Slide 1:
Transceiver Specifications
Vince Melendy
Application Engineer
Methode Electronics
Office: 508-281-3216
FAX: 617-531-2017
Email: 72540.1774@compuserve.com

Discussion:
I am going to raise questions regarding the proposed specifications. 

Slide 2:

Goals of Task Force
• Products to meet Operating Distance Goals
• Use Fibre Channel Physical Layer
- Common to Fibre Channel, ATM, 
and HIPPI Serial
• Use Short and Long Wavelength Options

	Low Cost Solution That Meet Above Specification Comparison

Discussion:
I believe that these are the goals of the Task Force. Also the use of common optics to the other major specifications will produce the lowest cost solution for Gigabit Ethernet and not create a Gigabit specific product.

Slide 3:

• Long Wavelength Single Mode Specification DECREASED 
Power budget
• Short Wavelength Multimode Specification INCREASED Power 
budget 


	 Inconsistent  --  WHY

Discussion:
(1) In reviewing the proposed specifications for the long wavelength and short wavelength I see that the long wavelength single mode specification decreased the power budget, where as the short wavelength product had its power budget increased. In my view why did the single mode budget get reduced, is not this one of the important specifications in determining link length. Why reduce the budget if you want greater distances.

Methode in the lab today utilizing the current Fibre Channel specification can operate at 30 km with ease on single mode fiber with the current Fibre Channel 2 km specification. With this data then there will be no problem operating at the 3 km distance.

(2) The Short wavelength specification the power budget was increased to extend the link length. I believe that other factors namely the fiber characteristics of dispersions dictate the link length not increasing the power budget.

In the lab Methode has taken data that exceed the 500 meter goal.

Slide 4:

Specification Changes
• Long Wavelength Changes
- Increased Spectrum	
- Launch Power Decreased
- Power Budget Decreased
- Added Rise and Fall Times
* Others unchanged

Slide 5:

Specification Changes
• Short Wavelength Changes
- RIN Changed					
- Receiver Sensitivity Increased
- Power Budget Increased
- Added Rise and Fall Times		
* Others unchanged

Slide 6:

Proposed Long Wavelength Specifications vs Fibre Channel Specifications

Slide 7:

Proposed Short Wavelength Specifications vs Fibre Channel Specifications 

Slide 8:

Proposal for Long Wavelength
• Go back to the standard 2 Km version of the Fibre Channel 
Specification


Results:	 LOWER COST through higher volumes manufactured. 
		NOT a Gigabit Ethernet Specific Product

Discussion:
I believe we should go back to the standard 2 km version of the Fibre Channel specification. As I stated previously Methode's data shows this specification already allows 30 km link lengths. Why should the specification change? If the specification changes this will create a higher cost Gigabit Ethernet Specific Product not a common product at lower cost.

Slide 9:

Proposal 1 for Short Wavelength
• Go back to the standard version of the Fibre Channel 
Specification 


Results:	 LOWER COST through higher volumes manufactured. 
 NOT a Gigabit Ethernet Specific Product

Discussion:
For the short wavelength specification I believe we should go back to the standard short wavelength version of the Fibre Channel specification. As I stated previously Methode's data shows this specification already allows link lengths exceeding the 500 meters in the lab. Why should the specification change? This will give the lower cost product and not create a higher cost Gigabit Ethernet Specific Product. Why should we cast aside the work the Fibre Channel Committee has done and reinvent a new Gigabit Specific optic module specification.

Slide 10:

Proposal 2 for Short Wavelength
• Launch Power
- Change to -4 to -8 dBm
• Receiver Sensitivity
- Decrease to -14 dBm


Result:	 LOWER COST

Discussion:
As I proposed at the October meeting if specification changes are made as suggested here the lowest cost product could be delivered for Gigabit Ethernet. This specification would maintain the 6 dB power budget which is more than adequate for the link lengths desired. This would then give you a product derived from the Fibre Channel product.

The data in the next slide shows that our product can maintain the above specification.

Slide 11:

Methode Launch Power Graph

Slide 12:

Conclusion
• Let's be Consistent
• Let's not create Gigabit Ethernet Specific Products
• Let's use Fibre Channel Specifications


Low Cost Solutions

Discussion:
In conclusion, we should be consistent in writing specifications. 

We should not create Gigabit Ethernet Specific Products. We should not allow a higher cost product be specified than is really necessary. If the specifications remain as they are then the product for Gigabit Ethernet will be a premium product over the Fibre Channel Product.

Let's go back the Fibre Channel Standards for the low cost solutions. Let's not throw out all the work the Fibre Channel Committee has done in the past.