Slashdot Mirror
AMD's Athlon XP 2700+
kraven_73 writes "According to some Taiwanese sources, AMD will officially reveal its Athlon XP 2700+ processor on the 7th of October. Most interesting is that this CPU will have a 333 MHz FSB. The first implementation of this increased FSB on Athlon platform. It is expected that the novelty will be based on the latest Thoroughbred core stepping 1, just like the current Athlon XP 2400+ and 2600+, and will work at 2.17GHz."
Right now the AthlonXP has a FSB of 133DDR or an effective 266. Increasing this to 166DDR or 333 gives an additional 25% bandwidth. This means that the IS a point to putting DDR333 ram in an athlon MB and seeing a real performance advantage. I personaly would like to see them skip 333 altogether and go to 400. This would bring them up to one generation behind the P4 in terms of FSB bandwidth and would even out alot of the test scores that ppl are complaining about right now.
In case you have been asleep for the last year the FSB is the AthlonXP's largest bottleneck!
As for overclocking: Remember when the 266 FSB came out and ppl were complaining about the low overclock potential on the new boards? well that will happen again, but the second generation of boards will ROCK for overclocking. I have my money on boards that will handle a 400 FSB within 6 months of volume market penetration for the 333FSB.
<This .sig left intentionally blank>
I'm not normal, but here's a few from my background
Video editing. Nothing out there is remotely fast enough for what I want to do, and what I want to do is pretty limited.
Computatational chemistry. Nothing out there (or scheduled for the next ~100 years) is fast enough to do the simulations people are really interested in.
License key cracking for those companies who decide to use encryption. :^)
"Seven Deadly Sins? I thought it was a to-do list!"
I'm sorry, but Intel does NOT use a 533Mhz FSB speed any more than AMD use a 333Mhz one. The "533" refers to 4x133Mhz (it's a 133Mhz bus with QDR tech) whereas the "333" refers to 2x167Mhz (it's a 167Mhz bus with DDR tech). Incidentally, I think that Apple is the first company (unbelieveably) to have implememted a 167Mhz FSB in their new "DDR" G4 designs. Shame the G4 chip isn't up to using the (otherwise fast) bus in DDR mode. Oh well, roll on the fabled MPC 7470!
That was classic intercourse!
I know you are going to get slammed by the 'Bill Gates said 640k was enough' crowd so I thought you would like to know that Doom 3 isn't going to require a Freon cooled GeForce 3000 XP+ SupraGamer to be playable ...
Doom III playable on current hardware says John Carmack in Interview with GameSpy
-DR
Depends what you define as a northbridge, and what you define as a FSB. The bus type (EV6, HyperTransport, whatever) is just a name for the signaling and protocol - the name of the bus itself can still be "Front Side Bus".
The "traditional" northbridge had a memory controller and an AGP controller, as well as a PCI controller. The PCI controller got moved completely off the North Bridge to the South Bridge and replaced with a proprietary interconnect in a lot of modern chips. The memory controller was moved on die, but the AGP controller is still off-die, and thus needs a chip for it. This chip could be called the "north bridge". It's just a name - AMD calls it the "HyperTransport AGP 3.0 Graphics Tunnel" (which doesn't really make much sense, as it also has a HyperTransport link to a south bridge - how does THAT relate to graphics?) but it's still a North Bridge, just without the memory controller.
There are two HT links on the system, which is why it makes sense to call it a "north bridge" and a "south bridge": there's a HT link from the CPU to the North Bridge (the AMD 8151) and a HT link from the North Bridge to the South Bridge (the AMD 8111).
So, yes, they do have a FSB, unless you want to call it something else: "highspeed HT link" and "lowspeed HT link" (for the North Bridge-South Bridge interconnect) maybe? Got me. It doesn't matter. The FSB has always been the high speed link out of the processor to a bridge chip, which then has a low speed link to another bridge chip which has all the PCI, LPC, ethernet, all that crap. Hammer doesn't change that, it just removes the memory controller from the North Bridge.
You're thinking too much along the lines of a home-user running Windows. There's a very real presence of Intel (and unfortunately, not so much AMD) CPUs in low cost workstations for business use. Linux is certainly part of the picture as well, as almost all EDA vendors have or are releasing Linux versions of their tools.
What do I use my 80x86 cpu for? Well, I work in a hardware engineering group which does ASIC and FPGA design. We have a CPU farm of about 30 machines with Intel P4s running RedHat 7.2. (May see AMD Hammer chips in the future - we are excited about this possibility). We run everything from RTL and gate level VHDL and Verilog simulations, to chip synthesis, to test insertion and fault grading simulations. One of the last chips I worked on required such a large set of ATPG vectors (and the design was just so huge), that it required breaking the test vectors into ten groups, and even then, just one file (10% of the total) required an 8GB Sun box to convert the vectors to the fabs tester format, and the gate-level simulation took 10 days. PER FILE. Yes, that was total of 100 CPU days of simulation time for one chip just for ATPG vectors. And these were running on 1.8GHz Pentium 4s with 3GB of RAM. Not surprisingly, leading edge tools in this field are starting to look at distributed simulation over high-speed backplanes (read: not ethernet).
Tomorrow's technology is designed and verified on today's hardware. Every generational step in every sector of industry leapfrogs like this. You can't design next year's high performance video card using 80286s. Definitely for ASIC/FPGA design, there isn't a system fast enough for how quickly we (the engineers) would like a simulation to finish in. Being able to run more simulations overall means a better design out the door. More stuff caught up front. The faster a simulation runs, the quicker it will finish, meaning we can get by with fewer high-priced licenses for our EDA tools. (Licenses are usually in the tens or hundreds of thousands of dollars EACH).
You can never run enough tests before a product is done. How many tests/simulations are run, depends on how long they take to run. Give me the fastest CPU you got, decked out with the most physical RAM it can handle. (Sadly, the 32bit limit on current 80x86 platforms is hurting us badly - go x86-64! go AMD! Capture the workstation market!)
You said it, you're an EE. Most of AMD's customers aren't and they need to sell chips. Like it or not Intel deliberately created a pipeline with too many stages so they could clock their chip high. People care about performance, AMD is telling people their chip performs like an Intel or better at some clock rate. Otherwise their customers wouldn't get it because they aren't EEs. You clearly know the real clock so there's no problem. Get over it, if AMD marketed like you suggest they might be out of business already.
Processor myth #1: AMD chips generate more heat then Intel chips.
Check the data sheets sometime, the amount of heat that an AthlonXP and a P4 put out is nearly identical. Both are also only in the 60-70W range, or about the same as your typical light bulb. We all know how turning on a single light in your house cause your AC bills to skyrocket!!!