Replies to post #6205 on Rambus (RMBS)

[smd1234]

infringeon - round 2:
The Verdict on HPQ and Grover Patents
by: infringeon2003 11/18/05 11:27 am
Msg: 826463 of 826741

The verdict on the HPQ patent has already been determined by the US Patent Office.

The Verdict on the Grover patent is that RAMBUS does not infringe. The Grover patent is much more complex than the RAMBUS patent.
Grover is a "closed loop" control timing control system while Rambus is an "open loop" timing control system.

Grover was intended to be used in telecommunication applications where the clock signal is a radio wave or someother part of the electromagnetic spectrum not an electricla signal on a copper trace. Grover would be useful for synchronizing groundsation to satellites. RAMBUS would be useless for that purpose.

-----------------------------
Is that why Grover was not cited as prior art?
-----------------------------
Re: The Verdict on HPQ and Grover Patent
by: infringeon2003 11/18/05 11:51 am
Msg: 826497 of 826741

Probably!

Also the '703 patent issued before the Grover patent which was tied up in intereference proceddings with the Hewlett Packard patent.

The Grover patentwas addressing mainly the "telecommunications" arena not the memory system design area. So they were in different patent categories.

-----------------------------------
If one diagrammed the two leaving off the feedback loop and not mentioning the physical tuning of the traces, the diagrams might appear very similar? Will Payne understand this?
-----------------------------------
Re: The Verdict on HPQ and Grover Patent
by: infringeon2003 11/18/05 11:44 am
Msg: 826490 of 826741

I don't think that Payne is into the "understanding" thing.

The clock circuit diagram for Rambus does not show that it attaches back to the source of the master clock. It is an "Open Loop" control system.

The Grover Patent digram shows that the return clock signal attaches back to the source of the master clock.

This is a HUGE DIFFERENCE from an engineering and manufacturing point of view!

Re: The Verdict on HPQ and Grover Patent
by: infringeon2003 11/18/05 11:46 am
Msg: 826492 of 826741

The Rambus "open loop" control system design
leads to the use of "latency registers" on the memory components attached to the RAMBUS bus.

[calbiker]

Re. the clock apparatus operates "open loop".

Infringeon isn't reading the patent correctly. Makes you wonder how on earth he did reporting the trial. ;-)

The free press will explain. Ha.

Let's look at the Grover patent:

A clock distribution apparatus for use in providing a common absolute time reference to spatially distributed application modules requiring synchronized clocks, said apparatus comprising:

means for transmitting an outgoing reference signal from a first site to a second site and a return reference signal from said second site to said first site upon arrival of said outgoing reference signal at said second site; and

For simplicity, let's call the devices that require synchronous operation DRAM.

Infringeon confuses the 'first site' with the oscillator or the clock origin. It's not. The clock signal travels down the transmission line and after a propagation delay arrives at the 'first site', or in this example DRAM #1. The clock signal continues onward to the 'second site' or DRAM#2. Each site is a DRAM. The clock never gets feedback to the clock generator. This is also quite obvious from Fig. 1. The schematic shows that the return clock is just terminated. Also, there was never any discussion within the patent that the return clock is used in any control of the oscillator. It's just wishful thinking on behalf of infringeon.

For that reason, his conclusions, based on incorrectly reading the patent are WRONG.

More choice comments from infringeon:

Grover was intended to be used in telecommunication applications where the clock signal is a radio wave or someother part of the electromagnetic spectrum not an electricla signal on a copper trace. Grover would be useful for synchronizing groundsation to satellites. RAMBUS would be useless for that purpose.

This is just nonscense! He's talking out of his A....

From the patent:

In the telecommunications industry and the computer industry, leading manufacturers are continuing to develop equipment designs employing increased clock rates while simultaneously pursuing architectures with hundreds and even thousands of distributed machine elements. These elements may be, for example, periphery interface modules, time-switch modules in a digital switching machine, individual processor elements in a "connectionist" type machine, pipeline floating point logic units in an array processing supercomputer, or systolic array processors for signal processing or radar applications.

For maximum performance and efficiency in these applications, each equipment module requires a clock signal that is phase-synchronous with the clock in every other module. The clocking of every clement of the machine synchronously at the highest possible rate provides the best chance of approaching system speeds equal to the switching speeds of individual logic elements of the technology employed.

Ya gotta admit, Infringeon is a great story teller. Ha.
I wouldn't base my investments on his inputs. Hope cal wasn't too rough on the guy. ;-)