Details on Nvidia’s Next-Gen Blackwell GPUs Appear to Have Leaked

Nvidia GeForce RTX 4070 Founders Edition
(Image credit: Tom's Hardware)

Nvidia's codenamed Blackwell family of graphics processors will contain five different GPUs and will lack a direct successor to Nvidia's highly-successful AD104 chip, according to leaks by Chiphell and kopite7kimi (via VideoCardz). The information is unofficial and may be inaccurate, but if true, Nvidia will have to address market segments currently addressed by AD104 with two different GPUs.

Apparently, Nvidia's Blackwell family of graphics processors contains five chips codenamed GB202, GB203, GB205, GB206, and GB207. Nvidia's Ada Lovelace family released to date also contains five processors (AD102, AD103, AD104, AD106, AD107), just like the company's Ampere lineup of GPUs (GA102, GA103, GA104, GA106, GA107) that still powers some of the best graphics cards. Meanwhile, Nvidia's Turing family comprised of three members, whereas the Pascal lineup contained five GPUs. 

It is unclear why Nvidia's Blackwell family is said to feature GB200-series GPUs, but not GB100-series graphics processors. Typically, 200-series represents re-spinned GPUs.

Historically, Nvidia's XXX04 served performance-mainstream segment of the market and contained 50% - 66% transistors of the top-of-the-range part. The gap between the high-end and the performance mainstream part was quite noticeable. To fill the gap between its GA102 and GA104 GPUs in the Ampere era, Nvidia introduced GA103 part and did the same with the Ada Lovelace family: there is the AD103 sitting between the AD102 and AD104.  

While the now Nvidia's lineup no longer has wide gaps, a cut-down AD103 would overlap with the full AD104, which means that Nvidia has to either throw away AD103 GPUs that have one or more defective streaming multiprocessors or even CUDA cores so as not to compete against AD104, or keep them and then use them quietly to substitute AD104, which means cutting them down substantially and not using the whole potential of AD103.

Apparently, the company wants to avoid such a situation in the future. As a result, its GB202 will keep addressing the highest-end of the market (e.g., GeForce RTX 5090, GeForce RTX 5090 Ti), its GB203 will address high-end and performance-mainstream segments (e.g., GeForce RTX 5080, GeForce RTX 5070 Ti, and GeForce RTX 5070), while GB105 will address mainstream part of the market (GeForce RTX 5060 Ti, GeForce RTX 5060). This will enable Nvidia to use all GB203 silicon that it has even with some defective SMs. Of course, we are speculating here.

Nvidia's next-generation Blackwell GPUs are expected to hit the market in late 2024 or early 2025, so the company's plans may change a lot between now and then. Therefore, take this information with a grain of salt for now.

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.

  • pixelpusher220
    I know it's common place but it's really sad they will physically disable existing already built functionality to sell it at a lower price.
    Reply
  • edzieba
    pixelpusher220 said:
    I know it's common place but it's really sad they will physically disable existing already built functionality to sell it at a lower price.
    Welcome to every binned die for the last... several decades. Per-core CPU binning has occurred since two CPU cores were placed on one die. Most of the time it's because the disabled area has a defect, but sometimes working cores are disabled in order to ensure a supply of a desired chip bin (e.g. the classic Athlon/Duron 'pencil trick' to re-enable disabled-but-good cores, or sometimes cause the machine to fail to POST due to attempting to enable a dead disabled core).
    Reply
  • DSzymborski
    pixelpusher220 said:
    I know it's common place but it's really sad they will physically disable existing already built functionality to sell it at a lower price.

    It's mostly because of binning that these products exist at these prices.
    Reply
  • DonQuixoteIII
    Either really, really crap editing, or new terms that I don't understand. Or, AI made article?? What is "re-spinned"... Actually, whole thing seems a bit 'off'.
    Reply
  • AgentBirdnest
    Admin said:
    will lack a direct successor to Nvidia's highly-successful AD104 chip
    Ummm... since when?
    DonQuixoteIII said:
    Either really, really crap editing, or new terms that I don't understand. Or, AI made article?? What is "re-spinned"... Actually, whole thing seems a bit 'off'.
    It's crap. It makes little sense. There is WAY too much speculation, based on such little info from a leak.
    But the strangest part is that bit about having to throw out AD103s if they have a single defective CUDA core. The author doesn't seem to have a full grasp of binning. The RTX 4080 already IS a cut-down AD103 die. It uses 76 of the AD103's 80 SMs (9,728 out of 10,240 CUDA cores.) If a single SM or CUDA is defective, they can use a different one that would have been fused off anyway.

    This is weird.
    Reply
  • DaveLTX
    DonQuixoteIII said:
    Either really, really crap editing, or new terms that I don't understand. Or, AI made article?? What is "re-spinned"... Actually, whole thing seems a bit 'off'.
    Nevermind that GTX900 series is made up of GM2xx cards (and GTX700 series being made up of GK2xx cards which was hardly a respin)
    Which is a optimization of the original Maxwell architecture
    Wow tomshw is certainly getting worse
    Reply
  • doomtomb
    DonQuixoteIII said:
    Either really, really crap editing, or new terms that I don't understand. Or, AI made article?? What is "re-spinned"... Actually, whole thing seems a bit 'off'.
    re-spins are basically just revisions to the silicon. In this case they mean the photo masks, which is to say the circuit layout. Since these haven't been manufactured yet (to our knowledge), re-spinning is unusual. Also, it means architecturally speaking, the devices would be similar to the ones before it (100 series) but maybe include bug fixes, enhancements, or performance improvements.
    Reply
  • Lucky_SLS
    I still remember when people were trying to flash RX 480/580 bios on RX 470/570 to try and unlock any locked down cores. and in rare cases it worked. People were sharing the lot or S/N of the GPUs in reddit to gain an ez boost...
    Reply
  • DaveLTX
    Lucky_SLS said:
    I still remember when people were trying to flash RX 480/580 bios on RX 470/570 to try and unlock any locked down cores. and in rare cases it worked. People were sharing the lot or S/N of the GPUs in reddit to gain an ez boost...
    That was a tradition that went back to at least HD6950
    Often times they simply weren't lasered at all (common theme with AMD going back to phenom 2/athlon 2 three cores being unlocked to quads, even the later quads were unlocked to hexa cores as well) and repurposing cast off bins therefore you could easily unlock them

    Ever since AMD got their act together they have been putting programmable fuses properly, before that they were more unintentional to end up with poorer bins rather than their aim to create more bins
    Reply