Slashdot Mirror
AMD Releases Barton: Athlon 3000+
Harle writes "Today AMD has introduced a new version of the Athlon, codenamed "Barton," that features twice as much L2 cache as previous chips. Along with the increase in L2 cache comes an increase in the Athlon's performance rating -- specifically the new 2.17 GHz chip is rated at 3000+.
The clockrate is actually slighly lower than the Athlon XP 2800+'s 2.25 GHz speed, so the question becomes "Does the cache improve performance enough to counter the loss in clockspeed?" For the most part, the answer seems to be "yes," however, it doesn't unilaterally stand up to the 3.06 GHz Pentium 4.
With the recent delay of the Athlon 64 to September, this is AMD's top desktop chip for some time to come. The reviews are starting to pop up at Ace's Hardware and Extremetech." There's also reviews on The Tech Report, SimHQ, HotHardware, EarthV, in Norwegian on Hardware.no, and last but not least AMD's press release. I'm sure there's many many more links, but I'm tired of pasting them all in here, so post 'em below. *grin*
There is 1.1 fps improvement for Jedi knight,
and this comes at 200$+ Isnt it better to invest this money in a graphics card...
Other benchmarks also dont show marked improvement. I guess this is due to delays introduced by the much larger die size...
AMD's botched it for sure this time. I hope they bring down the pricing to a more sensible level.
My Aurora : http://www.youtube.com/watch?v=o91ZsGwJYyg
FB : https://www.facebook.com/TanveersPhotography
IBM should have their Power970's out at least a month ahead...I don't remember where, but I heard they were going to release it in the 3rd Quarter of this year, presumably along with Apple's new release of a new (G5?) Power Macintosh, or XServe (just a rumor, don't flame me), although it may not go into full production until the end of the year. What about the Itanium 2? I haven't heard anything about that. Unless its not for desktops and only for servers. If that is the case, what is Intel coming out with to join the 64-bit desktop wars?
"Some fight for law. Some fight for justice. What will you fight for? One day, you will see."
From the article on Ace Harware
This exclusive 512 KB L2-cache works together with the 128 KB L1-cache (64 KB data, 64 KB instruction) to form one impressive 640 KB on-die cache. According to AMD, the extra 256 KB cache boosts, an 2170 MHz Athlon XP from a 2700+ level to a 3000+ one.
If this is the case why do AMD, and Intel for that matter not put ever larger amounts of cache on their existing chips to achieve better performance ? Does the cost implications completely prohibit this or do the performance benefits tail off too quickly. SUN seem to able to achieve impressive performance with lower far lower Mhz (I know its different architecture) but I get the impression the large amounts of cache (2-4 MB) they use contributes significantly to performance.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
Well it is quite clear that the Athlon
architecture is at the end of its useful life.
However,the performance difference is somewhat
"exaggerated" in favour of the P4. Most of the
"content creation" applications and games
are SSE2 enhanced while, on the other hand
3dnow is propably less popular. The hard fact,
of course, is that P4 needs less time in these
applications so it is faster (whatever the
reason).
However, for general purpose usage, I firmly
believe that the Athlon is faster, mostly
because everyday applications do not need
huge memory bandwidth and cannot be made to
run with SSE.
I'm thinking that someone should start doing
some "Open Source" benchmarks where the source
is available. A good idea would be to run
a set of:
a) Kernel compile (or gcc compile or something
like that) and perhaps even "make check"
gdb or gcc or some other application (libc!).
b) MP3 compression with lame
c) Video compression with xvid or ffmpeg
d) Linpack/POVray for fpu
e) Ecasound/LADSPA for sound processing
f) Maybe a perl/high-level bechmark for some
standard system tasks.
g) Cachegrind some of the above (have a look
at valgrind/cachegrind!!)
Anyway, if someone has anything above XP 2600+
let's gather some results.
P.
I'd say that was informative and maybe insightful. Someone needs to get their finger out and develop a comprehensive, relevant and useful set of Linux (and UNIX in general) benchmarks for these platforms, especially since Linux is gaining so much market share. Just how does a SQL server benchmark on Windows 2k relate to what I do on my AMD Slackware box? How does a Windows game using Direct X have any relevance to OpenGL applications? So, come on, who's going to do it? Who's going to give us some Free (GPL preferably) benchmarking software for Linux and other UNIX-like operating systems? I can write C. I'd gladly contribute a few hundred lines of code.
Stick Men
Wow! I suppose that 4 dollar difference according to pricewatch between the 3000+ and the 3.06 means a lot to you.
And the muscular cyborg German dudes dance with sexy French Canadians
It's pretty clear which one is a better measure of relative performance. Although Toms' Hardware would not admit it for some reason.
--
If cars were open sourced, there would be at least five steering wheels in the cockpit, each operating differently -- but you'd be able to shift gears with your car stereo.
"Fighting terrorists with millitary might is like killing a mosquitor on your Dad's forehead with a rifle."
If it is like this with the 2.17, then it will be sweet when the faster cores get the increased cache.
Although I'm not really sure why I care other than when these things come out the slower ones go down in cost and that makes building clusters cheaper.
Right now I feel that the 2000+ chips are the best bang for the buck (I can make a single node in a cluster with one of those and 256M ram for under $300) - but perhaps with this thing coming out the pricing structure will shift and I can get me something faster.
hot damn.
