Update on the thermal putty on Gigabyte graphics cards after further storage: Oil deposition on an alarming scale and silence as a strategy

The following photo documents the condition of the already familiar Gigabyte dispenser putties after a total of six weeks of storage on plain copy paper, laid out flat and in a completely horizontal position. My first test already clearly showed that the material marketed as “Server-Grade Thermal Conductive Gel” tends to migrate binder components under static load at room temperature – an effect that is referred to as bleeding in the specialist literature and indicates insufficient retention of low-molecular silicone oils in the polymer matrix.

What is now visible is a significant intensification of this phenomenon: a large, dark edge has formed around the amorphously deformed mass of the putty, documenting the absorption of volatile components by the paper. This oil-like zone is much more pronounced than in the first experiment. The spread occurs radially in a uniform manner and proves that a continuous leakage of unbound components continues to occur – and this under completely motionless laboratory conditions, without any external thermal or mechanical influence.

The rectangular sample below – a pre-formed semi-putty pad with high internal coherence – shows no separation, no material migration and no recognizable traces of oil film. This comparison once again impressively confirms that the Gigabyte material is a formulation designed purely for dispensability and not for long-term stability, as the diameter of the oil spot has increased dramatically. That’s just not possible, sorry.

Review of the original findings

I had already pointed out the following critical points in the first analysis:

1. Over-optimized viscosity: the putty was obviously formulated to remain as easy as possible to extrude through dispenser nozzles. For this purpose, highly volatile silicone oils are usually added to improve the shear behavior – but at the expense of structural integrity.

2. Weak adhesion and low matrix cohesion: Microscopic analysis showed a lack of wetting of metallic surfaces and insufficient bonding between the carrier matrix and filler.

3. Porosity and trapped air pockets: These favor sedimentation, delamination and increase oil separation under static or thermal load.

4. Unsuitable for vertical installation: Even a moderate inclination can lead to creep processes in which the material is slowly displaced from the contact zone by its own weight.

These theoretical weaknesses now manifest themselves once again in the form of a bleeding effect that can no longer be overlooked in practice (see photo).

Consequences and evaluation

The observed segregation is more than just a cosmetic problem. In real applications, such as on GPU VRM zones or between memory modules, this means

• Reduction of thermal coupling, as the oil reduces the effective contact area.

• Uncontrollable local heat distribution, which favors hotspots.

• Long-term material migration, which can lead to partial delamination or even short circuits if the oil migrates onto conductor tracks.

• Maintenance problem: The reconditioning of such systems is complex, as the material cannot be removed without leaving residues and binding agents remain in pores and on surfaces.

For a manufacturer that uses this material across the board in series production, this is simply a thermal risk with notice. In retrospect, the fact that Gigabyte nevertheless opted for such a solution appears to be not only negligent, but also a strategic mistake. The short-term advantage in the production process comes at the price of a structural problem that has a long-term impact on product reliability and ultimately also on trust in the brand.

Outlook and further development

I will continue to observe and document this process, as the visible oil separation provides a realistic picture of the development of degradation over time. It remains to be seen to what extent this effect escalates under thermal cycling or mechanical stress when installed, for example in compact ITX cases or vertically mounted systems, which are increasingly common in the end customer sector. Gigabyte has not done itself any favors with this use of materials, neither technically nor in terms of public perception. Anyone operating with the term “server-grade” should also deliver server-grade long-term stability. At the moment, however, the putty primarily delivers one thing: a textbook example of material failure due to misunderstood priorities in production.

In addition to the technical aspects of the thermal putties under investigation, a structural question is increasingly arising: Why has Gigabyte not contacted me at all to this day, even though I communicated my objections and initial findings early on, and did so objectively, verifiably and without polemical escalation? It would have been easy to respond to my comments, ask questions or at least seek an exchange. Instead, there is consistent silence, a practice that has unfortunately not been seen for the first time. It is understandable that a large company cannot or does not want to respond to every critic. However, in this case it is not a matter of vague assumptions or anecdotal user reports, but of laboratory-based material analyses with ASTM-compliant measuring procedures, documented microscopy and chemical traceability via LIBS. Anyone who simply ignores this type of criticism is also not taking the risks for their own customers seriously.

Perhaps my type of reporting does not fit in with the soft-focus press work that is apparently preferred in Gigabyte’s local PR department. I think that even at headquarters they prefer to trust the statements of internal technologists or sales partners of the expensively purchased machines, i.e. people whose job it is to sell or represent processes, not to scrutinize them. In such a felt structure, it is obviously easier to marginalize independent voices than to seriously engage with technical criticism.

The fact that I spoke out publicly about flaring power supply units at the time and hit a sore spot is still held against me today. But instead of engaging in an unemotional dialog, which I had initiated on my own initiative, I was blacklisted at the time as a supposedly uncomfortable journalist. This is not only short-sighted, but ultimately self-damaging. Because: I am not the problem. On the contrary: I am happy to share my findings, transparently, based on sources and with the aim of improving product quality. The only thing is that you have to ask. I don’t bite. I measure, document and publish. And anyone who claims to manufacture products for professional or even industrial users should also meet this standard. Not through wordless refusal, but through critical discourse.

It is not my fault that Gigabyte is not taking this path. I continue to work, find out what others prefer to keep quiet about and publish it. In the end, however, the company is only punishing itself with a loss of image that cannot be compensated for by marketing platitudes or neologisms such as “Server Grade Gel”. Technology does not forgive PR tactics. It reacts to physics, chemistry and stress – and that’s where the problem begins. More updates to follow.

Gigabyte’s so-called “Server-Grade Thermal Conductive Gel” and a degrading thermal paste on the Radeon RX 9070 Gaming OC