ABOUT BENCH FORUMS PODCAST LOGIN REGISTER
PC Components▼
Smartphones & tablets▼
Systems▼
ENTERPRISE & IT
GUIDES▼
DEALS

Processor Obfuscation and You: An Update on AMD and Intel CPUs

 by Kristopher Kubicki & Jarred Walton on May 17, 2005 12:00 PM EST
55 Comments | Add A Comment
55 Comments

Socket 939

The best alternative is all that remains, and that's Socket 939. With dual core Athlon 64 X2 chips planned at least through the 4800+ (2.4 GHz with 1MB L2), 939 has quite a lot of headroom remaining. At present, there are at least four variants of Athlon 64 processors for Socket 939, and two more are planned for the near future.

The oldest of the offerings is the Sledgehammer core, which was used in the FX-53 and FX-55. (The FX-55 used strained silicon, but we're not aware of any additional code name for that model.) Sledgehammer (which we often just interchange with Hammer) was also used for the original "CG" Opteron processors, and there was a slight modification made to allow the use of non-registered (unbuffered) RAM. Performance is very good; in fact, the Sledgehammer was, until very recently, the fastest processor around. The major drawback is the same as for Clawhammer - the core was expensive to produce, due to the 130nm process and the 105+ million transistors. That brings us to the next processor core.

Newcastle was available for either 754 or 939 and used a 130nm process.   Newcastle differs from Hammer primarily by the reduced L2 cache size and "CG" stepping. Since the 1MB cache was about half of the total CPU core, reducing the cache made the core substantially smaller and thus, cheaper to produce. What's interesting to note is that the Socket 939 CPUs had a huge price premium relative to their Socket 754 counterparts - a price premium that appears to be related to supply/demand and price/performance concerns more than anything else. Consider that both the 3400+ and 3500+ are basically identical, other than the socket. Both are 2.2 GHz 512K cache chips made on a 130nm process. However, the Athlon 64 3500+ carried a roughly 50% price premium. The end result is that Socket 939 chips tend to outperform their Socket 754 counterparts, due to the dual channel memory controller. The market basically dictates the prices that AMD and Intel can charge, and the market determined (with good reason) that Socket 939 was preferable to Socket 754.

The follow up to the Newcastle core is Winchester, which is more or less the same chip design, but only produced on a 90nm process instead of 130nm. We say "more or less" because there are measurable performance differences, and we're not entirely sure what caused the increased performance of the 90nm parts. More important than the performance difference - it's generally less than 5% and sometimes as little as 1%, which means that you won't notice it in practical use - is that the 90nm SOI parts run substantially cooler than the 130nm parts. Our system power draw tests have shown that even the fastest 90nm parts still draw less power than the slowest 130nm parts. In fact, the dual-core 90nm parts are also less power-hungry than most of the 130nm parts. But we're jumping ahead.

Newcastle 3500+ models are still available at a discount relative to the newer 90nm parts, again most likely due to marketing factors rather than the actual cost to produce the chips. AMD may also be trying to clear out the old inventory before making it obsolete. The end result is that Winchester 3200+ and 3000+ parts are cheap and readily available, while the 3500+ Winchester costs quite a bit more than the 3500+ Newcastle. For most people, we'd either go with the cheaper and slower 90nm cores or make the jump to one of the higher performance 90nm cores; the 3500+ 90nm just isn't a very good bargain right now.

Generally, you can determine a processor is a Winchester CPU if the SKU ends in "BI". Winchester processors only utilize the "D0" stepping that we mentioned earlier, so don't expect SSE3 either. Although Winchesters are certainly a step up from the 130nm Newcastles, they are not the only 90nm chips around.