There are some odd things afoot now, in the Villa Straylight.
With the proliferation of digital video cameras that conveniently download absolutely monstrous videos to your PC, full motion at 720x480, there is definitely a growing need for both more CPU power and more storage than is currently the norm. I have an Athlon XP 2000+ with 512MB of DDR with a pair of 60GB 7200RPM drives, and recently decided to compress a 60 minute video of a baby shower. Total compression time in Windows Media Encoder (compressing to WM9's video codec): Approximately 9 hours. (Actually a first run turned out to have a bad setting, so in real metrics it took about 18 hours) This was taking a direct several GB uncompressed AVI and verbatim compressing it to a 1.5Mbps video (which just fits on a CD-R). Of course doing any sort of processing or effects raises the bar even more if you're going to anything other than postage-stamp videos, so you could triple or quadruple if I were doing any major processing. If the Internet bandwidth were there for full-sized full-motion video again the current codecs are running at around 1:6th real time or less on current PCs.
I think right now the Athlon XP 3000+ based on the Barton CPU core is the right first step, but I think the CPU that will REALLY worry Intel will likely come later this spring when we see Barton core Athlon XP's that take full advantage of DDR400 DDR-SDRAM.
Remember, under pure-CPU tests the Athlon XP 3000+ has almost the same performance as the Intel Pentium 4 3.06 GHz CPU with HyperThreading; what will happen when the Athlon XP gets the Front Side Bus speed bump necessary to support DDR400 memory?
What I've heard is that the MP chips are the "cream of the crop" of the XPs - AMD manufactures a batch, and then picks out the best to be MPs. So you do get something for what you pay for.
It's hard to be religious when certain people are never incinerated by bolts of lightning.
That could be useful in embedded applications. Running the entire app in the cache, at higher speed than in main memory, could be a Very Good Thing.
Best Slashdot Co
Oh come on... haven't you figured out by now that all new chips start expensive, and in a year are the bottom of the barrel, bargain basement, can't buy anything slower deals? And that all top-of-the-line chips are a marginal improvement for way too much money?
Do you know what bottom-of-the-barrel is right now? It's like an Athlon 1800+ depending on where you shop. (Gee, I was just in there last week and they were still selling 1.2GHz Durons...) Do you know how much an Athlon 1800+ cost when it came out? Do you really think this price is permenent?
What's the alternative? Never introduct a chip until it's cheap? Doesn't work that way, for a whole lotta reasons.
Is it really 640Kbyte in total? That would happen only if every address in the L1 were not also in the L2. If it's possible for L1 and L2 caches to overlap, then it would be more accurate to say '512Kbyte of cache in total, of which 128Kbyte can run rather quickly'.
I mean, nobody says their system has '128.5 megabytes of memory in total' if they have a single 128 meg memory module and 512Kbyte of cache in the CPU.
-- Ed Avis ed@membled.com
depends on which review you read and with whose devils eye you read it with.
you do realise that p4 3.06 is supposed to be better/faster than p4 2.8 would be at 3.06 speeds too? does that make calling p4 2.8 a p4 2.8 wrong, or calling 3.06 a 3.06 wrong? you can't compare even intel products just by the number they're sold with.
it certainly doesnt get 'bitchslapped'.
intel is far worse. they sell celerons and p4's with essentially same 'horsepower' numbers(to the consumer). not to mention their ht tech they're trying to make look far more better than it is. if you think the common joe can identify the differences then you're really wrong and haven't do enough mandatory helpdesking for your relatives.
intel is like a car dealer selling cars by just telling consumers how high they can rev, amd is telling a performance number(though, mostly just comparable to other amd cpu's, and thats alright, why should they change their rating system according to what their competitors do.)
i'd prefer to know the actual kw rating of the engine rather than just it's displacement and how high it can rev.
world was created 5 seconds before this post as it is.
Holy cow! 60 minutes?
I have to admit, that may have something to do with the codec. However, last night I compressed a movie - split into 2 - into Divx 5.02 Pro, using 2 pass encoding, 1 hour of compressing over both passes took 2 hours. Add another 5-10 minutes to transform the audio to 256kbit mp3, and it was complete... well, the first half. The bitrate was equivalent - if I recall, around 1350kbps for the video.....
I've got a similar setup to you - XP2100+, 512mb 333ddr cas2 mushkin, and 2x80gb 7200 8mb WD drives striped. I'm just curious if the WME is what is taking so long.....
Karnal
This is much better. http://www.anandtech.com/printarticle.html?i=1783
--
If cars were open sourced, there would be at least five steering wheels in the cockpit, each operating differently -- but you'd be able to shift gears with your car stereo.
"Fighting terrorists with millitary might is like killing a mosquitor on your Dad's forehead with a rifle."
Power Consumption and Self Shutdown were both addressed with the Palomino (Athlon XP) core.
Also I find it hard to understand how 1) a CPU that follows a published standard (X86) could be an emulator and 2) how this emulator can outperform a "pure" system at a slower speed? This could be due to my own ignorance though, please educate me.
Dissolve... Resolve... Evolve...
A serious question: If I booted up vanilla DOS on this thing, would it pretty much all run inside of cache?
That would be insanely fast.