AMD's K10: a "dead" product or not?
by Johan De Gelas on May 12, 2008 12:00 AM EST- Posted in
- IT Computing general
A few years ago it was fashionable to bash Intel's Pentium 4 as a braindead architecture. The fact that the Pentium 4 Northwood (533 MHz FSB) was the best performing processor from mid 2002 until late 2003 in many applications, and that the Pentium 4 Northwood remained competitive until early 2004 was conveniently forgotten: nuances do not make good headlines.
It is now trendy to bash AMD. One" PC doctor" at ZDNet goes as far to say that:
"When I look at AMD’s current product line, all I see is a forest of
deadness. Intel has products trump every category of products
going. Server, desktop, mobile, low-end, high-end, dual-core,
quad-core. Intel has all these markets stitched up."
Nuances, who needs them when you can make a sensational headline? And indeed, the lastest desktop CPU articles here at Anandtech show that Intel's midrange CPU have a significant lead over the fastest Phenom processors.
Like any design, the K10 is a trade-off. And most trade-offs were made in favor of the applications in the server and HPC market, at the expense of games and other desktop applications.
First take a look at this page which compares a Core 2 Duo 4400 (2 GHz, 2 MB L2 and 800 MHz FSB) with a slower 1.86 GHz Core 2 Duo E6320 (4 MB of L2 and a 1066 MHz FSB). One thing is for sure: games prefer the larger L2 cache. Some of the games were up to 10% faster on the CPU which was clocked 7% lower but with twice the L2-cache. The fact that games prefer a 4 MB L2 is not going to change when you run it on a AMD CPU with integrated memory controller. A L2 can deliver the necessary data in 12-20 cycles, an IMC needs about 100 cycles.
Now, take a look at the Cache architecture of AMD's K10/Barcelona. If your run a single threaded game on it, it gets a fast 512 KB L2-cache and after that a relatively slow (44-48 cycles!) 2MB L3. If you know that the same game can benefit from more than 2 MB cache, it is pretty clear that the 512 KB L2 is not going to cope, you'll end up using the L3 a lot. A dual threaded game might need a little less per thread, but the same problem will happen again: it needs to go to that slow L3 cache all too often. Run that same game on Intel Core CPU and each thread of your dual threaded game gets a low latency 4 MB (or 6 MB) L2.
Now let us now imagine that we run 4 threads of an HPC workload on it. Each thread has a very limited number of instructions, which perfectly fit in each of the L2 caches. You get 4 threads which gets a total of 4x the bandwidth of L2. In case of Intel, each two threads have to share the available bandwidth of the L2. The amount of data is huge, so caching the data is hardly possible. The fast IMC does wonders for the K10 chip.Data that is shared between the 4 cores remains in the L3-cache and all L2 caches are kept coherent at a incredibly fast SRI. So your cache coherency overhead does not increase with the number of caches, it increases per socket. Going from 2 to 4 sockets means that you double the amount of cache coherency traffic. Compare that to the Intel platform where all L2 caches need to be kept coherent.
It is just one example why we could never expect the K10 chip to be a super desktop chip. But how is Barcelona doing in the server world? Is it limited to an HPC niche market? Well, let us see what Intel thinks. First of all, where do most of the 45 nm chips go? Just a few weeks ago, Anand reported that Intel had no intention of flooding the desktop with 45 nm Core 2 chips quickly.
Those 45 nm chips are going to the server market. Why? Several reasons.
First of all, the server market might be only 20% of Intel's revenue. But look at this:
CPU
ASP
Profit margin (estimate)
Percentage of revenue Intel Server CPU >$400 >$300
+/- 20%
AMD Server CPU
$300-$400
$220-$330
+/- 16%
Intel Mobile/Desktop CPU
$100
$40-$50
+/- 80%
AMD Mobile/Desktop CPU
$50-65
$5-$30
>80%
Secondly, Intel needs those 45 nm to be competitive in the HPC market. A 2 GHz Barcelona is capable of keeping up with the best 65 nm Xeons in those applications.
It is pretty clear why AMD focused on the server market. Without a complete redesign it is not possible to beat Intel's integer crunching power and the fast and big L2-cache and that is exactly what a modern game needs. Barcelona built further on the K8 architecture and inherited the relatively inflexible integer pipeline. While Core 2 has sophisticated reordering of loads and stores, Barcelona does a limited reordering of loads. While Core 2 offers a 32 entry queue to the integer units, Barcelona has 3 rather inflexible separated 8 entry queues.
So the right way forward for AMD was to focus on HPC and server applications where it could leverage it's strong points. We can bash AMD for being so late, and coming up with relatively low clocked CPUs, but even a 2.8 GHz Phenom would not have raise AMD's ASP significantly in the desktop market.
We are almost done with our first round of quad socket benchmarking and we can tell you that we are having a lot more fun than Anand: it is a good old exciting fight between AMD and Intel. Don't believe us? Let Intel do the talking again:
Yes, projecting the bad performance of the desktop chip to say that "AMD's products are a dead forest" is ... just silly. If you have missed the previous entries of our IT blog, just go to it.anandtech.com
Facebook
Twitter
RSS
74 Comments
View All Comments
Sunrise089 - Wednesday, May 14, 2008 - link
"But if you want a Mac with a PCI Express slot and don't want to pay $2300+, you're don't have *ANY* options"This is correct. I don't think you can say the same thing about your specifics though.
