▲ 423 ▼ DeepSeek ditches Nvidia for Huawei chips in V4 launch (cybernews.com) submitted 6 months ago by inari@piefed.zip to c/technology@lemmy.world 89 comments fedilink hide all child comments
[–] ag10n@lemmy.world 2 points 6 months ago* (last edited 6 months ago) (7 children) https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro permalink fedilink source parent hideshow 7 child comments replies: [–] theunknownmuncher@lemmy.world 2 points 6 months ago* (last edited 6 months ago) (6 children) You've proved my point that you don't know what you're talking about by blindly linking to the git repo. Couldn't find any source that supports your claim? I wonder why. Sure you can serve one request at a time to one patient user at a slow token per second rate, which makes running locally viable, but there is no RAM that has the bandwidth to run this model at scale. Even flash would be incredibly slow on CPU with multiple requests. You'd need the high bandwidth of VRAM and to run across multiple GPUs in a scalable way, it requires extremely high bandwidth interconnects between GPUs. permalink fedilink source parent hideshow 6 child comments replies: [–] ag10n@lemmy.world -3 points 6 months ago (5 children) Thank you for proving my point. It can be run on a cpu “It’s slow, it’s inefficient” it still runs It’s a foundational model just like R1 was. permalink fedilink source parent hideshow 5 child comments replies: [–] theunknownmuncher@lemmy.world 3 points 6 months ago* (4 children) Yes, you can run it at scale. at scale Shift those goalposts! We went from "at scale" to "it still runs" permalink fedilink source parent hideshow 4 child comments replies: [–] ag10n@lemmy.world 2 points 6 months ago (3 children) Quote me in full. You can run it at scale, on huawei. You can also run it on a cpu permalink fedilink source parent hideshow 3 child comments replies: [–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] theunknownmuncher@lemmy.world 2 points 6 months ago* (last edited 6 months ago) (6 children) You've proved my point that you don't know what you're talking about by blindly linking to the git repo. Couldn't find any source that supports your claim? I wonder why. Sure you can serve one request at a time to one patient user at a slow token per second rate, which makes running locally viable, but there is no RAM that has the bandwidth to run this model at scale. Even flash would be incredibly slow on CPU with multiple requests. You'd need the high bandwidth of VRAM and to run across multiple GPUs in a scalable way, it requires extremely high bandwidth interconnects between GPUs. permalink fedilink source parent hideshow 6 child comments replies: [–] ag10n@lemmy.world -3 points 6 months ago (5 children) Thank you for proving my point. It can be run on a cpu “It’s slow, it’s inefficient” it still runs It’s a foundational model just like R1 was. permalink fedilink source parent hideshow 5 child comments replies: [–] theunknownmuncher@lemmy.world 3 points 6 months ago* (4 children) Yes, you can run it at scale. at scale Shift those goalposts! We went from "at scale" to "it still runs" permalink fedilink source parent hideshow 4 child comments replies: [–] ag10n@lemmy.world 2 points 6 months ago (3 children) Quote me in full. You can run it at scale, on huawei. You can also run it on a cpu permalink fedilink source parent hideshow 3 child comments replies: [–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] ag10n@lemmy.world -3 points 6 months ago (5 children) Thank you for proving my point. It can be run on a cpu “It’s slow, it’s inefficient” it still runs It’s a foundational model just like R1 was. permalink fedilink source parent hideshow 5 child comments replies: [–] theunknownmuncher@lemmy.world 3 points 6 months ago* (4 children) Yes, you can run it at scale. at scale Shift those goalposts! We went from "at scale" to "it still runs" permalink fedilink source parent hideshow 4 child comments replies: [–] ag10n@lemmy.world 2 points 6 months ago (3 children) Quote me in full. You can run it at scale, on huawei. You can also run it on a cpu permalink fedilink source parent hideshow 3 child comments replies: [–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] theunknownmuncher@lemmy.world 3 points 6 months ago* (4 children) Yes, you can run it at scale. at scale Shift those goalposts! We went from "at scale" to "it still runs" permalink fedilink source parent hideshow 4 child comments replies: [–] ag10n@lemmy.world 2 points 6 months ago (3 children) Quote me in full. You can run it at scale, on huawei. You can also run it on a cpu permalink fedilink source parent hideshow 3 child comments replies: [–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] ag10n@lemmy.world 2 points 6 months ago (3 children) Quote me in full. You can run it at scale, on huawei. You can also run it on a cpu permalink fedilink source parent hideshow 3 child comments replies: [–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] theunknownmuncher@lemmy.world -1 points 6 months ago* (2 children) Quote me in full. Okay! You can run at scale, on huawei. You can also run it on a cpu Yeah, that is absolutely not what you argued. Anyway, you've conceded that I'm correct that you cannot run it at scale on a CPU, because running on CPU is too slow and inefficient, and that they instead use GPU hardware like Huawei GPUs to run the model at scale. That's good enough for me! permalink fedilink source parent hideshow 2 child comments replies: [–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent [–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent
[–] Diurnambule@jlai.lu 2 points 6 months ago Okey, then priced to just screenshot the part after the initial argument. Dude do more efforts. permalink fedilink source parent
[–] ag10n@lemmy.world 2 points 6 months ago Your interpretation of the English language has won you an argument! Huzzah So good of you to concede it runs on cpu permalink fedilink source parent