Affordable Dual Core from AMD: Athlon 64 X2 3800+
by Anand Lal Shimpi on August 1, 2005 9:36 AM EST- Posted in
- CPUs
AMD's Efficiency Advantage?
Before we get to the actual barrage of performance tests, there is one issue that we have been wanting to tackle for quite some time now.AMD has often argued that their dual core architecture is inherently more efficient than Intel's, primarily because of their System Request Queue (SRQ). All core-to-core transfers occur via this queue instead of over a main, shared FSB, which is the case in the Pentium D.
Johan put AMD's architecture to the test by measuring the latency of cache-to-cache transfers in AMD's dual core chips vs. Intel's. The results were quite impressively in favor of AMD's architecture. Cache-to-cache transfers on Intel's dual core CPUs took over twice as long as on AMD's dual core CPUs, but at that time, we could not find any real world benefit to the architecture.
Armed with a bit more time, we went through all of our benchmarks and specifically focused on those that received the most performance gain from dual core architectures. Using these multithreaded and/or multitasking benchmarks, we looked at the performance improvements that the dual core processors offered over their single core counterparts. For AMD, making this comparison was easy; we took the Athlon 64 X2 3800+ and compared it to its single core equivalent, the Athlon 64 3200+. For Intel, the comparison is a bit more complicated. The inclusion of Hyper Threading makes the single-core to dual-core jump a little less impressive in some cases, thanks to the fact that virtually all single-core Pentium 4 processors these days can execute two threads simultaneously. Thus, for Intel, we had to look at HT enabled, dual core and dual core with HT enabled, all compared to single core performance to get a complete picture of Intel's multithreaded performance scaling.
Remember that all performance increases are with reference to a single core processor, and in the case of Intel, we are talking about a single core Pentium 4 with HT disabled. More specifically, we used a Pentium D 830 (3.0GHz) for the dual core tests and compared it to its single core counterpart - the Pentium 4 530 (3.0GHz).
First, we have our Winstone 2004 benchmark suite; we omitted Business Winstone 2004, since it shows virtually no performance boost from dual core CPUs and instead, focused on Multimedia Content Creation Winstone 2004 and the Multitasking Winstone tests.
It is also worth noting that although Hyper Threading improves performance with a single core, enabling HT on the dual core CPU actually yields lower overall performance than if we had left it off (+24.1% vs. +39.1%). Johan explained exactly why situations like this exist on the Pentium D in his "Quest for More Processing Power".
Next up is the SYSMark 2004 suite. In all but two of the tests, AMD scales slightly better than Intel when going to dual core. The scaling advantages aren't huge, but they are tangible in some of the tests.
Once again, while Hyper Threading itself tends to impress, HT + dual core gives us a mixed bag of results, sometimes outperforming dual core alone while falling behind other times.
Finally, we have our application-specific benchmarks; here, we have AMD scaling better than Intel in 3 out of the 5 tests, but then in the remaining 2, Intel scales better.
While the Athlon 64 X2 does have much better cache-to-cache transfer latencies than the Pentium D, it appears as if for the most part, those advantages don't surface in real-world desktop usage. That being the case, the Athlon 64 X2 3800+ must outperform the Pentium D 830 based on the performance advantages of its individual cores in order to win this battle, not based on any dual core architectural efficiencies. So, does it?
Facebook
Twitter
RSS
109 Comments
View All Comments
masher - Wednesday, August 3, 2005 - link
> " believe he meant to post this review showing the X2 3800 matching or besting the XE840 in all but a handful of tests:"Let's see...the 840 wins 3DMark, half the Cinebench, Lame Encoding, DivX encoding, MPA encoding, SiSoft FP, Sisoft Multimedia, and Sphinx.
The 3800 wins the 3D shooters, PoVRay, ScienceMark, picColor, and half the Cinebench.
The Intel results are "all but a handful"?
PrinceGaz - Monday, August 1, 2005 - link
I was surprised that no attempt was made (or posted in the review) to overclock the X2 3800+, given that most people reading it here would want to know what headroom it had.Fortunately most other good sites did overclock it and post their results, and the general consensus is 2.4-2.5GHz was as much as they could do with air-cooling at reasonable voltages. Looks like the X2 3800+ really is all the speed-binned rejects, at least from the samples AMD sent out to review sites.