The most recent addition to AMD's lineup is the new Venice core. It appears to be the same as Winchester, only with SSE3 extensions enabled. Applications that take advantage of SSE3 are still few and far between, and even then, the difference isn't tremendous. It's sort of like the x86-64 advantage when using less than 2 GB of RAM. The extra registers might improve performance by 5% or so, but it's not a huge deal. One thing that does differentiate Venice from Winchester is that Venice parts are also available in 3800+ models. Winchester tops out at the 3500+ (2.2 GHz) part right now, which is odd, considering that many overclockers are reaching 2.6 GHz and beyond quite easily. Given the recent overlap in chip names from AMD, perhaps we'll even see 2.6 GHz Venice parts in the future as an alternative 4000+, but that's pure speculation.

San Diego is the follow-up to Sledgehammer; the 1MB cache version of Venice. The new Athlon 64 3700+ and Athlon 64 4000+ parts use San Diego, as will the FX-57 when it's launched sometime in the next few months. San Diego chips have already begun to show up at retail (albeit in small quantities), and the prices are pretty attractive, for a high-end processor. Judging by the performance comparisons between Newcastle and Winchester, we would expect San Diego to outperform Sledgehammer clock for clock by a few percent - this is mostly due to caching algorithm optimizations, and the SSE3 additions are merely one more point in favor of the new core. We also expect the power consumption of San Diego to be significantly lower than Hammer, and quite a bit lower than Newcastle.

Venice and San Diego both use the "E" stepping cores, although there is actually a distinction between the steppings this time around. Venice, denoted by the "BP" at the end of the SKU, use the "E3" stepping. San Diego, on the other hand, use the "E4" stepping and are denoted by "BN" in the SKU.

Looking forward to later this year, we'll see the arrival of the Manchester and Toledo cores. Toledo is the dual core version of San Diego and it will power the Athlon 64 X2 4400+ and 4800+ parts (1024K L2 per core); the Manchester core will be used in the 4200+ and 4600+ (512K L2 per core), though we may also see "rejected" Toledo cores in those chips if the past is any indication. We've covered dual core performance in several recent articles: Intel Part I and Part II, along with the nForce4/855 comparison, and then the AMD Opteron and Athlon X2 articles. The basic summary is that anyone who runs a lot of applications at the same time - especially CPU intensive applications - will definitely benefit from dual core processors. The only fly in the ointment for AMD is that their dual core parts are currently targeting a much higher price range than their single core processors. In fact, the cost of a dual core chip like the 4200+ (2.2 GHz and 512K cache per core) is actually more than the cost of two 3500+ 130nm chips. That's definitely not the way to spur market adoption of dual cores, but it may be a business decision caused by the limited manufacturing capacity of AMD.

As a final topic that we didn't address yet, let's talk about overclocking for the various Socket 939 cores. Newcastle and Sledgehammer both top out around 2.5 GHz for the best chips, and many chips won't even reach that high. The heat and power requirements certainly play a role, although it's interesting that the FX-55 core with strained silicon added into the mix manages to run at 2.6 GHz, with overclocking attempts getting it as high as 3.0 GHz. The better choice for overclocking is definitely the 90nm parts. As usual, nothing is guaranteed with overclocking, and it's almost equal parts luck, components, and - for lack of a better term - artistry that determine your final clock speed. If you know what you're doing, we would venture to say that 2.5 GHz is achievable on all of the single core 90nm parts, though the 9X multiplier of the 3000+ may make it more difficult to reach. Early previews of the X2 chips are also showing results above 2.5 GHz, and when you think about it, it makes sense.  The thermal density is actually the bigger problem with CPUs these days, so having a core that's twice as large in transistor count as well as die size basically means that the highest attainable clock speed (with a sufficient heat sink) will be similar. We won't know for sure until the retail launch when we can see results for more than a few hand-selected CPUs, but even without overclocking, the performance offered by the Athlon X2 is impressive.

