Intel Will Counter Barcelona with 1600MHz FSB Xeons

[DanTheBanjoman]

After hearing nothing about the 1600MHz FSB Xeons for several months VR-Zone tells us they will appear in Q4 this year. There will be 2 quad core Harpertown models at 2.8 and 3 GHz and a single dual core Wolfdale at 3.4 GHz. Additionally the cache is increased by 50% to 3MB per core. The faster FSB and increased cache will raise the bar even higher for Barcelona, however since both CPU's aren't available it is yet to be seen if this will be enough.

To support these new processors Intel will release the Seaburg chipset, it is not mentioned if the current chipsets will officially support the faster bus.
For a full list of new Wolfdales and Harpertowns visit VR-Zone.

Source: VR-Zone

[kwchang007]

Is the higher fsb meant to help in the server segment?

[mandelore]

hmm, dont it make u think, they need this to counter the barcelona chip?

[kwchang007]

Originally Posted by mandelore

hmm, dont it make u think, they need this to counter the barcelona chip?

Meh, it doesn't really boost performance speed of the CPU unless you're running into a fsb bottleneck. The only time I can think of that happening is in a multi-server cpu socket solution.

[mas0n]

In other news, Intel still uses the old FSB architecture...

[jocksteeluk]

even though this is server hardware i am pretty sick of the ever changing hardware specifications of Intel cpu's because i have no doubt that the 1600mhz bus will filter down to the next revision of Intel cpus making yet another batch of general consumer electric items out of date very soon after release. Upgrade-ability and backward compatibility is certainly something Amd should be praised for.

[mas0n]

I don't really think either of them should be praised for it. If Intel would hurry up and just integrate the Northbridge into the CPU then these guys could just use the same socket again, remember Socket 7?

We should be able to drop any recent processor into any recent board, but these bastards have us ditching whole systems to switch sides.

[tigger]

intel maybe using old FSB architecture,but their still kicking amd's ass.same for the on chip memory controller,i dont think it showed any advantage at all really.

[Sovereign]

Originally Posted by tigger69

intel maybe using old FSB architecture,but their still kicking amd's ass.same for the on chip memory controller,i dont think it showed any advantage at all really.

Thats just stupid. Thats like a kid saying "Hey! My dad is bigger than your dad!" and then runs off to hide behind him.

[Steevo]

Leapfrog.

[FR@NK]

Originally Posted by jocksteeluk

even though this is server hardware i am pretty sick of the ever changing hardware specifications of Intel cpu's because i have no doubt that the 1600mhz bus will filter down to the next revision of Intel cpus making yet another batch of general consumer electric items out of date very soon after release. Upgrade-ability and backward compatibility is certainly something Amd should be praised for.

I really dont see why this would be a problem since Intel's current mid range chipset(P35) can easily do 2000Mhz Bus.



The X38 would most likely hit 2400Mhz easy. The issue I can see is since the bus will be rated higher on these chips, the multi would be set lower which would limit the max overclocking compared to a 1333 or 1066Mhz chip.

[Grings]

Originally Posted by tigger69

intel maybe using old FSB architecture,but their still kicking amd's ass.same for the on chip memory controller,i dont think it showed any advantage at all really.

it was a disadvantage in the end, as they had to change from 939 to am2, when they had no real improvement other than ddr2 support to warrant a socket change (and lost me as a customer in the process)

[DanTheBanjoman]

Originally Posted by FR@NK

I really dont see why this would be a problem since Intel's current mid range chipset(P35) can easily do 2000Mhz Bus.



The X38 would most likely hit 2400Mhz easy. The issue I can see is since the bus will be rated higher on these chips, the multi would be set lower which would limit the max overclocking compared to a 1333 or 1066Mhz chip.

Current generation is Intel 5000 series, not P35 or X38. Also, with 2 sockets (and buses) the strain on the northbridge is higher. Besides, 1600MHz settings do exist in current chipsets, the BIOSes simply don't support it. With the official release the code will appear and might be injected into the current BIOSes like they did with the PC-DL.

Originally Posted by mas0n

In other news, Intel still uses the old FSB architecture...

This is no news, the roadmap clearly shows that Tigerton will use CSI. CSI will come to lower end platforms as well. Besides,these platforms have a FSB per socket, so it isn't as crippled as Tulsa for example is.

[Wile E]

Originally Posted by Grings

it was a disadvantage in the end, as they had to change from 939 to am2, when they had no real improvement other than ddr2 support to warrant a socket change (and lost me as a customer in the process)

Intel is no better. Look how many people had to buy a new board for Core 2, despite having a socket 775 board already.

[WarEagleAU]

Agreed Wil E. Also....the integrated memory controller is still a good thing. The DDR2 support wasnt able to be integrated into socket 939s controller from what was stated. Youd have to redevelop the chipset and that would just lead to confusion in the 939 world. It isnt that bad really going from 939 to AM2 (940)

[msimax]

the imc doesnt suck that bad

http://www.xtremesystems.org/forums/...&postcount=136

[KennyT772]

holy f***ing s***. I thought I was doing good at 8000mb/s