Houdani - Monday, August 1, 2005 - link
Well, Anand did to a tiny bit of overclocking, he just didn't do any benches with the overclock. The above snip was taken from the final page of the article.
[L]http://www.anandtech.com/cpuchipsets/showdoc.aspx?...[/L]
PrinceGaz - Monday, August 1, 2005 - link
Apologies to Anand, I was skimming through the article rather than reading it in full, and missed that. Seems like 2.4-2.5GHz is all the current batch of X2 3800+ can do, which is less than impressive. Ordered a X2 4400+ today as a result.Hacp - Monday, August 1, 2005 - link
They had a DDR2 533 1gb kit for 80 bucks at monarch I think. But I think the DDR2 667FSB chips are more expensive...Joepublic2 - Monday, August 1, 2005 - link
"Except that by the time you get a new motherboard that supports that cheaper Intel chip and the pricier DDR2 memory, any price advantage Intel had is already gone. And the AMD system still outperforms the Intel one 15% in for the 830D and 20% in the 820D. So choose wisely...""From what I can see, Intel 955 chipset motherboards are around $180 - $220, which is a lot more expensive than a decent Socket 939 motherboard. SO overall thge system cost is much cheaper than the Intel 830, and very close to the 820."
You guys obviously haven't checked prices lately. At newegg, you can get a nice ASUS 945 board for $130, which while being more expensive than an EPOX 939 at $85, is much lower than "$180-220". Also, 2GB of DDR is $177.98, and 2GB of DDR2 is $186.00. But I agree, it's worth the extra $50 over the Pentium D 820 platform.
krisia - Monday, August 1, 2005 - link
Hmmm, zipzoomfly has a 945 mobo for $110 and 1GB DDR2 ram for $92.Looks reasonable to me...
http://www.zipzoomfly.com/jsp/ProductDetail.jsp?Pr...">http://www.zipzoomfly.com/jsp/ProductDetail.jsp?Pr...
http://www.zipzoomfly.com/jsp/ProductDetail.jsp?Pr...">http://www.zipzoomfly.com/jsp/ProductDetail.jsp?Pr...
Hacp - Monday, August 1, 2005 - link
IMO Anand's article was the fairest. He compared the Intel 830 to the AMD 3800. Some sites compared the 820 to the 3800+ and it obviously was not fair to Intel to do that.Also, if you want an entry level processor and want multitasking, you should go with an HT enabled Intel. At 240 dollars, you should be able to buy a 3.2 GHZ Intel processor with HT enabled, which will perform WAY better at single threaded tasks and perform well in some multithreaded applications. Thats what ENTRY is. You sacrafice future performance for performance right now.
Finally, Anand, I would have liked to see the 3800+ compared to the 3.2 GHZ Intel Dual core processors. Why? Because AMD's PR numbers for single threaded applications indicate that the 3800+(2.0 GHZ with 512 Cache AKA the 3200+) should be comparable with the 3.2 GHZ processor. You shouldn't soely compare the processors on price, but also the PR system :). A 3.0 and 3.2 Intel Dual core comparision with the 3800+ X2 would have been a nice......
cryptonomicon - Monday, August 1, 2005 - link
that chip looks awesome! how come they don't have that benchmark called um,Comparison over number of Threads
or something, where they run some sort of bench over 2, 3, 4, 5, 6, threads. I always thought that was an interesting one to see how effective each processor handled multiple threads.
Samus - Monday, August 1, 2005 - link
There is no reason at this point to go with an Intel based system for general productivity (business, graphics, compression, a/v editing, 3D modeling, CAD) AND gaming (where AMD had always excelled)I couldn't agree with Anand more. Where you loose single threaded/task performance (from clockspeed) you more than make up for it in multi threaded/task performance. The overclocking potential is also fairly good, 2.46GHz is nothing to sneeze at when your talking about a 2.0GHz budget chip for 350 bucks.
Also, comparing 512KB dual cores to 1MB dual core's at the same clock speed, there is little performance difference because since the cores communicate directly via L2, not via HT (or bus like Intel) your theoretical performance-based L2 is 1MB and 2MB respectively. AMD's memory controller is so efficient to begin with that the difference between 1MB and 2MB L2 cache is very little.
The cache's are also independent (like they always have been, ie L1 and L2 being non-consecutive) so they don't hold the same information.
Boy, they've come a long way.