Socket A

We don't consider Socket A/462 to be a valid platform for anyone anymore, at least for new purchases. If you're still running a socket A system and you want to prolong its life, your best bet is to grab an Athlon XP Mobile processor and overclock it. The Athlon XP-M 2500+ [RTPE: AXMH2500FQQ4C] has basically been at a flat price of $90 for a while now, and we don't see it dropping any further. The 2600+ increases the clock speed slightly, but since both CPUs have unlocked multipliers and can overclock to 2.2 to 2.4 GHz with relative ease, there's not much point in spending any extra money (though, it is only $6). If you're already running an Athlon XP 3200+, you're pretty much maxed out on the CPU and the only real performance gain will come by purchasing an Athlon 64 motherboard and CPU. Performance of a 2.4 GHz Barton core is roughly equal to that of the Athlon 64 3000+ - the integrated memory controller essentially improves performance by 20% relative to the older Athlon XP design.

As for the individual cores, they all used 130nm process technology, and the maximum clock speed was similar - around 2.2 GHz. The added cache of the Barton core helped performance quite a bit, as did the increase in bus speed that came with the later parts. Modern Sempron chips for socket A use the Thoroughbred core, due to its smaller size. The one exception is the Sempron 3000+, which uses a Barton core clocked at 2.0 GHz. It's interesting to note that the Barton 2.0 GHz would have originally qualified as something like an Athlon XP 2800+, only 200 points below the Sempron rating. The Thoroughbred 2.0 GHz, on the other hand, was labeled as an Athlon XP 2400+ and later the Sempron 2800+, a difference of 400 points. Most of us realize that the performance ratings of the later Athlon XP chips were more marketing than reality. Joe Consumer figures a 3200+ is the same as a 3.2 GHz, right?  The Sempron chips are now supposed to be compared with Intel's Celeron line rather than Pentium 4, and their model numbers seem to reflect true performance ratings better.

AMD Processors Intel Processors
PRINT THIS ARTICLE
Comments Locked

55 Comments

View All Comments

  • JarredWalton - Thursday, May 19, 2005 - link

    justly, that's why the detailed AMD specs weren't put out in the article. I dislike both naming schemes to a large extent, but I'm something of a techno-snob. Heheh. I liked getting FSB, core speed, cache amounts, etc. as the CPU name rather than a generic "3200+" or "Pentium 531". The main idea was to explain what each of the specific CPU cores offers. Some people will prefer AMD's PR ratings to Intel's ratings. Either way, it gives guys like me something to write about. :D

  • justly - Wednesday, May 18, 2005 - link

    Jarred, you seem to be a overly concerned about the AMD numbering scheme. While it does have some flaws it still seems better to me than Intels system, for most people at least.

    In fact I would go as far as to say that AMDs PR numbers where better even before Intel stopped using their MHz/GHz numbering policy.

    You say that its your friends and relatives that you worry about, well how many of them (without your help) know the difference between a 505, 520, 530, 530J, and a 630? of course they could ask a retail sales accociate (now there is something to be worried about).

    I remember a few months ago (maybe a year) going to Newegg and doing a search for Intel and 2.4 GHz and recieving no less than 8 uniquely different CPUs in the result. Yes this was a combination of Celerons, P4s and Xeons but even then I think at least 4 of them where P4s. They varied in bus speeds, HT support and even cores and cache if I remember correctly.

    Ok, so there are 10 CPUs that use the 3000+ designation. I'll agree that using the Sempron name for socket "A" AND 754 shouldn't have been done. On the other hand I see no problem using the same PR for Athlon 64 and Athlon XP, as these are totally different platforms (no worse than seperating performance by using a letter behind the GHz of a CPU like Intel has done in the past with their 400, 533 and 800 MHz versions).
    It would also have been nice to see more emphisis placed on how the Athlon XP and 64 processors are rated (many people still think that a Thunderbird is used as the base CPU for the A64 rating). Maybe I am wrong (its been a long time) but I seem to recall reading (on Anandtech I believe) that the A64 is based on a different set of benchmarks than the XP and uses the 1800+ XP as a baseline.
    Anyway, if we eliminate the models that compete in different classes (mobile and entry level either by design or age) that only leaves 3 models of the 3000+ (Newcastle, Winchester and Venice) only 2 of those share the same platform and they have the same clock speed, cache and bus speed.

