Further Camino delay will mean Intel compromises
Chipzilla may have to light other backburners
As reported here yesterday, the Camino (i820) chipset is now unlikely to arrive until the end of September at the earliest. (Story: Rambus Meltdown a sorry tale of fudge, mudge and kludge) And that is likely to mean more compromises to ensure that Direct Rambus RIMMs, processor and board all work together. Our inside information now is that Intel's PC partners -- in particular IBM, Compaq and Dell -- are very unhappy about the delays to the technology after a meeting in California last week. If the technical problems continue, we understand that Intel may drop back to AGP 2x, and major PC OEMs do not wish to ship 300MHz Rambus RIMMs, outperformed as they are by even 100MHz SDRAMs. The 133MHz FSB is also threatened. Our source said: "Politically, IBM have to deny that they're unhappy [with Rambus]. Intel, whether it likes it or not, will probably have to endure PC-133." As we reported from Computex, Intel is attempting to put pressure on Via to delay its PC-133 offerings until the chip behemoth gets a chance to put its marchitecture together. The world+dog is unlikely to see Rambus RIMMs at speeds of over 300MHz when Camino does have its bugs sorted out, he added. One of the problems with Direct Rambus is "cache trashing", described by Mosel Vitelic at a press conference in Taiwan last week. ® See also Computex 99 coverage Many stories relating to PC-133 and Direct Rambus, including admissions that the Ram bus is slowed by speed and other considerations Pentium 550MHz is an emergency Plan A didn't work, so Plan B's kicked in...
Sponsored: Benefits from the lessons learned in HPC