and the upcoming Athlon 64 X2s in June but AMD usually comes out with pretty good products at pretty good prices. Go down to the store or website to buy a new system for your business or home, though, and you will probably be choosing between a P4 running at 2.8 Ghz or a P4 at 3.2 Ghz for your new desktop system. If it's a server, you will probably get some sort of dual-Xeon system because that mostly what's for sale. Either way, you will probably buy something branded by Dell or HP.
AMD has turned out a lot of good products over the years but they never sell very well (fill in your own reason why here) and the new dual-core stuff won't be any different. That's reality.
Idiot. look again.
Just because it'll priortize, etc. doesn't mean that it's running the threads simultaneously which is what "at once" actually means. The only way is to have Hyperthreading or SMP for that. In the case of the SMP machine, it'll priortize the threads and divvy them up across the CPUs/Cores on the machine, to be executed as in-parallel as is possible.
On a non-Hyperthreading, non-SMP machine, it's going to execute only as fast as the one-legged man is able to get to kicking asses...
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
OK, but how fast can it run Final Cut Pro?
Ah, it can't? Suck.
...for a Java virtual machine, let alone an x86 segmented architecture. Only time will tell if superscalar chip fabs are O(log n).... but I seriously doubt, SERIOUSLY doubt! that **any** pointer arthimetic will be performed faster on such a set up.
I am very small, utmostly microscopic.
I like to call them "benchmark monkeys."
Sites like these, all they know how to do is run a couple of very common benchmarks. But they have no ability to determine if the benchmarks are appropriate for the system, nor are they at all qualified to interpret the results of the benchmark and draw any sort of meaningful conclusion.
Any monkey can look at a bar graph and say, "ooh-ooh, BIGGER!" Hardly anyone writing for these "enthusiast" sites can tell you how important that bigger number really is and what other factors in system design may have a larger impact on different types of workloads.
"An alternative approach is to have "virtual" cores - have a stack of registers and pools of computational elements. This does require some extra element of sophistication, to share out resources, but if you have two programs with very different CPU needs, both programs should run faster. "
It's called hyperthreading. It sucks because multiple threads doing different things isn't the common case (unless you run BeOS). It also sucks because the L1 cache gets thrashed a lot, which is why it's actually bad to have it on for some activities.
--
Internet Explorer (n): Another bug -- that is, a feature that can't be turned off -- in Windows.
Good god, why did they choose Matlab? It's single threaded:
/ 1- WNXI8.html?solution=1-WNXI8
"Currently MATLAB is only able to utilize one processor."
http://www.mathworks.com/support/solutions/data
Is this real enough for you?
e n& q=AMD+outsells+Intel&ie=UTF-8&oe=UTF-8
http://www.google.com/search?client=safari&rls=
I googled and couldn't find a good answer - how is dual core different than dp (other than it being one processer instead of two) and what are the benefits of it?
A non-technical response would be much appreciated.
What is this supposed to tell me?
How does a dual core opteron stack up against a dual core pentium? Seems a simple question. But I guess they thought a better comparison would be to compare two dual core opterons against a single dual core pentium...brilliant.
This just re-inforces my belief that review sites never tell you anything of any use. Horrendous graphs (like showing scores of 1660 vs: 1640, but starting the graph at 1600 so the 1660 looks 150% times better than 1640, instead of 2%).
The hyperbole also gets to me....chips/systems are always 'blowing away the competition' when they are like, 5% faster. Hey...get real. When I upgraded my P166 to an Athlon 2400+, did I really care that it might have been 5% faster/slower than it's competition? It was over 10x faster than the system it replaced! Let's get some reality in here....
But of course, we know it's all about scoops and page views for advertising anyway...integrity is gone, if it ever existed in the first place.
the
canonical
anonymous
coward
my ed session only uses 16KB!
Apocalypse Cancelled, Sorry, No Ticket Refunds
found at The Tech Report:
- x75/index.x?pg=1/
http://www.tech-report.com/reviews/2005q2/opteron
"...that's as white as it gets; all the bits are on..."
I'd also love to see how well these overclock, since I have a athlon64 3000 (1.8ghz) that does 2.5ghz, which would put it at about 4100 AMD's PR scale.
If a ~$150 1.6ghz 242 can do the same speeds as the ~$700 2.4ghz 250, a 50% overclock, then we might have another Celeron 300A on our hands :)
my karma will be here long after I'm gone
hope this link works:
X JsX3Jldmlld19JRD0xMTI5/
http://www.hexus.net/content/reviews/review.php?d
"...that's as white as it gets; all the bits are on..."
There used to be talk that the V. 34+ speeds of 33.6 kbps represented the very fastest that US phone lines could ever handle. It just wasn't scientifically possible to go any faster.
The 33.6 limit was NOT due to the phone lines, it was due to the equipment at the end of the phone lines at the telephone exchange.
At the telco end, telephone audio would be converted to and from analog and digital so that an analog signal could be presented to your phone, but on the other side a digital signal could become part of a digital packet switched network to efficiently be routed to where your phone call needed to go.
Upgrade that equipment or allow the elimination of conversion between analog and digital and you upgrade the limits. The limits of the copper lines themselves are much higher than 33.6kbit/s.
War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
Nearly all popular linux distributions now come on more than one CD (even if you ignore the source code) and the default installations are WAY bigger than that of Windows XP.
I remember Mayer Amschell Rothschild continuing the same Satanic drivvel that is rebutted and thoroughly abated down to the inhabitants of the depths of hell where not even Christ would be willing to redeem.
So sayeth Our Lord's Servant, Ben-Jammin' Frankfooter the slick Will of Jefferson Township;
"/~/He that desires a little space over a little freedom deserves neither space or freedom./~/"
NRAdude, I momentarily AM
Dude, seriously.. Go grab a PII 266.. No, I'll give you 366MHz to play with..
Try, just ^try^ to get by for a week.
With a dual-core machine, your dedicated box down at the co-lo will spit out pr0n and process CC payments faster.
Who is suggesting that these boxes are for consumers, or worse, college students?
Because there are no motherboards for dual 860's!
:(
Intel has no need to produce such as the chip is available in such low quantities