Mitas Electronics

Mitas Electronics High Speed Innovation for High Volume

THE POINT IS NOT THAT CPO FAILSThe point is that the boundary moves.This week, we examined a more precise description of...
10/17/2026

THE POINT IS NOT THAT CPO FAILS

The point is that the boundary moves.

This week, we examined a more precise description of co-packaged optics.

CPO creates genuine architectural value by moving the optical engine closer to the host ASIC. Substantial electrical distance can disappear. Loss and impedance discontinuities can be minimized. The dependence on signal conditioning may be reduced.

But a shorter electrical path remains.

That is not a criticism of CPO.

It is part of the architecture CPO creates.

When conversion moves closer to silicon, the surrounding electrical, optical, thermal, packaging, attachment, test and manufacturing relationships move with it.

The question is therefore not whether electrical or optical transport wins.

It is:

Where should electrical transport end?

Where should optical transport begin?

And what changes when that boundary moves?

Next week, we will examine where the boundary has moved—and what now exists within the shorter, more integrated path surrounding it.

Follow the boundary.

Mitas Electronics
Engineering the future of high-speed data transport.

CPO MOVES THE BOUNDARY
What must the system preserve across it?

CPO DOES NOT ELIMINATE THE ELECTRICAL PATHThis week, we examined what co-packaged optics removes—and what it retains.CPO...
10/16/2026

CPO DOES NOT ELIMINATE THE ELECTRICAL PATH

This week, we examined what co-packaged optics removes—and what it retains.

CPO creates real architectural advantage by moving the optical engine closer to the host ASIC. Substantial host-board reach can be reduced. Electrical connector and module-board sections may be removed. Loss and impedance discontinuities can be minimized.

But the electrical path does not become zero.

The host ASIC must still communicate electrically with the drivers, modulators, photodetectors and receiver circuitry associated with the optical engine.

The path becomes shorter.

The integration becomes tighter.

And the electro-optical boundary moves inside the package architecture.

Our new anchor article examines the conventional pluggable path, the electrical regions CPO removes or compresses, the channel it retains and the broader system conditions surrounding the point of conversion.

The central question is not whether optics works.

It is:

What must the complete system preserve as information approaches, crosses and leaves the point of conversion?

Conversion occurs at a boundary.

The path remains a system.

Read the full article: https://www.mitaselectronics.com/cpo-does-not-eliminate-the-electrical-path/

Mitas Electronics
Engineering the future of high-speed data transport.

CPO MOVES THE BOUNDARY
What must the system preserve across it?

A 50 MM PATH IS STILL A PATHCPO does not eliminate the electrical path. It shortens and relocates it by moving the optic...
10/15/2026

A 50 MM PATH IS STILL A PATH

CPO does not eliminate the electrical path. It shortens and relocates it by moving the optical engine closer to the host ASIC and integrating the connection within the package.

In its Co-Packaging Framework Document, the OIF describes an illustrative CPO operating point with approximately 50 mm of first-level packaging substrate and approximately 10 dB of loss at Nyquist, including package parasitics from the ASIC and optical-engine assembly.

That is a defined electrical channel. It has physical length, reference planes, parasitics and finite loss that must be designed, measured and budgeted.

CPO changes where the electrical path exists, not whether it exists. As the electrical-to-optical boundary moves into the package, signal integrity, loss margin and measurement discipline remain essential across the channel.

The path is shorter.

The need for disciplined engineering remains.

Mitas Electronics
Engineering the future of high-speed data transport.

CPO SHORTENS THE PATH
It does not eliminate the need to engineer signal integrity, margin and measurement across the boundary.

WHAT CPO REMOVES—AND WHAT IT RETAINSA conventional pluggable optical path contains a long sequence of electrical regions...
10/14/2026

WHAT CPO REMOVES—AND WHAT IT RETAINS

A conventional pluggable optical path contains a long sequence of electrical regions:

ASIC package.
Host board.
Layer transitions.
Connector.
Module board.
Electronic conversion circuitry.

Co-packaged optics changes that architecture.

By moving the optical engine closer to the host ASIC, CPO can remove or compress much of the board reach, connector path and module-board distance that electrical signals previously had to cross.

But several electrical regions remain:

The ASIC’s electrical I/O.
The die-to-package transition.
Substrate or interposer routing.
The optical-engine attachment.
The electronic interface to the photonic devices.

This is not evidence of a weakness in CPO.

It is a more precise description of how CPO creates its advantage.

Substantial electrical distance disappears.

A shorter, more integrated electrical path remains.

And that remaining path ends at a boundary where electrical information becomes optical information.