    Sure, it can be confusing for some but no worse than comparing 400MHz FSB 2.4GHz P4 using single channel pc133 memory to a 800MHz FSB 2.4GHz P4 W/HT running with duel channel pc3200 memory. In fact all Athlons recieved far less of an impact from varying memory or bus speed, so the average (non-geek)
    consumer was more likely to get preformance close to what they expected from AMDs PR numbers.

    I know it must sound like I am biased (I probably am, even though I try not to be) but if anyone wants to know exactly how any of these CPUs (AMD or Intel) compare a crib sheet is almost a necessity.

    I'd have to say the best thing about these numbering schemes has more to do with the transition to AMDs 939/754 and Intels LGA 775 since now its more difficult to make a faster CPU perform pooly becase of the wrong choice of chipset or memory.

    All in all, both systems work and both have flaws. There seems to be only two answers to this problem, be informed or trust some one else to give you the correct information.

  • Tujan - Wednesday, May 18, 2005 - link

    Well JarredWalton, bottom line is MAKE IT A 939 OR 775 PCI EXPRESS. AND MAKE IT FOR LESS THAN 1300 DOLLARS. Tell me it can be upgraded and let me be happy about it. They are not 'value machines,they are not 'low-end machines. If the criteria for a low end machine is 256MB of RAM,then we are really in trouble for words arent'we.
    You and I can put them together,. So why would I be insulted by somebody showing me some 4000$ rig asking me for my money on it ? WE KNOW THAT.Back off so somebody can enjoy some technology. Dont need all those PCI slots,or usb slots,or duel gigabit ethernet for example. Or 'top not firebreathing graphics card. AND YOU KNOW WHAT I MEAN.
    Why ride this technology around when nobody can get on.Bring on the power for your money,its never been here before.A great opportunity to put it where your mouth is for real. Like everybody we are hungry and I know you got it.

    And for the AMD numbering.If you could be satisfied to referencing simply socket #s this would be welcome enough. Then when you think about it,I believe Intel changed to numbering scheme so that AMD would have nothing to refer to and they dont.
    AMD Winchester...
    Intel Prescott w,hyperthreading....
    both winners.

    939s w PCI-e, 915s w PCI-e......

    Less than 1300s and they are really good technology. Worth the money.

    Tell me Im on a salvage crew.

  • JarredWalton - Wednesday, May 18, 2005 - link

    If you want to download an Excel spreadsheet containing the table from my post #41, here it is:

    http://images.anandtech.com/reviews/cpu/amd/cores/...

  • JarredWalton - Wednesday, May 18, 2005 - link

    37 - I'm not sure what that was about. Sorry.

    JohnsonX: You can still find the 939 3400+ around at a few places. Check out our pricing engine:
    http://labs.anandtech.com/links.php?pfilter=2051

    There were three retail outlets when I started the article, and now Axion Tech is the only one showing up. Oh, well.

    True, the equivalently rated Athlon 64 chips are also equivalent performance, give or take. Here's something I was considering including in the article that I decided to cut. Remember: we're the "informed public" - it's my friends and relatives that I worry about.

    -----------------------------
    If you *don't* think AMD's names are confusing, than you're a geek like me and you know your CPU cores by heart. However, consider the following table of AMD processors: [I'm going to put this up as a downloadable file in a minute, as it doesn't look good without HTML markup.]

