Slashdot Mirror
AMD Sued Over Allegedly Misleading Bulldozer Core Count
An anonymous reader writes: A class action suit accuses AMD of misleading buyers about the number of cores in its Bulldozer-based CPUs. The complaint claims that the chips effectively had only four cores, while AMD claims there are eight. According to Ars: "AMD's multi-core Bulldozer chips use a unique design that combines the functions of what would normally be two discrete cores into a single package, which the company calls a module. Each module is identified as two separate cores in Windows, but the cores share a single floating point unit and instruction and execution resources. This is different from Intel's cores, which feature independent FPUs. The suit claims that Bulldozer's design means its cores cannot work independently, and as a result, cannot perform eight instructions simultaneously and independently. This, the claim continues, results in performance degradation, and average consumers in the market for a CPU lack the technical expertise to understand the design of AMD's processors and trust the company to give accurate specifications regarding its CPUs."
http://www.extremetech.com/extreme/217672-analysis-amd-lawsuit-over-false-bulldozer-chip-marketing-is-without-merit
AMD's CPU architecture has a similar purpose as hyperthreading -- to share hardware resources between what looks to the OS like independent cores -- but the tradeoff is different. Intel's hyperthreading approach only works to cover memory latency, because the hyperthreads share so many physical resources (I think basically everything except register files and hyperthreading-related state). AMD's is somewhat different in that each "module" has two independent integer ALUs, register files, and L1 data caches. The module has one L1 instruction cache, one L2 data cache, one FPU, and one instruction fetch/decode unit.
But AMD has always been pretty up-front about this architecture. There is maybe a cause of action against resellers who package the AMD chips into systems and do gloss over which aspects each "core" shares with another core, but AMD publicly presented the core-vs-module distinction well before the chips were released.
Why is this even reported? This suit isn't going to go anywhere (unless AMD's lawyers are extremely incompetent, and the judge is extremely incapable of understanding basics about computer architecture and ISAs).
The AMD cores shared an FPU, sure, but sharing a resource doesn't mean that cores cannot execute simultaneously. The AMD cores still have independent integer-based execution units (instruction registers, register files, ALUs, branch counters, etc.), after all, and are fully capable of executing integer instructions simultaneously (which accounts for the vast majority of instructions under typical loading).
"Is not a sentence" is not a sentence. Well damn.
The sooner AMD goes down the toilet, the sooner someone competent takes over, the sooner Intel and Nvidia get competitive again.
At the moment, consumers and society as a whole are suffering from this beating of a dead horse. We need its death accelerated to the whole industry back on its toes again.
Why do you think AMD going bust would magically mean someone new stepping up to the plate? There are effing ginormous obstacles for a start-up to come and compete in the x86-scene, so much so that it's nearly impossible, and if Intel got a monopoly on the market even for a brief moment the situation would become even worse!
No, the better option is that AMD gets their shit together, never giving Intel full monopoly on the x86-market even for a bit.
I was shopping for VCRs about 20 years back and asked the Future Shop guy how much better it was for having (quoting from the card beside the VCR) a "19 micron tape head". Turns out they ALL had 19 micron tape-heads (whatever the hell that *meant*) as it was the spec for a VCR tape head, at the time, at least. It was just another bit of science-y sounding technobabble to put on the card.
Buying based on core count is like buying for the 19-micron thing; it's either a fast machine for your purposes or not. Absolutely the only way to tell that for sure is a test. The only thing that was ever useful with, say, "megahertz" was that it had for a decade or so there a correlation with the performance you'd get in real use. I've never found "cores" to have anything of the sort.
Concur. The design of the bulldozer modules was abundantly clear. The fact that the "cores" share an FPU was clearly disclosed and part of any diagram of the parts. The shared FPUs played a huge role in assessing bulldozer-based CPUs for high performance computing workloads. For HPC, the usual benchmark is HPL (a.k.a., high performance LINPACK), which is a measure of double precision floating point performance for a particular matrix operation called DGEMM. The fact that the FPU was doing double-duty for two "cores" on a module meant that the peak theoretical performance was limited by the number of FPUs in a CPU, not the number of cores or modules or anything else.
As others have noted, hyperthreading via Intel can have exactly the same impact: the threads share various components, including the FPU.
Another aspect that can have a major impact on performance is the number of memory channels, and how things like cache coherency is handled. Among other things, AMD's hyper-transport exhibits different scalability characteristics depending on the number of sockets. In a four or eight-socket configuration, latency due to cache coherency operations can have a big impact on performance.
- gbn
AMD has always been behind Intel in the performance area for most of its life.
Well, there was the Athlon - era where they were sweeping the floors with Intel; the classic Athlons and Athlon XPs were phenomenal CPUs at the time and highly overclockable. It was glorious, but yeah, I think that was pretty much the only time they beat Intel.
Lets hope AMD gets their shit together. As I've said before. The fat lady hasn't sung yet on AMD but she is warming up in the bullpit. Lets hope that as she waddles up on stage AMD pulls a rabbit out of the hat and she falls off the stage into a tuba.
I really, *REALLY* hope they can manage to do it, but.. I just haven't heard any promising news in that regards anywhere. There's quite literally nothing to indicate that AMD has in any way or form stopped digging even further down the hole they are already in. I do dread the day when Intel becomes the sole x86 - vendor and can practically demand whatever they want, do whatever they want and laugh all the way to the bank.
Err except with HT there is one execution unit, and with AMD there are two.
They are real cores, that share instruction decode and a FPU.
Actual execution is parallel (unlike HT which is more interleaved).
HT isn't similar to Bulldozer modules in any sense.
Mod parent up. It is correct. Some i5's do have 2 cores. The original AC post near the top is also correct, which I notice the mods have knocked to -1 in the usual frenzy of "I disagree, although I am uninformed about the matter."
For a couple decades, "core" meant CPU. 4004, 1802, 8080, 6800, 6809, Z80, 68000... etc.
Then it meant CPU+MMU.
What's funny is that "core" now seems to mean CPU+MMU+FPU (which is great... I love me an FPU per core.) But some (often aimed at mobile or tablets or phones) also have GPUs.
So how long before we won't accept "core" except as a label for CPU+MMU+FPU+GPU?
Then it might be "Nah, that thing has no NPU (neural processing unit), that's not a "real core": CPU+MMU+FPU+GPU+NPU
Then perhaps it'll be "Nah, that thing has no QPU (quantum processing unit)", that's not a "real core": CPU+MMU+ FPU+GPU+NPU+QPU
Then... well, I have no idea. But I do think it'll be something. It's always something. :)
I've fallen off your lawn, and I can't get up.