Nvidia Quietly Rolls Out GeForce RTX 4090 With New Die

News
By
published

But what does this mean?

AD102
(Image credit: cavitysearch123/Reddit)

Nvidia has quietly started using a new version of its AD102 graphics processing unit for its GeForce RTX 4090 Founders Edition graphics cards, according to a Reddit post (via VideoCardz). The new version is marked as the AD102-301, and it is believed that it will have a minor impact on the bill-of-materials costs of the graphics boards, though the information is unofficial.

"I received one of the new batches of the [GeForce RTX 4090] Founders Edition Cards and discovered it is a new production GPU die AD102-301-A1," wrote cavitysearch123 in a Reddit post. "[The] board also has a new designation 16F4. There is no news about this, only [GeForce RTX] 4080-series receiving a new [AD103-301] revision."

One of the reasons why the Reddit user started to research which GPU version and revision they have is because the AD102-301 comes with a limited maximum voltage of 1.070V, whereas the AD102-300 supports a maximum voltage of 1.1V. While the difference does not seem to be significant from a regular user point of view, it is important from overclocking perspective. Yet, the user managed to push his GPU all the way to 3 GHz with liquid cooling, which makes the GeForce RTX 4090 — which is the best graphics card available today — even faster.

"I can get 3015mhz stable at 1.070V, but having the extra headroom is always welcome for stability," the user said. "Although I am not sure in the stock configuration, this is under a waterblock. I would guess worse performance [with stock cooling] since temperatures would be higher."

Since the printed circuit board of this AD102-301-based GeForce RTX 4090 FE card has a different ID than a PCB of an AD102-300-powered RTX 4090 FE product, it is impossible to flash it with the original firmware with higher maximum voltage.

Earlier this year Nvidia quietly started to use AD103-301 and AD104-251 chips in graphics processors for GeForce RTX 4080 and GeForce RTX 4070 add-in-boards, respectively. As reported, these GPUs integrate a comparator that ensures that fans rotate at the right speed by comparing the fans' PWM signal with actual values. Previous configurations — AD103-300 and AD104-250 — had their comparator circuit disabled and this required makers of actual graphics cards to add a comparator circuit to their cards.

The difference between the AD102-300 and AD102-301 is unknown, but if the latter integrates the aforementioned circuitry, this will have an impact on GeForce RTX 4090 BOM cost. Meanwhile, keeping in mind that Nvidia's GeForce RTX 4090 Founders Edition starts at $1,599, a slightly lower BOM would hardly have any impact on its retail price. Yet, if the change between the AD102-300 and AD102-301 is not a working comparator but rather something else, then we could only guess how this impacts the performance or costs of GeForce RTX 4090 boards.

Anton Shilov
Anton Shilov
Contributing Writer

Anton Shilov is a contributing writer at Tom’s Hardware. Over the past couple of decades, he has covered everything from CPUs and GPUs to supercomputers and from modern process technologies and latest fab tools to high-tech industry trends.

See more GPUs News
15 Comments Comment from the forums
  • InvalidError
    I doubt Nvida would spend the millions of dollars needed to re-tapeout an IC only to integrate ~$1 worth of external circuitry. There must be more to it. Having a new chip and board ID to drop max Vcore by 30mV seems excessive, unless Nvidia expects RMA liabilities to exceed costs or has other things wrapped into that presumed re-spin too.
    Reply
  • JamesSneed
    InvalidError said:
    I doubt Nvida would spend the millions of dollars needed to re-tapeout an IC only to integrate ~$1 worth of external circuitry. There must be more to it. Having a new chip and board ID to drop max Vcore by 30mV seems excessive, unless Nvidia expects RMA liabilities to exceed costs or has other things wrapped into that presumed re-spin too.