    CPU Name Core Name Clock Speed L2 Cache Bus Speed Socket Process Notes
    --------------------------------------------------------------------------------------------------
    Sempron 3000+ Thoroughbred 2.0GHz 512K 333 462 130nm
    Sempron 3000+ Palermo 1.8GHz 128K 400 754 90nm SSE3
    Athlon XP Mobile 3000+ Barton 2.2GHz 512K 266 462 130nm
    Athlon XP 3000+ Barton 2.16GHz 512K 333 462 130nm
    Athlon XP 3000+ Barton 2.1GHz 512K 400 462 130nm
    Athlon 64 3000+ Newcastle 2.0GHz 512K 400 754 130nm
    Athlon 64 DTR 3000+ Clawhammer 1.8GHz 1024K 400 754 130nm
    Athlon 64 Mobile 3000+ Clawhammer 1.8GHz 1024K 400 754 130nm
    Athlon 64 3000+ Winchester 1.8GHz 512K 400 939 90nm
    Athlon 64 3000+ Venice 1.8GHz 512K 400 939 90nm SSE3
    --------------------------------------------------------------------------------------------------
    Athlon XP 3200+ Barton 2.2GHz 512K 400 462 130nm
    Athlon 64 3200+ Clawhammer 2.0GHz 1024K 400 754 130nm
    Athlon 64 3200+ Newcastle 2.0GHz 512K 400 754 130nm
    Athlon 64 DTR 3200+ Clawhammer 2.0GHz 1024K 400 754 130nm
    Athlon 64 Mobile 3200+ Clawhammer 2.0GHz 1024K 400 754 130nm
    Athlon 64 3200+ Winchester 2.0GHz 512K 400 939 90nm
    Athlon 64 3200+ Venice 2.0GHz 512K 400 939 90nm SSE3


    Ten (TEN!?) CPUs all bearing the "3000+" designation, and seven with 3200+. Hell, there are six and five (respectively) "Athlon 64" chips with each name. And I probably missed one or two less common variations.

    Yes, I "unfairly" included Athlon XP and Sempron into those charts. The problem is, I hear from people on a regular basis that think the model number is all that matters. "Celeron, Pentium, Athlon, Sepron... it's all the same, right?"

    Curse the marketing departments. Although, it keeps guys like me employed, so maybe I shouldn't complain too much? Anyway, maybe it really doesn't matter that much to Joe and Jen Consumer what type of chip they actually get, but at some point they're bound to have a case of someone telling them that, no, their system isn't as good as another system.

    At that point, they're in bad shape, because they really have no idea what they're looking for, and unscrupulous sales people can now convince them that they "need" a new computer. Especially if they get conned into buying a 256MB RAM system initially, because when they feel it's too slow, the salesperson isn't going to say, "Oh, you just need to spend $50 on more memory."
    -----------------------------

  • johnsonx - Wednesday, May 18, 2005 - link

    A final point of interest:

    On their SKU list, AMD no longer admits the existence of the "obscure 2.2 GHz 512K 939 chip that is limited to an 800 MHz HyperTransport link".

    I know it did exist, NewEgg sold them. It was a Socket 939 chip called a 3400+. AMD now says ALL 939 chips support 1000Mhz HT, and says ALL 3400+'s are socket 754.

  • johnsonx - Wednesday, May 18, 2005 - link

    I'd also point out, despite what poster 21 said, there don't appear to exist any Paris core Semprons below 3000+ or above 3100+; D0 and E3 Palermos are the only Semprons for 2600, 2800, and 3300. The not yet available 2500+ will be E3 only, while the 3000+ and 3100+ are covered by all three Sempron cores (though the only Paris core I can actually find at NewEgg is the 3100+).

    My guess is the 3000+ Paris cores are the ones with some bad cache.

    BTW, have I complained again that AMD selling 64-bitless Semprons is a big mistake?

  • johnsonx - Wednesday, May 18, 2005 - link

    AMD's model scheme is designed to allow AMD to bring different processors into and out of production without changing marketing names. Changing marketing names confuses the masses; AMD doesn't have to worry about confusing the enthusiasts - enthusiasts will figure it out anyway.

