I'd talk to John Larkin at Highland Technologies before you try to develop your own photo diode interface to anything. Perhaps Dr. Phil Hobbs at ElectroOptical Consulting. You can spend three months verifying photodiode designs or you can just buy the design expertise or devices.
The difficult part is getting the Transimpedance amplifier and Comparator right. You have to get from the photodiode signal to ECL or PECL logic levels. Then from ECL/PECL to what ever the chip needs. This requires some form of Transimpedance converter,
and one of very small physical size and high speed, because of propagation delays. The reason for this is to select a constant point on the edges of the photodiode signal to trigger from.
If you have a reasonable signal repetition rate of tens of Hertz or more you can use a CFD to find the pulse edge and condition the signal.
If your application is single shot you need either modules made for CAMAC crates, or some other specialized signal processing hardware.
Re-inventing the wheel is not advised in this area, you are dealing with signals of Gigahertz Bandwidth, and that requires "Controlled Impedance" circuit design techniques. A small error in circuit board design can set you back months at this pulse rise time.
This is not expensive if you find an expert. it is widely available for high speed fiber optics techniques such as SONET.
Even if you cannot afford their services, Mr. Larkin or Dr. Hobbs are very likely to point you in the right direction.
http://www.highlandtechnology.com/
http://www.electrooptical.net
www.ThinkSRS.com Stanford Research Systems also works in this area.
Steve