Don't copy core counts into the other vendor's field when merging GPUs

The merged duplicates stored AMD stream processors as cuda_cores and NVIDIA
CUDA cores as stream_processors. Every keeper already had the count in its
own field, so the copies are dropped.

Author: Seungpyo1007

data/gpu/amd/2021/enterprise/instinct-mi250x.json

@@ -5,7 +5,7 @@
   "architecture": "CDNA 2",
   "release_date": "2021-11-08",
   "msrp_usd": 12000,
-  "cuda_cores": 14080,
+  "cuda_cores": null,
   "stream_processors": 14080,
   "rt_cores": null,
   "tensor_cores": null,

data/gpu/amd/2023/enterprise/instinct-mi300x.json

@@ -5,7 +5,7 @@
   "architecture": "CDNA 3",
   "release_date": "2023-12-06",
   "msrp_usd": 15000,
-  "cuda_cores": 19456,
+  "cuda_cores": null,
   "stream_processors": 19456,
   "rt_cores": null,
   "tensor_cores": null,

data/gpu/amd/2025/consumer/radeon-rx-9060-xt-8gb.json

@@ -5,7 +5,7 @@
   "architecture": "RDNA 4",
   "release_date": "2025-06-05",
   "msrp_usd": 299,
-  "cuda_cores": 2048,
+  "cuda_cores": null,
   "stream_processors": 2048,
   "rt_cores": 32,
   "tensor_cores": 64,

data/gpu/amd/2025/enterprise/instinct-mi350x.json

@@ -5,7 +5,7 @@
   "architecture": "CDNA 4",
   "release_date": "2025-06-12",
   "msrp_usd": 25000,
-  "cuda_cores": 16384,
+  "cuda_cores": null,
   "stream_processors": 16384,
   "rt_cores": null,
   "tensor_cores": null,

data/gpu/amd/2025/enterprise/instinct-mi355x.json

@@ -5,7 +5,7 @@
   "architecture": "CDNA 4",
   "release_date": "2025-10-01",
   "msrp_usd": 30000,
-  "cuda_cores": 16384,
+  "cuda_cores": null,
   "stream_processors": 16384,
   "rt_cores": null,
   "tensor_cores": null,

data/gpu/nvidia/2016/consumer/geforce-gtx-1060-3gb.json

@@ -6,7 +6,7 @@
   "release_date": "2016-08-18",
   "msrp_usd": 199,
   "cuda_cores": 1152,
-  "stream_processors": 1152,
+  "stream_processors": null,
   "rt_cores": null,
   "tensor_cores": null,
   "memory_gb": 3,

data/gpu/nvidia/2016/consumer/geforce-gtx-1060-6gb.json

@@ -6,7 +6,7 @@
   "release_date": "2016-07-19",
   "msrp_usd": 249,
   "cuda_cores": 1280,
-  "stream_processors": 1280,
+  "stream_processors": null,
   "rt_cores": null,
   "tensor_cores": null,
   "memory_gb": 6,

data/gpu/nvidia/2025/consumer/geforce-rtx-5060-laptop.json

@@ -6,7 +6,7 @@
   "release_date": "2025-05-01",
   "msrp_usd": 1099,
   "cuda_cores": 3328,
-  "stream_processors": 3328,
+  "stream_processors": null,
   "rt_cores": 26,
   "tensor_cores": 104,
   "memory_gb": 8,

data/gpu/nvidia/2025/consumer/geforce-rtx-5060-ti-8gb.json

@@ -6,7 +6,7 @@
   "release_date": "2025-04-16",
   "msrp_usd": 379,
   "cuda_cores": 4608,
-  "stream_processors": 4608,
+  "stream_processors": null,
   "rt_cores": 36,
   "tensor_cores": 144,
   "memory_gb": 8,

data/gpu/nvidia/2025/consumer/geforce-rtx-5070-laptop.json

@@ -6,7 +6,7 @@
   "release_date": "2025-04-01",
   "msrp_usd": 1299,
   "cuda_cores": 4608,
-  "stream_processors": 4608,
+  "stream_processors": null,
   "rt_cores": 36,
   "tensor_cores": 144,
   "memory_gb": 8,