What remains is not an abstract electrical residue.

It is a real, engineered part of the architecture.

Mitas Electronics
Engineering the future of high-speed data transport.

CPO MOVES THE BOUNDARY
What must the system preserve across it?

CPO’S ADVANTAGE BEGINS BY SHORTENING THE ELECTRICAL PATHNot by pretending the electrical path never existed.Co-packaged ...
10/13/2026

CPO’S ADVANTAGE BEGINS BY SHORTENING THE ELECTRICAL PATH

Not by pretending the electrical path never existed.

Co-packaged optics moves the optical engine closer to the host ASIC.

That shorter reach can reduce electrical-channel loss, reflections and the signal conditioning required before data becomes optical.

The OIF describes this rationale precisely: proximity allows electrical-channel losses and impedance discontinuities to be minimized.

Minimized—not eliminated.

The ASIC still communicates electrically with the drivers, modulators, photodetectors and receiver circuitry surrounding the optical engine.

CPO makes that path shorter and more integrated.

That is not a failure of CPO.

It is part of how CPO creates its architectural advantage.

But it also leads to a more precise question:

If the electrical path remains, where should it end—and what must the system preserve before data crosses into the optical domain?

Mitas Electronics
Engineering the future of high-speed data transport.

CPO MOVES THE BOUNDARY
What must the system preserve across it

WHAT IF THE ARCHITECTURE IS WAITING FOR A DIFFERENT PATH?Many boundaries in computing systems appear fixed.Memory sits w...
10/10/2026

WHAT IF THE ARCHITECTURE IS WAITING FOR A DIFFERENT PATH?

Many boundaries in computing systems appear fixed.

Memory sits where the physical path can reach it.

Compute is organized around the latency and bandwidth that path can deliver.

Packaging accommodates the routing, power, thermal and signal-integrity requirements needed to connect them.

The boundary between electrical and optical transport is placed where the surrounding technologies can support it.

These arrangements can begin to look like inherent requirements of the architecture.

But some may be accommodations to the way data must currently move.

That distinction matters.

The memory wall is not only a question of memory capacity or device bandwidth.

Bandwidth must cross a physical path.

That path imposes conditions on the components and architectures on both sides.

Changing the path could change those conditions.

But a different path creates architectural value only if the complete system retains the benefit.

The information must survive.

The path must remain physically viable.

The handoff must preserve the advantage that justified changing physical regimes.

And the improvement must not simply reappear elsewhere as additional area, latency, power, thermal load, control or validation.

If those requirements can be satisfied while asking less from the surrounding system, the architecture may gain choices it does not have today.

Memory and compute could have different placement options.

Computation could occur closer to the information it requires.

Functions could be partitioned differently.

Physical boundaries could move.

New combinations of reach, density and topology could enter the feasible design space.

None of these outcomes should be assumed.

They must be demonstrated through evidence that extends from physical behavior to architectural consequence.

But they reveal why transport deserves to be considered before architectural boundaries are accepted as permanent.

Mitas is examining where electrical and optical physics naturally belong as computing systems scale—and how changes in the transport path could alter the architectural assumptions those systems must carry.

Because the most important result may not be a better connection.

It may be an architecture no longer required to organize itself in the same way.

What if the architecture is waiting for a different path?

Mitas Electronics
Engineering the future of high-speed data transport.

THE VALUE OF A NEW PATH IS THE ARCHITECTURE IT MAKES POSSIBLEA transport path can become faster, longer, denser or more ...
10/09/2026

THE VALUE OF A NEW PATH IS THE ARCHITECTURE IT MAKES POSSIBLE

A transport path can become faster, longer, denser or more energy efficient.

But a better path is not automatically a better architecture.

If the improvement requires additional area, latency, power, thermal capacity, control, packaging or validation, the surrounding system may absorb the advantage.

The more consequential question is:

What can the architecture now do that it could not do before?

Can memory and compute be placed differently?

Can computation occur closer to the information it requires?

Can unnecessary data movement be reduced?

Can functions be partitioned differently across packages or modules?

Can the boundary between electrical and optical transport move without recreating the original constraint elsewhere?

These are the outcomes that reveal whether a new path has expanded the feasible design space.

The appropriate transport boundary will not be the same for every architecture. It depends upon the workload, physical scale, bandwidth density, latency, energy, thermal environment, manufacturability and operating requirements of the complete system.

That is why transport cannot be evaluated independently of the architecture and components it connects.

In our latest Mitas Intelligence article, we examine why the value of a new path must ultimately be measured by the architectural choices it makes possible.