    From all the testing I've seen over the years, there is little arguement that all the various processors in the same family with the same model numbers do in fact perform fairly close to each other by most measurements. Sure, there are always a few benches that favor the big-cache versions, and a few that favor extra clock speed, but overall, the various A64 3200+ chips perform about the same, and likewise the 3000+ and 3400+ variants.

    If you are a typical end user, Athlon64 3200+ tells you all you need to know... any extra digits would only confuse the issue.

  • Tujan - Wednesday, May 18, 2005 - link

    Ok. So I'll be a little critical here. Even though adjoining 'marketing,in order to keep your presentations advertisers consistent,and happy,it is still 'slightly 'goody two shoes,to constently only do reviews with the'top notch processors and motherboards. Then leave the stats for 'the rest of us out,while we could use some of that 'being informed,just as well as the benifits of using the 500.00,and 1000$ processors.
    True,you've got business quarters,deadlines and the likes,and perhaps ...maybe its not your place to create the relationship of an 'informed buying public. While you do anyway,there is always the 'nitch of us that are trapped with our technology wich now,we might be considered 'loosers to be advocating using,buying,or convey as 'good technology.
    Just a little historical reference first for this rubber band train. The AMDs - the Socket 775s didn't have what ? PCI-express,this to them is only several months past if they have it at all. The same with the 939s,they really have been 'marketed for less than a year !! They really really have PCI-express slots on them. But alas,they are using DDR400 memory,and still am I advocating 'new technology,or market speak to say I dont think DDR2 is a looser.

    Then Intel,the 865s...they finally got dual-channel memory. But they didn't have pci-express.And the memory is still DDR400.
    Go to the 915s,they have (some of them)and finally PCI-express. With a large assortment of hyperthreading cpus (thats two processors working in one package). And the 925s well here finally all the stops are in,but now,we have not only dual core 'hyperthreading,but we have affordable dual core ACTUAL processors. And your investments of perhaps the past several months are still performers.
    Comparitively speaking,Intels processors will outperform any of the non FX AMD processors besides in multitasking. And monitarily speaking,the AMD processors have been very expensive. Then seeing as 939 has only been around for less than a practical year you still would be hardpressed to find the difference created in either virtual-cores,or multitasking using them.
    Well the difference between the platforms isn't what Im server to get to.Fact is the show has always been the fastest most powerful computer there.With a front page show,of something that costs thousands of dollars.Frankly Ive got to tell you,that today,if that 3 or 4 thousand dollar computer was on the front page I would like to fire the salesman.
    Simply on a basis of fact,that there are performers there,of wich plenty of people would like to participate to.And now they dont have to spend the for the top notch ticket to do so.
    A 939 mb with a gig of ram,a so so AMD processor,gives an upgrade path far into the future.And is very good participation to the technology.
    An Intel platform,utilizing any of the 2.8 or better 775s is as well 'good enough a welcome to someone wanting to participate as well. And you dont have to spend 4 grand to do so.
    Both platforms can have very competitive systems for just under 1300s. AND THEY ARE WORTH THE MONEY!!!!.
    And they are BETTER THAN PREVIOUS TECHNOLOGIES!!!.
    They should be on the front page.Why is it that such systems by OEMs are saying,oh low end,'low cost,value machines ? These are performers make no doubt about it.

    You can take a closer look at this than me.But I would prefer being an informed happy customer.Than someone who is said to be a looser becuase tommorow,my system is shut down.

    Really dont think its like that.But I dont see that anybody is confirming this to me.We need the systems,now,.You know how to build them.So a suggestion is to just shuck corn like that,and see what the gross is.

    Mean these systems are really atractive !!!.Back those expensive things to a later page or something.


  • Quanticles - Wednesday, May 18, 2005 - link

    I'm pretty sure AMD wants you to be confused. If they didnt, they'd just call the parts by the rev's. =)
Copyright © 2024. All rights reserved.
BENCH
TOPICS
FOLLOW
ABOUT

Log in

Don't have an account? Sign up now