"For instance, you can't buy a $500 budget Core 2 Mac tower."
Does the PC OEM world sell $500 Core2Duo machines in full-size chassis' often?
"You can't buy something with the iMac's specs that lets you swap out the video card or that allows you to choose what monitor you prefer."
Does the PC OEM world sell all-in-one PCs at any price with 24" displays and user-upgradeable video? Do they even sell all-in-ones with 24" displays?
"You can't buy a sub-$2000 notebook nor a 13" subnotebook with a discrete GPU."
Does the PC OEM world sell a 13" notebook with a discrete GPU (they probably do, but all I normally see are 15" models)? Does the PC OEM world sell sub-$2000 notebooks with high-end Core2Duo's, 2gigs of memory, and other higher-end features with extensive software suits with discrete GPUs?
Pirks - Tuesday, May 13, 2008 - link
I was talking about the myth "3x the price for THE SAME hardware" - that one has died a long agohighlandmoose - Monday, May 12, 2008 - link
In fact for typical HPC codes Barcelona annihilates the Intel chips. We have just ordered a new HPC machine at my university and I had all the fun of running the benchmarks. For a single thread Intel was up to 50% faster, but for anything with a large memory footprint the Intel chips (even 45nm ones) with the FSB just fall over. For problems using 8GB RAM (8 threads) the K10's were twice as fast comparing 2.2 GHz Barcelona vs 3 GHz Harpertown chips. I just wish we could have had 3GHz Barcelonas...As for Nehalem I think it will help intel an awful lot, but if AMD can put on a 64 stream graphics processor with double precision capability then again performance for HPC will be stellar (compare say 64 Gigaflops [on-die ATI coprocessor @ 1 GHz] vs 48 Gigaflops [8 core Nehalem @ 3GHz]. Interesting times are ahead...
Itany - Tuesday, May 13, 2008 - link
The "raw" capability of GPU could not turn into the real application performance. For AMD HD 3k family GPU users, what's the contribution of the GPU to performance except games and HD decoding?Under the condition that GPU suitable for acceleration, Intel Nehalem+NV CUDA is the best choice. Few would prefer the AMD plantform.
Griswold - Wednesday, May 14, 2008 - link
"Under the condition that GPU suitable for acceleration, Intel Nehalem+NV CUDA is the best choice. Few would prefer the AMD plantform."And you base that claim on what? The non-existence of this combo or your apparent fanboyism? Most people live in "here and now" not in "what could be in 6-12 months if all goes well".
Locutus465 - Monday, May 12, 2008 - link
And to be perfectly honest for the money it doesn't dissapoint. What I wanted was a reasnobly priced system to get me up to date with system speed wise with an emphasis on reliability. The solution? Just go AMD. Irronically enough, AMD has been kicking some major @ss in teh driver department, an area where nVidia used to dominate. Their Vista 64 drivers are complete, fast and reliable, nVidia has only been getting to that point more recently. In fact my nvidia based laptop currently has issues handling the new Zune video's which is dissapointing :PAlso, while yeah I could build a faster intel system my current AMD build doesn't dissapointe. Running a Phenom 4x 9850, AMD Radeon 3870, 4GB OCz memory and a high end Asus board (the one with built in wifi) I'm able to play EVERY game available on the market right now (inc Crysis) with good quality settings and more than acceptable frame rates. Pretty much all of my "last gen" PC games (Doom 3, quake 4, WoW, Half Life 2 & Episode 1) are getting at, just over or near 100FPS with all quality settings maxed and FSAA jacked up played @ 1280x1024 (my 19" LCD's native res). I'm very very happy with that. Crysis I can run with many of the graphics settings set to high (not all though) and still maitain smooth game play. I don't have Crysis FPS numbers because I don't remember how to run the FPS counter in the game, nor do I know how to run a benchmark. If anyone has that info I'd be happy to post back with hard numbers.
antifanboys - Sunday, May 18, 2008 - link
Lol. You got to be joking. If you want a fast computer then get intel, that's what good right now. If you're only going to game at 1280x1024 then you have no business talking about anything. Face it, AMD is doing bad for the moment, I hope they can rebound. But your post made no sense. Are we supposed to buy the second rated processors because you dont run a FPS counter?plonk420 - Tuesday, May 13, 2008 - link
if you bought AMD (procs) ONLY to game, that was a pretty dumb choice (unless you already had "good enough" memory and previously had a compatible expensive motherboard) unless you like throwing money at AMD. and don't bother with benches. AMD is beat in all but one or two gamesi personally am an AMD fanboy, but at LEAST had a reason to buy a Phenom: x264 video encoding. for the first 10-15 days i've had mine, i've used all 4 cores at LEAST 80% of any given 24 hours. it keeps up in a linear fashion (or better) price-to-performance-wise to an Intel Quad. AND the x264 team said there's more optimizations they can do, as well (IIRC).
Ensoph42 - Wednesday, May 14, 2008 - link
Just wanted to emphasis that gaming on a phenom is perfectly reasonable. When comparing CPUs for gaming, benchmarking is done at CPU bound resolutions, typically 1024x768. If you're still gaming at these resolutions I would think that your money would be best spent elsewhere and not on a CPU. Once you get into high resolutions you less and less differences between AMD and Intel CPUs.Locutus465 - Wednesday, May 14, 2008 - link
Ditto... Just because you want to game on your PC doesn't mean you should avoid AMD... They're worth considering for sure, the platform is great.