Read the full article: https://www.mitaselectronics.com/the-value-of-a-new-path-is-the-architecture-it-makes-possible/

Mitas Electronics
Engineering the future of high-speed data transport.

WHEN TRANSPORT REQUIRES LESS, SYSTEM BOUNDARIES CAN MOVESystem boundaries often appear fixed because the technologies ar...
10/08/2026

WHEN TRANSPORT REQUIRES LESS, SYSTEM BOUNDARIES CAN MOVE

System boundaries often appear fixed because the technologies around them have evolved together.

Where memory is placed.

Where compute is placed.

Which functions remain within a package or module.

Where the signal changes physical form.

How far information must travel between those points.

These are architectural decisions.

But they are also responses to what the transport path can physically support.

If the path requires a particular reach, geometry, power budget, thermal environment or signal margin, the system organizes itself around those conditions.

When those requirements change, the feasible design space can change with them.

Memory and compute may have more placement options.

Packaging functions may be partitioned differently.

The point at which one physical transport regime hands off to another may be reconsidered.

Routing, power delivery and thermal capacity may be allocated differently across the system.

None of this means that moving a boundary eliminates its cost.

A boundary moved elsewhere can simply relocate latency, power, thermal load, complexity or validation burden.

The complete architecture must still satisfy the full requirement set.

But if a changed path preserves continuity while requiring less from the surrounding system, boundaries that once appeared fixed may become architectural variables.

That is the more consequential possibility.

The value of a new path may not be limited to how well it transports data.

Its greater value may be the architecture it makes possible.

Mitas Electronics
Engineering the future of high-speed data transport.

A NEW PATH SHOULD EXPAND THE FEASIBLE DESIGN SPACEArchitectural freedom is not the number of configurations that can be ...
10/07/2026

A NEW PATH SHOULD EXPAND THE FEASIBLE DESIGN SPACE

Architectural freedom is not the number of configurations that can be imagined.
It is the number that can actually work.
A system configuration becomes feasible only when its requirements can be satisfied simultaneously.
Bandwidth.
Latency.
Reach.
Power.
Thermal behavior.
Signal margin.
Physical density.
Reliability and manufacturability.
An improvement in one transport metric may move one boundary while tightening another.
Greater reach may be less useful if it requires disproportionate power.
Lower loss may not create architectural value if the complete path consumes additional area or thermal capacity.
Successful conversion may not expand the design space if the handoff introduces new latency, control or validation burdens.
The relevant question is therefore not simply whether the path performs better.
It is whether previously excluded combinations of placement, distance, density and topology become technically viable.
That is how architectural freedom can be evaluated.
A new path expands the feasible design space when it gives system architects more configurations capable of meeting the complete requirement set—not merely more theoretical options.
This creates a demanding but useful standard:
Which architectural choices become viable that were not viable before?
And what happens when those new choices allow the system’s existing boundaries to move?

Mitas Electronics
Engineering the future of high-speed data transport.

WHAT COULD THE ARCHITECTURE BECOME FREE TO DO?Engineering decisions are often made within established constraints.Compon...
10/06/2026

WHAT COULD THE ARCHITECTURE BECOME FREE TO DO?

Engineering decisions are often made within established constraints.

Components are placed where the path can reach them.

Routing is organized around available area and signal margin.

Power and thermal budgets are allocated partly to support data movement.

System topology adapts to what the physical interface can reliably deliver.

Over time, these accommodations can begin to look like fixed architectural requirements.

But they may not be fixed.

Some may be consequences of what the transport path currently demands.

If a future handoff could preserve information, physical viability and architectural value while requiring less from the surrounding system, resources could return to the architecture.

Less area consumed by transport could create placement freedom.

Lower power and thermal burden could preserve capacity for memory or compute.

Greater practical reach could permit different forms of separation and system organization.

Reduced routing, control or validation burden could expand the range of viable topologies.

None of these outcomes should be assumed.

But they reveal the more important question.

Which decisions in today’s systems are genuine requirements of the workload?

And which are architectural accommodations to the way data must currently move?

The value of requiring less from the architecture is not merely efficiency.

It is the possibility of giving the architecture choices it does not have today.

How should that freedom be measured?

Mitas Electronics
Engineering the future of high-speed data transport.

Address

1750 E Red Bud Lane, Round Rock
Austin, TX
78664

Opening Hours

Monday 8am - 6pm
Tuesday 8am - 6pm
Wednesday 8am - 6pm
Thursday 8am - 6pm
Friday 8am - 6pm

Website

Alerts

Be the first to know and let us send you an email when Mitas Electronics posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Share

Category