    Exactly this. They found some kind of defect, vulnerability, or something major. If it was only the BOM they would simply have done a super refresh with said changes.
    Reply
  • InvalidError
    JamesSneed said:
    Exactly this. They found some kind of defect, vulnerability, or something major. If it was only the BOM they would simply have done a super refresh with said changes.
    Doesn't have to be something 'major' as that would definitely have to make it into notices somewhere. Could be more benign stuff like improved yields, such as the new chips being able to hit the same performance with 30mV less Vcore which could pave the way for a 4090 Ti/Super with the old 1.1V Vcore and 10% higher stock clocks.
    Reply
  • ManDaddio
    My opinion: I highly doubt it was because of a defect. There's revisions made all the time to gpus. AMD does the same thing. Imagine doing years of testing of a GPU (not a simple device) and then you release it with a defect.
    And don't bring up memory and capacitors. Sometimes that's just on the other end with bad quality control.
    The GPU itself would be working most likely near perfection before they would consider releasing it to the public.
    And we're talking about a company that's dedicated to video graphics. It is their specialty. It's unlikely they're going to have a defect. At least to make a change like this.
    It could just simply mean they're using different parts now from different manufacturers and had to make some slight changes to accommodate.
    It's funny how people just default to "it must be a defect".
    Only an obvious AMD shill would say that. But if I'm wrong then I'm not sure what kind of mind you have.
    Simply saying something without any kind of verification is kind of misinformation.
    Reply
  • InvalidError
    ManDaddio said:
    And we're talking about a company that's dedicated to video graphics. It is their specialty. It's unlikely they're going to have a defect.
    Chips launch with defects all of the time, many of which don't get in-silicon fixes until generations later if ever, which is why we get firmware, driver, OS, application and compiler-level work-arounds.
    Reply
  • Friesiansam
    Whatever the reason for the revision, the high cost of 4090 means that, for the overwhelming majority of GPU buyers, the change is irrelevant.
    Reply
  • -Fran-
    Friesiansam said:
    Whatever the reason for the revision, the high cost of 4090 means that, for the overwhelming majority of GPU buyers, the change is irrelevant.
    I was going to comment something very in-line with yours, haha. Like... "Look, the beach has one less grain on sand in it!".

    And adding it to the above posts/comments as well, nVidia is definitely being sus here. Not that they have any obligation to say/disclose anything, but it would still be interesting to know more. Maybe someone can check the naked dies and compare if there's more changes than just power regulation?

    I'll try my best to be optimistic and just think of the few bucks people in the market for a 4090 will get to save. Like for a cup of coffee or something, right?

    Regards.
    Reply
  • InvalidError
    -Fran- said:
    more. Maybe someone can check the naked dies and compare if there's more changes than just power regulation?
    If you don't mind sacrificing two GPUs for science to strip their dies down to signal layers.

    The other changes may be too subtle to find unless you know exactly where and what to look for. At the lowest end of the spectrum, it could simply be because TSMC published new primitives libraries and Nvidia decided the gains were worth a refresh.
    Reply
  • btmedic04
    ManDaddio said:
    My opinion: I highly doubt it was because of a defect. There's revisions made all the time to gpus. AMD does the same thing. Imagine doing years of testing of a GPU (not a simple device) and then you release it with a defect.
    And don't bring up memory and capacitors. Sometimes that's just on the other end with bad quality control.
    The GPU itself would be working most likely near perfection before they would consider releasing it to the public.
    And we're talking about a company that's dedicated to video graphics. It is their specialty. It's unlikely they're going to have a defect. At least to make a change like this.
    It could just simply mean they're using different parts now from different manufacturers and had to make some slight changes to accommodate.
    It's funny how people just default to "it must be a defect".
    Only an obvious AMD shill would say that. But if I'm wrong then I'm not sure what kind of mind you have.
    Simply saying something without any kind of verification is kind of misinformation.

    Intel is in the business of making cpu's, and it's obvious you weren't in the industry or around when the pentium FDIV bug happened. https://en.m.wikipedia.org/wiki/Pentium_FDIV_bug
    Reply
  • digitalgriffin
    I found this most curious.

    A comparitor circuit is really cheap. Couple bucks at most. A revision on a tape out mask...easily tens of thousands on something this complex. I've seen respin tapeouts run over $1,000,000 before.
    Reply